Quantitative iTRAQ LC-MS/MS Proteomics Reveals the Proteome Profiles of DF-1 Cells after Infection with Subgroup J Avian Leukosis Virus

Avian leukosis virus subgroup J (ALV-J) is an avian oncogenic retrovirus that can induce various clinical tumors and has caused severe economic losses in China. To improve our understanding of the host cellular responses to virus infection and the pathogenesis of ALV-J infection, we applied isobaric tags for relative and absolute quantification (iTRAQ) labeling coupled with multidimensional liquid chromatography-tandem mass spectrometry to detect the protein changes in DF-1 cells infected and mock-infected with ALV-J. A total of 75 cellular proteins were significantly changed, including 33 upregulated proteins and 42 downregulated proteins. The reliability of iTRAQ-LC MS/MS was confirmed via real-time PCR. Most of these proteins were related to the physiological functions of metabolic processes, biosynthetic processes, responses to stimuli, protein binding, signal transduction, cell cytoskeleton, and so forth. We also found some proteins that play important roles in apoptosis and oncogenicity. The differentially expressed proteins identified may provide valuable information to elucidate the pathogenesis of virus infection and virus-host interactions.


Introduction
The J subgroup of avian leukosis virus (ALV-J), which belongs to the Retroviridae family, was first isolated from whitemeat-type chickens in the United Kingdom in 1988 [1]. It can predominantly lead to myeloid leukosis (ML) and immunosuppression effects in both naturally and experimentally infected chickens [2,3]. In China, ALV-J-associated myeloid leukosis in chickens was first reported in 1999 [4]. ALV-J can induce various tumors, growth retardation, and production problems. In addition, in recent years, it has become widespread in many parts of our country and leads to severe economic losses in the poultry industry.
The pathogenesis of virus infection and the mechanism through which the virus interacts with host cells remain unclear. During virus infection, the proteins of host cells may be significantly changed. It is now possible to use proteomic techniques to identify the changes in protein abundance that indicate host cellular responses to virus infection and provide useful information to obtain a better understanding of the pathogenesis of virus infection [5][6][7][8]. Kvaratskhelia et al. [9] applied enzymatic digestion coupled with mass spectrometry (MS) to detect the sites of glycosylation on the surface of avian leukosis virus subgroup A (ALV-A) and found that carbohydrates may play an important role in receptor binding.
To explore the possible mechanisms of virus infection, we used isobaric tags for relative and absolute quantification (iTRAQ) combined with multidimensional liquid chromatography (LC) and tandem MS analysis to perform a 2 BioMed Research International quantitative proteomic analysis of DF-1 cells infected with ALV-J [10]. To the best of our knowledge, no previous study had used the iTRAQ LC-MS/MS proteomics strategy to investigate the differently expressed proteins in ALV-Jinfected DF-1 cells. The iTRAQ labeling technology could greatly increase the identification sensitivity and quantitation accuracy of proteomic analyses through a multiplexed quantitation strategy [11]. The results showed that 75 proteins were significantly changed after ALV-J infection. These changed proteins may provide valuable information to study the molecular mechanisms underlying ALV-J pathogenesis. supplemented with 10% fetal bovine serum (FBS) and 100 g/mL streptomycin and penicillin at 37 ∘ C in a 5% CO 2 atmosphere. ALV-J strain HPRS-103 (GenBank: Z46390) was kindly provided by Professor Venugopal Nair. DF-1 cells cultured in flasks to approximately 80% confluence were infected with 0.5 mL of 10 3.5 /mL 50% tissue culture infectious doses (TCID50) of ALV-J for 144 h. Uninfected DF-1 cells served as mock-infected cells.

Indirect Immune Fluorescence Assay (IFA).
At 144 h after infection, the infected DF-1 cells were washed twice with PBS and fixed with anhydrous ethanol for 20 min. The fixed cells were then incubated with mouse anti-P27 monoclonal antibody (prepared in our lab) at 37 ∘ C for 60 min. After washing three times with PBST (0.01 M PBS, pH 7.2, 0.05% Tween 20), the cells were incubated with goat anti-mouse IgG conjugated to FITC (Sigma, USA) at 37 ∘ C for another 60 min. Finally, the cells were observed under a Carl Zeiss Vision microscope (ZEISS Axio Observer D1) after three washes with PBST.

Protein Extraction, Digestion, and Labeling with iTRAQ
Reagents. Infected and mock-infected DF-1 cells were washed twice with PBS. The cells were lysed in a lysis buffer (9 M urea, 4% CHAPS, 1% DTT, and 1% IPG buffer). The mixtures were centrifuged at 15,000 g and 4 ∘ C for 15 min. The supernatant was collected, and the protein concentration was determined using the Bradford protein assay [12] (Bio-Rad Laboratories). Then, 100 g of protein was mixed overnight with four volumes of cold (−20 ∘ C) acetone and then dissolved using the dissolution buffer. After being reduced, alkylated, and digested with trypsin, the samples were labeled following the manufacturer's instructions Table 1   T i m e  0  2 4  3 0  3 1  3 8  B %  3  1 6 3 0 9 0 9 0 described in the iTRAQ protocol. The labeled samples were pooled for further analysis. The data were acquired at 38 min. The spray voltage was 2.0 KV, the capillary temperature was 320 ∘ C, the collision energy was 30, and the acquisition quality range was 300-1400 da.

LC-MS/MS and Database
The relative quantification and protein identification were performed with the Protein Discoverer software (version 1.2) using the built-in mascot as the search engine.
2.6. Real-Time PCR. The primers ( Table 2) were synthesized by BoShi Biotechnology Company (Harbin, China). The gene was amplified from the genomic DNA of DF-1 cells by polymerase chain reaction (PCR). The PCR-amplified products were separated in a 2% agarose gel and then purified using a DNA gel extraction kit (Axygen Biotechnology Limited, Hangzhou City, China). The products were then ligated into the pZeroBack/blunt vector (Tiangen Biotech Co., Ltd., Beijing, China), and the sequence was verified. The plasmid DNA was used as the standard to construct the standard curve via SYBR Green real-time PCR. The total cellular RNA of the infected or mock-infected DF-1 cells was extracted using the RNeasy Mini Kit (QIAGEN, China)

Bioinformatics Analysis.
The functional annotation of the 75 proteins in DF-1 cells that were significantly changed after infection with ALV-J was performed using the GOSlimViewer tool of the AgBase database (http://www.agbase.msstate.edu/) [13]. In addition, we aimed to determine how ALV-J interacts with the host cellular proteins and how it affects the function of host cells. The identified proteins were inputted into the STRING database to obtain the protein-protein interaction network [14,15] (http://string.embl.de/).

Confirmation of ALV-J Infection in DF-1 Cells by IFA.
To confirm that the DF-1 cells were infected by ALV-J, IFA was used to detect the viral P27 antigen. The results showed clear green fluorescence in ALV-J-infected DF-1 cells 144 h after infection, whereas the uninfected DF-1 cells exhibited no green fluorescence ( Figure 1).

Protein Profile Obtained by iTRAQ LC-MS/MS Analysis.
To explore the differences in the protein expression levels after virus infection, the total proteins of ALV-J-infected and mock-infected DF-1 cells were extracted for iTRAQ-LC-MS/MS analysis. A total of 1091 proteins were detected, including 75 proteins in DF-1 cells that were significantly changed infection with ALV-J for 144 h (Table 3). These differently expressed proteins were divided into two clusters: upregulated and downregulated. The number of upregulated proteins was 33, whereas the number of downregulated proteins was 42.

Functional Classifications of the Identified Proteins.
To annotate the functions of the 75 significantly changed proteins identified in our study, the proteins were submitted to GORetriever (http://www.agbase.msstate.edu/) for analysis. Three types of annotations were obtained using the website: molecular functions, biological processes, and cellular components.

Validation of the iTRAQ Data by Real-Time PCR.
To confirm the results of the differentially expressed proteins identified by iTRAQ LC-MS/MS analysis, real-time PCR was performed to detect the transcript expression levels of the genes after ALV-J infection. We generated four standard curves to determine the gene expression of BLOC1S5, keratin, HMG14, and AACS in ALV-J-infected and mock-infected DF-1 cells. The results showed that HMG14 was upregulated (Figure 3), whereas BLOC1S5, AACS, and keratin were downregulated ( Figure 3). The RT-PCR results were consistent with the results of the iTRAQ LC-MS/MS analysis (Table 3), confirming that the iTRAQ data were reliable.

Protein-Protein Interaction Analysis.
The mechanism through which the virus interacts with host cells remains unclear, and oncogenicity is an important index of the pathogenicity of ALV-J. During virus infection, some proteins of host cells may be significantly changed. As a result, the functions of the changed proteins will also be altered. In our study, we aimed to determine whether the significantly changed proteins that were identified have some relationship with apoptosis or ALV-J-induced oncogenicity. We searched the STRING database to analyze the protein-protein interactions between the differently expressed proteins and PARK7, PTENP1, AKT1, PIK3CA (PI3K), and VDAC ( Figure 4). These proteins are known to have some relationship with tumor-associated process and apoptosis. The protein-protein interaction networks may provide valuable information to further investigate the possible mechanism of ALV-J-induced oncogenicity.

Discussion
Proteomics is a relatively novel technology that has been used for the detection of the host cellular proteins response to virus infection [16,17]. Isobaric tags for relative and absolute quantification (iTRAQ) combined with multidimensional liquid chromatography (LC) and tandem MS analysis are a powerful tool for quantitative proteomic analysis that has   been widely applied in many studies [18][19][20]. In this study, we first applied the iTRAQ approach to identify the differential protein expression profiles of DF-1 cells infected with ALV-J. Using the iTRAQ LC-MS/MS technology, the significantly changed proteins were mostly associated with metabolic process, signal transducer activity, cell cytoskeleton, oxidoreductase activity, response to stimulus, and immune responses.
In addition, some apoptosis and tumor-associated proteins (VEGF-A, ACTN4, and METAP2) were also identified by the iTRAQ LC-MS/MS technology.

Alterations of Tumor-Associated Proteins.
Vascular endothelial growth factor A (VEGF-A) is an important inducer of angiogenesis [21]. As has been shown in many reports, upregulated VEGF-A can induce tumor formation via some unique signaling pathways [22,23]. In addition, VEGF-A, which is known as a positive regulator, contributes to tumor growth and promotes tumor formation [24,25]. Previous studies described a threshold level of proteins to promote tumorigenesis, which indicated that the expression level of one protein needs to reach the threshold level before promoting tumorigenesis [26,27]. Studies in our lab showed that the increased replication of ALV-J increased the expression of VEGF-A, indicating an increased opportunity for ALV-J to push the expression level of VEGF-A to reach the threshold level to promote tumorigenesis [27].
In this study, we found that VEGF-A is overexpressed in DF-1 cells after infection with ALV-J. The results further suggested that VEGF-A is closely associated with ALV-J-induced tumorigenesis and may also suggest a novel molecular mechanism for better understanding of the higher oncogenicity of ALV-J. Alpha-actinins (ACTNs) were classified into cytoskeleton proteins, while ACTN4 has some other unique functions, such as signal transduction, protein expression regulation, and nuclear transport. Histological analyses of cancer tissues showed a strong correlation between ACTN4 expression and tumorigenesis in several types of cancers [28][29][30]. Furthermore, upregulated ACTN4 in cancer cells has been suggested as a biomarker for drug resistance and malignant cell invasion [31][32][33][34]. Previous studies showed that ALV-J infection in DF-1 cells led to rapid increase in Akt phosphorylation and the phosphorylation of Akt was PI3K-dependent [35]. PI3K/Akt pathway also regulates viral replication of ALV-J [35]. Furthermore, AKT interacts with ACTN4 and ACTN4 is a functional partner of AKT [36]. Therefore, the upregulated ACTN4 observed in this study may be associated with tumorigenesis induced by ALV-J through PI3K/Akt pathway. This may provide useful information to elucidate the mechanism of ALV-J induced tumorigenesis and may also become potential therapeutic targets to control ALV-J infections.
METAP2 was considered to have some relationship with angiogenesis inhibition [37]. In addition, METAP2 can block B cell differentiation into plasma cells [38]. Some viruses, whose primary target cells are B cells, can clinically induce tumor formation. Therefore, downregulation of METAP2 in this study may influence the function of B cells, which may provide evidence to explain why ALV-J infection can result in immune suppression and tumorigenesis.

Redox
Regulation. Peroxiredoxins (PRDXs), a family of peroxidases as antioxidant enzymes, can support tumor maintenance and survival through protecting cells from apoptosis induced by oxidative stress [39][40][41]. A previous study indicated that liver cells transfected with PRDX6 siRNA resulted in an increase in peroxide-induced cytotoxicity by apoptosis, which implies that decrease of PRDX6 promotes apoptosis [42]. Therefore, downregulated PRDX6 in this study suggests that ALV-J infection may weaken the antiapoptotic function of PRDX6.
In addition, PRDX1 was found to be upregulated in this study. A previous study indicated that the mice lacking    PRDX1 have several malignant cancers, including sarcomas, carcinomas, and lymphomas [43]. These malignancies are associated with low expression of PRDX1, which suggests that PRDX1 may function as a tumor suppressor [43]. Studies also indicated that PRDX1 interacts with the c-Myc oncogene and can inhibit its transcriptional activity [44] and high expression of PRDX1 appears to be associated with less aggressive breast cancers [45]. Therefore, upregulation of PRDX1 in this study may result from the defense of host cells responses to the ALV-J infection.

Cytoskeleton Proteins and ALV-J Infection.
Cytoskeleton proteins are involved in the maintenance of cell morphology, regulation of protein synthesis, endocytosis, cell movement, and cell-to-cell attachment [46,47]. As determined through iTRAQ LC-MS/MS analysis, some cytoskeleton proteins were identified to be significantly changed in DF-1 cells after infection with ALV-J. Isoform 2 of the F-actin-capping protein subunit beta isoforms 1 and 2 (CAPZB) can regulate the growth of actin filaments, and actin filaments play a vital role in the maintenance of cell morphology [48]. Furthermore, actin-related protein 3 (ACTR3) and actin-related protein 5 (ACTR5) were also found to be changed. The low expression of these proteins revealed that the cytoskeletal proteins were disrupted during infection with ALV-J. In addition, the differential expression of these proteins may be due to the interaction between the virus and host cellular proteins after infection with ALV-J.
It has been reported that keratins have become the standard detection marker for tumor cells and were also the most common marker to identify tumor cells [49]. Previous studies showed that before tumor cells got the ability to migrate and invade the host, they need to undergo epithelialmesenchymal transition, during which the cytoskeletons are rearranged and epithelial markers, such as keratins, claudins, and E-cadherin, are observed to be downregulated [49][50][51][52]. Immunohistochemical analysis showed that low expression of keratin was associated with a higher tumor grade in breast cancer [52]. Previous study indicated that acetoacetyl-CoA synthetase (AACS) was found in tumor tissues and plays important roles in metabolic processes of tumors [53].
Whether or not the downregulated keratin and AACS in this study were associated with tumorigenesis induced by ALV-J infection needs to be further investigated.

Conclusions
In summary, our study was the first to use iTRAQ LC-MS/MS to detect cellular responses to ALV-J infection in DF-1 cells. A total of 75 significantly changed proteins were identified. These differently expressed proteins may provide useful information for elucidating the molecular mechanism underlying the interaction between ALV-J and DF-1 cells and will also facilitate our understanding of the pathogenesis of ALV-J infection.