*Article* **Metabolomics Analysis of PK-15 Cells with Pseudorabies Virus Infection Based on UHPLC-QE-MS**

**Panrao Liu <sup>1</sup> , Danhe Hu <sup>1</sup> , Lili Yuan <sup>1</sup> , Zhengmin Lian <sup>1</sup> , Xiaohui Yao <sup>1</sup> , Zhenbang Zhu <sup>1</sup> and Xiangdong Li 1,2,\***


**Abstract:** Viruses depend on the metabolic mechanisms of the host to support viral replication. We utilize an approach based on ultra-high-performance liquid chromatography/Q Exactive HF-X Hybrid Quadrupole-Orbitrap Mass (UHPLC-QE-MS) to analyze the metabolic changes in PK-15 cells induced by the infections of the pseudorabies virus (PRV) variant strain and Bartha K61 strain. Infections with PRV markedly changed lots of metabolites, when compared to the uninfected cell group. Additionally, most of the differentially expressed metabolites belonged to glycerophospholipid metabolism, sphingolipid metabolism, purine metabolism, and pyrimidine metabolism. Lipid metabolites account for the highest proportion (around 35%). The results suggest that those alterations may be in favor of virion formation and genome amplification to promote PRV replication. Different PRV strains showed similar results. An understanding of PRV-induced metabolic reprogramming will provide valuable information for further studies on PRV pathogenesis and the development of antiviral therapy strategies.

**Keywords:** pseudorabies virus; metabolomic analysis; UHPLC-QE-MS; PK-15 cells

### **1. Introduction**

The pseudorabies virus (PRV) is a member of the *Herpesviridae* family, which causes Aujeszky's disease (AD) in pigs [1]. Pigs are the main host of PRV, and pigs of different ages can be infected with PRV. AD leads to high mortality and symptoms related to the central nervous system in piglets, respiratory disease in adult pigs, and decreased reproduction in sows, which have resulted in great economic losses for the pig industry [2]. PRV also infects other mammals, such as ruminants, carnivores, rodents, and even humans [3–5], posing a concern for public health. PRV virions are composed of double-stranded DNA genomes, capsids, teguments, and envelopes. The genome is approximately 150 kb and encodes over 100 proteins [1]. PRV was discovered in the 1900s and then widely distributed in the world. Although PRV has been eradicated in some Western countries (such as the United States, Germany, and Canada), it is still prevalent in many countries [6]. The Bartha K61 strain is one of the classically attenuated PRV strains, known as Bartha K61, which was isolated in 1961 and attenuated by a series of passages in embryos and chicken cells [7]. As an attenuated live vaccine, Bartha K61 can induce effective immune responses against PRV in pigs [8,9]. However, in 2011, PRV variant strains emerged in Northern China, then caused appalling outbreaks in swine farms, including in PRV-vaccinated swine farms [10,11]. Although much progress has been made in the research on the pathogenesis of PRV, the detailed mechanisms of the interaction between PRV and host cells remain unclear.

Metabolomics is a new subject developed after genomics and proteomics. Nowadays, metabolomics has been applied to disease diagnosis, pharmaceutical research and

**Citation:** Liu, P.; Hu, D.; Yuan, L.; Lian, Z.; Yao, X.; Zhu, Z.; Li, X. Metabolomics Analysis of PK-15 Cells with Pseudorabies Virus Infection Based on UHPLC-QE-MS. *Viruses* **2022**, *14*, 1158. https:// doi.org/10.3390/v14061158

Academic Editor: Douglas Gladue

Received: 8 May 2022 Accepted: 25 May 2022 Published: 27 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

development, nutrition science, environmental science, botany, and other fields closely related to human health [12,13]. Metabolomics focuses on low-molecular-weight metabolites (MW < 1 KD, such as sugars, lipids, amino acids, and vitamins) in various metabolic pathways, and it can reflect the changes in the metabolic response of cells or tissues to external stimulation or genetic modifications, which contribute to reveal the mechanism of interaction between host cells and external factors [14,15].

Viruses are intracellular parasites and cannot proliferate independently. They must hijack and rely on the metabolic mechanisms and resources of host cells for their own replication [16,17]. Virus infection remodels the metabolic machineries in host cells to deal with the higher metabolic demands during virus replication [18–20]. Zika virus infection increased the glucose utilization in the tricarboxylic acid (TCA) cycle in HFF-1 cells and elevated the AMP/ATP ratios, which led to cell death [21]. Influenza virus was shown to affect host metabolic pathways to ensure the production of viral particles. Glucose uptake and aerobic glycolysis were increased, while fatty acid β-oxidation were decreased in cells infected with influenza virus [22]. Human cytomegalovirus (HCMV) increased glycolytic flux to replenish the TCA cycle, and herpes simplex virus type-1 (HSV-1) induced the elevation of pyrimidine nucleotide components [23]. In addition, SARS-CoV-2 infection induced sphingolipid metabolism reprogramming, which was required for viral replication. The levels of glycosphingolipid and sphingolipid (sphingosine, GA1, and GM3) were markedly increased in cells and the murine model after SARS-CoV-2 infection [24]. However, a decrease in cholesterol and high- and low-density lipoproteins was induced in the blood of patients with COVID-19, which may be potential markers for monitoring the disease [25]. Therefore, metabolomics is widely used as an important tool to investigate complex virus–host interactions. In this way, these studies provide an insight into the pathogenic mechanisms and novel therapeutic methods of the virus.

The metabolic alterations induced by different viruses are distinct. PRV is one of the main pathogens of pigs, thus understanding the changes of PRV in host cell metabolism is necessary. Few studies have been performed on the host metabolism of PRV. Gou et al. established that the metabolic flux derived from glycolysis, the pentose phosphate pathway, and glutamine metabolism for nucleotide biosynthesis was necessary for PRV replication [26]. The changes of PRV infection in immortalized porcine alveolar macrophages (iPAMs) on glycerolipids, fatty acyls, glycerophospholipids, and sphingolipids have also been determined [27]. In this study, we analyze the metabolic alterations in porcine kidney cells (PK-15) infected with a PRV variant strain and Bartha K61 strain using ultra-highperformance liquid chromatography/Q Exactive HF-X Hybrid Quadrupole-Orbitrap Mass (UHPLC-QE-MS). The results show that plenty of metabolites and metabolic pathways are significantly changed during PRV infection, when compared to uninfected cell groups. It suggests that those alterations may be in favor of better viral replication. These findings may be helpful to understand the host response to PRV infection and development for this disease control.

### **2. Materials and Methods**

### *2.1. Cell Culture and Virus Infection*

PK-15 cells were purchased from the American Type Culture Collection (ATCC) and cultured in Dulbecco's modified Eagle medium (DMEM) (Gibco, Waltham, MA, USA) containing 10% fetal bovine serum (FBS, Thermo Fisher Scientific, Waltham, MA, USA) at 37 ◦C with 5% CO2. PRV variant strain JS21 (abbreviation of PRV-G) and PRV Bartha K61 strain (GenBank accession no. JF797217; with abbreviation of PRV-K) were preserved in our laboratory. PRV titers were determined as the median tissue culture infective doses (TCID50) on PK-15 cells.

### *2.2. Virus Infection*

PK-15 cells were cultured overnight at 37 ◦C with 5% CO2. When the cells' density reached approximately 80%, they were infected with PRV at a multiplicity of infection

(MOI) of 1 and incubated at 37 ◦C for 1 h. After washing with phosphate-buffered saline (PBS), the cells were incubated in DMEM supplemented with 2% FBS. PK-15 cell samples were harvested at 0, 6, 12, and 24 h post infection (h.p.i.).

### *2.3. Western Blot and Immunofluorescence Assay*

PK-15 cells were infected with PRV (PRV-G or PRV-K) at MOI of 1. Cell samples were harvested at 6, 12, and 24 h.p.i. for immunoblotting analysis, which was performed as previously described [28]. Briefly, cells were lysed with 200 µL lysis buffer (Beyotime, Shanghai, China) for 15 min on ice. Following centrifugation, the supernatant of the lysates was denatured. Then, the samples were subjected to SDS-PAGE and transferred to nitrocellulose membranes (Sigma-Aldrich, Whatman, MA, USA). Following the incubation of antibodies of anti-PRV gB protein mAb (1:1000, preserved in our laboratory) and anti-β-actin (1:1000, Cell Signaling Technology, Danvers, MA, USA), the bands were visualized using an enhanced chemiluminescence reagent kit (Share-bio, Shanghai, China) and analyzed using ImageJ software.

PK-15 cells were inoculated on coverslips in the 6-well plate and infected with PRV (PRV-G or PRV-K) at MOI of 1. Cell samples were harvested at 6, 12, and 24 h.p.i. for immunofluorescence assay (IFA), which was performed as previously described [28]. Finally, the slides were placed on the cover glass with antifade mounting medium and visualized using an LSM 880 Zeiss confocal microscope (Carl Zeiss, Jena, Germany).

### *2.4. Sample Preparation and Extraction and UHPLC-QE-MS Analysis*

PK-15 cells were infected with PRV (PRV-G or PRV-K) at MOI of 1. Cell samples were harvested at 0, 6, 12, and 24 h.p.i. There were approximately 1 <sup>×</sup> <sup>10</sup><sup>7</sup> cells per sample. The samples of 0 h were used as control, named Mock group. Other groups were named G6, G12, G24, K6, K12, and K24, respectively. Three replicates per group were set. Cells were washed with precooled PBS, and the supernatant was cleared by centrifugation for 10 min at 13,000 rpm at 4 ◦C. The cell pellet was frozen in liquid nitrogen for 30 s. Following freezedrying, the samples were dissolved in sterile water and ultrasound treatment in ice water. Following centrifugation at 12,000 rpm for 15 min at 4 ◦C, the supernatant was extracted with 1 mL of methanol/acetonitrile/water (2:2:1, *v*/*v*/*v*) containing isotope-labeled internal standard mixture. Following ultrasound treatment, the samples were incubated at −40 ◦C for 1 h, and centrifuged at 12,000 rpm at 4 ◦C for 15 min. The supernatant was used for the UHPLC-QE-MS analysis. All samples were obtained and mixed in equal amounts as quality control (QC) samples before testing. Then, experimental samples and QC samples were tested on the machine.

A UHPLC system (Vanquish, Thermo Fisher Scientific, Waltham, MA, USA) with a UPLC BEH Amide column (2.1 mm × 100 mm, 1.7 µm) coupled to a Q Exactive HFX mass spectrometer (Orbitrap MS, Thermo Fisher Scientific, Waltham, MA, USA) was used for LC-MS/MS analysis. Liquid chromatography phase A is an aqueous phase, containing 25 mmol/L ammonium acetate and 25 mmol/L ammonia water, and phase B is acetonitrile. The QE HFX mass spectrometer was used for acquiring full scan MS/MS spectra on information-dependent acquisition (IDA) mode under software (Xcalibur, Thermo Fisher Scientific, Waltham, MA, USA) control.

### *2.5. PCA and OPLS-DA Analyses*

The raw data were converted to the mzXML format using ProteoWizard software. Then, the data were processed by R package analysis for peak identification, extraction, alignment, and integration. Principal component analysis (PCA) and orthogonal projection to latent structures discriminant analysis (OPLS-DA) were conducted [29,30]. The data were logarithmically (LOG) transformed and centered (CTR) formatted using SIMCA software (V16.0.2, Sartorius Stedim Data Analytics AB, Umea, Sweden), followed by PCA modeling analysis. OPLS-DA modeling analysis was performed on the first principal component and a 7-fold cross-validation in the SIMCA software was performed throughout the analysis. The R2X or R2Y (interpretability of the model for the categorical variable) and Q2 (the predictability of the model) were used to evaluate the model validity.

### *2.6. Total RNA Extraction and Quantitative Real-Time PCR (qPCR) Analysis*

Cell samples were harvested at 24 h.p.i. after PRV (PRV-G or PRV-K) infection with different MOI; meanwhile, uninfected cells were used as control. Total RNAs from cell samples were extracted using TRNzol (TIANGEN, Beijing, China) and reverse transcribed to cDNA using a HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China), according to the manufacturer's instructions. The primer sequences of the target genes to be detected were designed, and are shown in Table S1. ACTB was used as an internal reference gene. qPCR was performed using Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) and an ABI QuantStudio 3 Real-Time PCR (96-Well) Detection System. The reaction parameters were: 95 ◦C, 30 s; 95 ◦C, 10 s, 60 ◦C, 30 s, 40 cycles.; 95 ◦C, 15 s, 60 ◦C, 60 s, 95 ◦C, 15 s. All experiments were performed in triplicate. The mRNA levels of genes were quantified relative to ACTB using the comparative threshold cycle (2-∆∆CT) method [31].

### *2.7. Statistical Analysis*

The first principal component of variable importance in the projection (VIP > 1) and Student's *t*-test (*p* < 0.05) was set as the standard to screen the differential metabolites. Then, the data was subjected to the KEGG Metabolome Database for identification of metabolites. Additionally, the metabolites were analyzed further via online statistical analysis (MetaboAnalyst, http://www.metaboanalyst.ca/ (accessed on 16 November 2021)) for identifying the altered metabolic pathways caused by PRV infection [32].

GraphPad Prism 7.0 software was used for the statistical analyses. *p*-values less than 0.05 were considered statistically significant. The values were expressed as the mean ± standard error of the mean. The significance in figures was indicated as follows: \*, *p* < 0.05; \*\*, *p* < 0.01.

### **3. Results**

### *3.1. Replication of PRV in PK-15 Cells*

To confirm PRV replication in PK-15 cells, the cells were infected with PRV-G or PRV-K strain at MOI = 1, respectively. Additionally, the expression levels of PRV-gB protein were determined by Western blot and IFA. In the uninfected cells, there was no signal of viral protein to be detected. As shown in Figure 1A,B, the gray values of PRV-gB protein notably increase over time. Additionally, the amount and the intensity of fluorescence in PRVinfected cells were progressively strong and reached a high level at 24 h.p.i. (Figure 1C,D). These results indicate that both the PRV-G and PRV-K strains could effectively replicate in PK-15 cells within a 24 h infection.

### *3.2. Multivariate Analysis of PK-15 Cell Metabolites*

The UHPLC-QE-MS used positive and negative ion (POS and NEG) switching modes and full-scan assay to screen and identify the numerous metabolites. After obtaining the data, we performed a series of multivariate pattern recognition analyses to evaluate the differences between the samples. PCA and OPLS-DA were performed to obtain more reliable information on the correlation between group differences of metabolites and experimental groups. The PCA-score scatter plot of all samples (including QC samples) is shown in Figure 2A,B. Each scatter represented a sample, and the color and shape of the scatter signed different groups. The results of the PCA score scatter plot show that all samples are in the 95% confidence interval. The OPLS-DA model for different groups versus the mock group was analyzed, and the R2X, R2Y, and Q2 of samples (POS, NEG) are shown in Figure 2C, in which the values of three parameters are close to 1. These results indicate that these different groups are clearly distinguished, and these models are efficient and reliable.

**Figure 1.** Infections of different PRV strains in PK-15 cells. (**A**,**B**) The cells were infected with PRV variant strain (abbreviation of PRV-G) and PRV Bartha K61 strain (abbreviation of PRV-K) at MOI = 1, and cell samples were collected at 6, 12, and 24 h for immunoblotting detection. (**C**,**D**) PK-15 cells were infected with PRV-G and PRV-K at MOI = 1 for 6, 12, and 24 h at 37 ◦C with 5% CO<sup>2</sup> . The expression levels of PRV-gB protein detected by IFA. Scale bars = 200 µm.

**Figure 2.** Score scatter plots of PCA and OPLS-DA of PRV-infected and uninfected cells. (**A**,**B**) Score scatter plot of the PCA model for the different infection groups versus mock group. Electrospray ionization served as the source of UHPLC-QE-MS, including positive and negative ion modes (POS and NEG). (**A**) was derived from POS and (**B**) from NEG. The lines denote 95% confidence interval Hotelling's ellipses. (**C**) OPLS-DA model for the different PRV-strain infection group versus mock group.

### *3.3. Differentially Expressed Metabolites during PRV Infection*

Based on OPLS-DA analysis, the VIP > 1 and *p* < 0.05 were set as the standards to screen the differential metabolites. We found that a great number of metabolites were altered during PRV infection, which were summarized and shown in Venn diagrams. The numbers of differential metabolites in PK-15 cells infected with PRV-G were 430, 426, and 606 at 6, 12, and 24 h.p.i., respectively. Additionally, the numbers of differential metabolites changed by PRV-K infection were 556, 425, and 535 at different time points (Figure 3A,B). Compared to the mock group, a total of 375 and 194 metabolites were significantly upregulated in PRV-Gand PRV-K-infected cells, respectively (Figure 3C,D). Furthermore, the different changes in metabolites were due to the different times of PRV infection. In addition, these differential metabolites were classified and analyzed. As shown in Figure 3E–H, lipids and lipid-like molecules, organic acids and derivatives, nucleosides, nucleotides, analogues, and organic oxygen compounds account for nearly 80% in the PRV-infected cells. It is worth noting that, among these differential metabolites caused by both PRV strains, lipid metabolites accounted for the highest proportion: around 35%. These results suggest that the lipid metabolism of the host cell may play an important role in PRV replication.

**Figure 3.** Analysis of differentially expressed metabolites in PK-15 cells infected with different PRV strains. (**A**,**B**) Venn diagrams between PRV-infected groups (6H, 12H, 24H) and mock group. (**C**,**D**) Numbers of differentially expressed metabolites upregulated (red) and downregulated (blue) in infected groups. (**E**–**H**) Pie charts and the histogram graphs showing proportions of different categories among differentially expressed metabolites in PRV-infected PK-15 cells.

**Figure 4.** Heatmap analysis of 103 and 136 metabolites among PRV-G, PRV-K, and mock groups. Rows: metabolites; columns: samples. The color of each rectangle represents the relative level of the differential metabolites. Red: upregulated; blue: downregulated. (**A**) PRV-G; (**B**) PRV-K.

To study the crucial metabolites related to the PRV replication process, common differential metabolites among the three comparisons, i.e., 6 h.p.i. vs. mock, 12 h.p.i. vs. mock, and 24 h.p.i. vs. mock were screened. A total of 103 and 136 metabolites were obtained in PRV-G and PRV-K groups, respectively, and presented in the heatmap of hierarchical clustering analysis (Figure 4). As expected, the metabolites changed along with the virus infection. Many metabolites were significantly downregulated after PRV infection, especially lipids and lipid-like molecules. Glycerophospholipids and sphingolipids are important phospholipid molecules. Glycerophospholipids are divided into many categories according to the substitution groups, such as phosphatidylglycerol (PG), phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylinositol (PI), and cardiolipin (CL). During the early stage of PRV infection, some of the glycerophospholipids

increased. At 24 h.p.i., PRV-G infection induced a decrease in the levels of 78 species of glycerophospholipids, including 53 species of PC, 18 species of PE, 3 species of PS, and 2 species of PI and others (Figure 4A and Table S2). Meanwhile, PRV-K infection caused a decrease in the levels of 95 species of glycerophospholipids, including 64 species of PC, 21 species of PE, 3 species of PS, and 3 species of PI and others (Figure 4B and Table S3). In addition, few species of sphingolipids significantly decreased in PRV-G- and PRV-Kinfected cells, respectively. These results indicate that, in the late stage of virus infection, PRV needs to consume a large amount of lipids in the host cell to ensure its replication.

### *3.4. Metabolic Pathway Analysis of Metabolites*

These differential metabolites were annotated by using the KEGG Metabolome Database and further comprehensive analysis, including enrichment analysis and topological analysis, was conducted to find the metabolic pathways with high correlations. The results are shown in a bubble plot (Figure 5). In PRV-G vs. mock, differential metabolites were mainly enriched in arginine and proline metabolism; glycerophospholipid metabolism; glycine, serine, and threonine metabolism; purine metabolism; pyrimidine metabolism; and sphingolipid metabolism (Figure 5A–C). For PRV-K vs. mock, the metabolic pathways of the differential metabolites contained thiamine metabolism, purine metabolism, arginine and proline metabolism, glycerophospholipid metabolism, pyrimidine metabolism, and sphingolipid metabolism (Figure 5D–F).

**Figure 5.** The KEGG-enrichment pathway analysis of differentially expressed metabolites for PK-15 cells infected with different PRV strains in different time courses. (**A**–**C**) PRV-G (6H, 12H, 24H); (**D**–**F**) PRV-K (6H, 12H, 24H).

In summary, the results present more visualized profiles of the metabolite changes in PK-15 cells infected with two different PRV strains (Figure 6). The metabolic pathway contained a TCA cycle, lipid metabolism, amino acid metabolism, purine, and pyrimidine metabolism. The levels of adenosine at 12 and 24 h.p.i. in PRV-G/PRV-K-infected cells were more upregulated than that in the mock group as well as the levels of dTMP, which indicated that they may be required in the virus replication cycle. Adenosine is an important intermediate for the synthesis of adenosine triphosphate (ATP), adenine, and adenylate. Additionally, dTMP is a basic unit for deoxyribonucleic acid, which is the material basis for DNA synthesis. The major metabolic pathways of glycerophospholipids and sphingolipids and fatty acids during PRV infection were also reprogrammed, and the levels of PC, PE, ceramide, and sphingomyelin were consumed with the PRV replication process. Additionally, some amino acids were altered after PRV infection. These results indicate that two different PRV-strain infections led to the metabolic reprogramming of PK-15 cells to benefit self-replication.

**Figure 6.** Schematic overview of altered metabolic pathways in PK-15 cells infected with different PRV strains. The metabolites were shown in different colors according to their changes. Black: unchanged; red: upregulated; blue: downregulated.

### *3.5. Validation of Metabolomic Data by qPCR*

Given the possible roles of lipid metabolism during PRV infection, PK-15 samples were harvested at 24 h after PRV (PRV-G/PRV-K) infection with different MOI (1, 5, and 10, respectively) to further validate the metabolomic data. The mRNA levels of enzymes related to glycerophospholipid metabolism and sphingolipid metabolism were analyzed by qRT-PCR (Figure 7), including sphingosine kinase (SPHK1/2), sphingomyelin synthase (SGMS1/2), serine palmitoyltransferase small subunit A/B (SPTSSA/B), sphingomyelin phosphodiesterase 1,2,3 (SMPD1/2/3), fatty acid synthase (FASN), 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), phosphate cytidylyltransferase (PCYT1A/2), and phosphatidylserine decarboxylase (PISD). We found that SPTSSB, SMPD3, and PCYT2 were significantly increased in PRV-G-infected PK-15 cells (Figure 7A,B). Moreover, the PRV-K strain could apparently upregulate these three genes (Figure 7C,D). These results suggest that some pathways of lipid metabolism in host cells are promoted during PRV infection to facilitate viral replication.

**Figure 7.** The mRNA levels of sphingolipid- and glycerophospholipid-metabolism-related enzymes after different PRV-strain infections. PK-15 cells were harvested with different MOI (1, 5, 10) at 24 h.p.i after PRV (PRV-G or PRV-K) infection, while non-infected cells were used as control. The mRNA levels of different enzymes were determined by qRT-PCR. β-actin was used as an internal reference gene. (**A**,**B**) PRV-G; (**C**,**D**) PRV-K. The significance in the figure was indicated as follows: \*, *p* < 0.05; \*\*, *p* < 0.01.

### **4. Discussion**

Recently, a growing number of studies on the combination of metabolite profiling with disease have been reported [33,34]. Through the detection and analysis of endogenous small molecules in cells, researchers evaluated the biochemical differences between healthy and pathological organisms to obtain an insight into the pathology, etiology, and possible treatment options of diseases. In our study, we analyzed the metabolomic profiles of PRVinfected PK-15 cells based on UHPLC-QE-MS, and these findings provide new viewpoints on the interactions between PRV and host cells, which will help future studies on PRV.

This study involved a global metabonomic analysis of PK-15 cells that were infected with two different strains of PRV. Both OPLS-DA and the heatmap of hierarchical clustering analysis indicated that, compared to the mock group, PRV-infected cells showed significantly different metabolic profiles. We determined that PRV infection broke the metabolic homeostasis of PK-15 cells, caused metabolic reprogramming, and significantly affected the metabolism of lipid metabolism and nucleotides metabolism (Figure 6). The alterations of these metabolites and pathways reflected the cellular responses to PRV infection or the nutritional needs in virus replication.

The metabolic alterations caused by the infection of two different strains of PRV were different, which were caused by the characteristics of the virus itself. The sequence homology of the two strains was 96%, but the pathogenicity was significantly different. Following two different strains of PRV infection, PRV-G could cause a more serious typical cytopathic effect than that caused by PRV-K 24 h post infection. Therefore, PRV-infected cells showed strain-specific metabolic characteristics. However, there were some overlaps between these pairwise comparisons in the metabolic pathways (Figure 5), such as glycerophospholipid metabolism, sphingolipid metabolism, purine metabolism, and pyrimidine metabolism. Lipids are the structural basis of the cell biofilm, and are biologically active molecules, which participate in a variety of cellular processes and immune functions. We have known that palmitoyl-oleoyl-phosphatidylglycerol and PI inhibited inflammatory sequelae and the

infection of respiratory syncytial virus and influenza A virus, by destroying the binding of virus particles to plasma membrane receptors of host cells [35]. Flaviviruses also increased lipid synthesis to expand the surface area of membranes for better replication [36]. Dengue virus (DENV) infection required the manipulation of cellular fatty acid synthesis and cholesterol biosynthesis and transport [37,38]. It has been previously found that sphingolipids were essential for the successful completion of the viral life cycle, which is involved in attachment, membrane fusion, intracellular replication, assembly, and release of several viruses [39–41]. In our study, as an enveloped virus, PRV also required a large amount of lipids to participate in the formation and release of virions. Consistently, we found that many lipids (including glycerophospholipids and sphingolipids) were consumed in the late stage of virus infection (Figures 4 and 6). Purine metabolism and pyrimidine metabolism are the basic steps for nucleotide synthesis. In COVID-19 patients, some metabolites related to purine metabolism show an upward or downward trend compared to healthy controls, and correlation analysis showed a close correlation between these metabolites and proinflammatory cytokines/chemokines [42]. Previous studies have been reported to inhibit DENV replication with inhibitors (methotrexate and floxuridine) of the thymidine synthesis pathway [43]. Tiwari SK et al. utilized nucleoside metabolic inhibitors fluorouracil and floxuridine to inhibit Zika virus in human microglial cells [44]. Following PRV infection, nucleotide metabolism was markedly changed, suggesting an unknown role in PRV replication (Figure 6). This may expand the novel possibilities for the development of antiviral therapies.

Few studies have been performed on the host metabolism of PRV. Gou et al. explored the metabolic networks in PK-15 cells infected with PRV using gas chromatography-mass spectrometry (GC-MS) analysis. They reported that the metabolic flux derived from glycolysis, the pentose phosphate pathway, and glutamine metabolism for nucleotide biosynthesis was necessary for PRV replication [26]. In addition, Yao et al. indicated changes of PRV infection in iPAM on glycerolipids, fatty acyls, glycerophospholipids, and sphingolipids [27]. Our results are somewhat different from the previous two reports due to differences in the testing methods (GC-MS vs. LC-MS) and cells (iPAM vs. PK-15).

In our results, we found the mRNA level of PCYT2 was significantly increased in PK-15 cells infected with two PRV strains (Figure 7). PCYT2 is an important enzyme for the biosynthesis of PE from ethanolamine and diacylglycerol. Previous studies have shown treatment of PRV-infected PK-15 cells with meclizine, an inhibitor of PCYT2, led to decreased PRV infection and replication [28]. The result showed that glycerophospholipid metabolism was essential for PRV replication, but the mechanism was unclear.

Other herpesviruses, such as HSV-1, HCMV, Kaposi's sarcoma-associated herpesvirus (KSHV), and Epstein–Barr virus (EBV), have shown a notable ability to reprogram the host's metabolism for viral replication. HSV-1 infection led to increased levels of phosphoenolpyruvate, deoxypyrimidines, and pentose phosphate pathway intermediates [23]. Treatment with inhibitor of glucose metabolism or nucleoside analogs decreased the cell-tocell spread and production of HSV [45,46]. HCMV infection markedly increased glucose uptake and glycolysis flux and promoted flux through the TCA cycle and fatty acid biosynthesis pathway [23,47]. Following the inhibition of fatty acid biosynthesis by drugs, the level of HCMV replication was suppressed [47]. KSHV caused changes in many metabolites of glycolysis, the pentose phosphate pathway, amino acid metabolism, and lipogenesis [48]. Moreover, latent KSHV-infected endothelial cells depended on glutamine and glutaminolysis for survival [49]. This evidence showed that metabolic changes caused by virus infection played an important role in viral replication.

In conclusion, the metabolic profiles of PK-15 cells infected with different PRV strains were analyzed to display the metabolic changes by UHPLC-QE-MS. There were significant differences in lipid metabolism and nucleotide metabolism between the PRV-infected groups and the mock group. The ability of viruses to actively modulate host metabolism is crucial for the successful completion of the viral life cycle. Many inhibitors of lipid metabolism and nucleotide metabolism are already used against some viral infections; therefore, identifying metabolic targets for antiviral therapy may be a promising strategy. Our study provides much information for a further understanding of PRV pathogenesis and drug intervention for disease control.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/v14061158/s1: Table S1. Primer sequences used for qPCR; Table S2. Identification and characterization of the metabolites in PK-15 cells infected with PRV-G; Table S3. Identification and characterization of the metabolites in PK-15 cells infected with PRV-K.

**Author Contributions:** P.L. and X.L. conceived and designed the experiments; P.L. and D.H. performed the experiments, analyzed the data, and drafted the manuscript; L.Y., Z.L., X.Y. and Z.Z. contributed reagents/materials/analysis tools; P.L. and X.L. thoroughly revised the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was funded by the National Natural Science Foundation of China (Nos. 32102637 and 32172823) and the Project of the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

**Institutional Review Board Statement:** Not applicable for studies not involving humans or animals.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The metabolomic data is available with the link: https://pan.baidu. com/s/1SRndDrNqckJ7DxofO5zWeA (Password: hv45, accessed on 24 May 2022).

**Acknowledgments:** We want to express our gratitude to Chan Ding from Shanghai Veterinary Research Institute, CAAS for his valuable suggestions for this project.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


## *Article* **Proteomic Analysis of Vero Cells Infected with Pseudorabies Virus**

**Xintan Yang † , Shengkui Xu † , Dengjin Chen, Ruijiao Jiang, Haoran Kang, Xinna Ge , Lei Zhou , Jun Han , Yongning Zhang , Xin Guo \* and Hanchun Yang**

> Key Laboratory of Animal Epidemiology of the Ministry of Agriculture, College of Veterinary Medicine, China Agricultural University, Beijing 100193, China; s20193050751@cau.edu.cn (X.Y.); skxu0721@163.com (S.X.); chendengjin-cau@foxmail.com (D.C.); wn2jjj@163.com (R.J.); 18260068857@163.com (H.K.); gexn@cau.edu.cn (X.G.); leosj@cau.edu.cn (L.Z.); hanx0158@cau.edu.cn (J.H.); zhangyongning@cau.edu.cn (Y.Z.); yanghanchun1@cau.edu.cn (H.Y.)

**\*** Correspondence: guoxincau@cau.edu.cn; Tel.: +86-10-62732875

† These authors contributed equally to this work.

**Abstract:** Suid herpesvirus 1 (SuHV-1), known as pseudorabies virus (PRV), is one of the most devastating swine pathogens in China, particularly the sudden occurrence of PRV variants in 2011. The higher pathogenicity and cross-species transmission potential of the newly emerged variants caused not only colossal economic losses, but also threatened public health. To uncover the underlying pathogenesis of PRV variants, Tandem Mass Tag (TMT)-based proteomic analysis was performed to quantitatively screen the differentially expressed cellular proteins in PRV-infected Vero cells. A total of 7072 proteins were identified and 960 proteins were significantly regulated: specifically 89 upregulated and 871 downregulated. To make it more credible, the expression of XRCC5 and XRCC6 was verified by western blot and RT-qPCR, and the results dovetailed with the proteomic data. The differentially expressed proteins were involved in various biological processes and signaling pathways, such as chaperonin-containing T-complex, NIK/NF-κB signaling pathway, DNA damage response, and negative regulation of G2/M transition of mitotic cell cycle. Taken together, our data holistically outline the interactions between PRV and host cells, and our results may shed light on the pathogenesis of PRV variants and provide clues for pseudorabies prevention.

**Keywords:** pseudorabies virus; Vero cell; TMT-based proteomic analysis; differentially expressed proteins

### **1. Introduction**

Pseudorabies (PR), also known as Aujeszky's disease (AD), is one of the most notorious swine diseases and causes enormous economic losses to the pig-raising industry [1]. Typical clinical symptoms of PR include respiratory distress, nervous disorders, and reproductive failures in sows [2,3]. PR is caused by pseudorabies virus (PRV), also called suid herpesvirus 1 (SuHV-1), which belongs to the subfamily of *Alphaherpesvirus* in the family of *Herpesviridae*. The genome of PRV is about 175 kb in length and encodes over 70 viral proteins contributing to neuronal latent infection and immune modulation [4,5].

Since the first report of PR outbreak in the 1950s, PRV has spread through China over the past 70 years [6]. The intensive herd vaccination by attenuated live vaccine Bartha-K61 facilitates PR eradication, whereas strong immune pressure may accelerate the virus's evolution and pave the way for the emergence of variants. In 2011, large scale outbreaks of PR caused by PRV variants swept China [6,7]. Subsequent studies showed that the emerging variants had higher pathogenicity, and the typical vaccine Bartha-K61 only provided limited protection against PRV variants infections [7,8]. Despite Jianle Ren et al. reporting that glycoproteins C and D of PRV variant strain HB1201 contribute individually to the escape from Bartha-K61 vaccine-induced protection [9], the pathogenesis of PRV variants remains largely unclear.

PRV has a wide host range and is capable of infecting numerous animals. Increasing evidence suggested that the newly emerged variants from 2011 were the most prevalent

**Citation:** Yang, X.; Xu, S.; Chen, D.; Jiang, R.; Kang, H.; Ge, X.; Zhou, L.; Han, J.; Zhang, Y.; Guo, X.; et al. Proteomic Analysis of Vero Cells Infected with Pseudorabies Virus. *Viruses* **2022**, *14*, 755. https:// doi.org/10.3390/v14040755

Academic Editors: Yan-Dong Tang and Xiangdong Li

Received: 8 March 2022 Accepted: 31 March 2022 Published: 4 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

genotypes worldwide and most frequently involved in cross-species transmission [10]. Next-generation sequencing and regular polymerase chain reaction (PCR) confirmed the presence of PRV genomes in cerebral spinal fluid from a 43-year-old patient [11]. In addition, a patient who presented with encephalitis and pulmonary infection also tested PRV positive in his cerebrospinal fluid and vitreous humor [12]. More severely, a PRV strain was isolated from an acute human encephalitis case in 2019, confirming the interspecies transmission between pigs and humans and the replication capacity of PRV in human [13]. Other animals, including bovine and wolf, were also reported to be infected by PRV [14,15]. Understanding the interactions in-depth between PRV infection and host may provide ideas for interspecies transmission prevention.

Innate immunity is the host's first line of defense against virus infection. When invading host cells, pathogens are recognized by specific pattern recognition receptors (PRRs) and then trigger immune responses [16]. To establish efficient infection, PRV has evolved various strategies to evade immune clearance. For example, PRV US3 degrades Bcl-2 associated transcription factor 1 to impair type I interferon production and benefit virus replication [17]; UL50 induces the degradation of type I interferon receptor via lysosomal pathway to antagonize interferon response [18]. Although NF-κB signaling pathway is activated during PRV infection, the expression of pro-inflammatory genes was inhibited [19]. Additionally, Wang et al. found that UL24 protein could abrogate tumor necrosis factor alpha (TNF-α)-mediated NF-κB activation [20]. We previously reported that PRV could dramatically enhance the dephosphorylation of eIF2α and thus promote host cell translation efficacy to facilitate its replication [21]. Higher pathogenesis and crossspecies transmission ability of PRV may partly attribute to the enhanced immune evasion of PRV variants. Despite several decades of intensive study, the underlying mechanisms of PRV pathogenesis and immunomodulation still remain elusive. Hence, it is imperative to investigate the host factors involved in virus infection.

To date, proteomics is broadly applied to hunt for host factors relevant to virus infection [22]. Various animal viruses had been subjected to proteomic analysis to dissect the host factors involved in virus infection, such as porcine epidemic diarrhea virus (PEDV) [23], porcine reproductive and respiratory syndrome virus (PRRSV) [24] and porcine deltacoronavirus [25]. Tandem Mass Tag (TMT) technology, developed and launched by Thermo, is one of the most powerful quantitative methods for protein expression analysis with the highest throughput, the lowest systematic error, and the most powerful functions. In this study, TMT-based quantitative proteomics was employed to analyze protein profiles in mock- and PRV-infected Vero cells to gain insights into the virus-host interactions.

### **2. Materials and Methods**

### *2.1. Cell Lines, Viruses, Chemicals, and Antibodies*

African green monkey kidney cell (Vero), the immortalized porcine alveolar macrophage (CRL-2843), and porcine kidney cell (PK-15) were all cultured in Dulbecco's modified Eagle's medium (DMEM: Invitrogen, Carlsbad, CA, USA) containing 10% (*v/v*) fetal bovine serum (FBS, Thermo Fisher, Waltham, MA, USA) in a humidified 37 ◦C incubator with 5% CO<sup>2</sup> and stored in our lab. PRV HB1201 (GenBank accession number: KU057086.1) was a variant strain isolated from a pig in He Bei in China. 40 , 60 -diamidino-2-phenylindole (DAPI) and TMT 16Plex were purchaseded from Thermo Fisher Scientific (Waltham, MA, USA). The primary antibodies used in this study were specific for XRCC5 (16389-1-AP, Proteintech, Rosemont, IL, USA), XRCC6 (10723-1-AP, Proteintech, Rosemont, IL, USA), β-actin (66009- 1-Ig, Proteintech, Rosemont, IL, USA), VP5 (prepared in our lab), and gB (prepared in our lab). The HRP-labeled secondary antibodies against rabbit (ZB2301) and mouse (ZB2305) were all purchased from ZSGB-BIO (Beijing, China).

### *2.2. Virus Inoculation and Protein Preparation*

Vero cells were grown to monolayers in 10 cm cell culture dishes and then were inoculated with PRV HB1201 at 0.1 MOI for 1 h. Sustaining culture medium DMEM

containing 2% FBS was added for another 24 h. Three independent experiments were conducted as biological replicates. The protein extraction procedure is as follows: at 24 h post-inoculation (h p.i.), the medium was removed and washed with 5 mL pre-cooling PBS twice; mock- or PRV-infected Vero cells were collected using a cell scraper and piped into 1.5 mL EP tubes; protein lysate (8 M urea, 1% SDS containing protease inhibitor) was added to lyse cell membrane and sonicated for 2 min to solubilize protein further; cell lysate was used to treat protein for another 30 min on ice and centrifuged (12,000 rpm for 15 min at 4 ◦C) to remove cellular debris. The protein concentration was analyzed by Bradford protein assay and SDS-PAGE was performed to evaluate the overall protein quality.

### *2.3. Reductive Alkylation and TMT Labeling*

Protein reductive alkylation and TMT labeling procedures were conducted according to the instructions as follows. Briefly, 100 µg protein was treated with triethylammonium bicarbonate buffer (TEAB) to the final concentration of 100 mM, and then Tris (2-carboxyethyl) phosphine (TCEP) was added to make the final concentration 10 mM for 60 min at 37 ◦C; 40 mM iodoacetamide was added to the final concentration and reacted in a dark room for 40 min at room temperature (RT); ice-cold acetone was added (v:v = 6:1) and reacted for 4 h at −20 ◦C, and the liquid was removed after centrifugation at 10,000× *g* for 20 min; sediment was dissolved with 100 µL 100 mM TEAB and digested with trypsin (m:m = 1:50) fully overnight at 37 ◦C; finally, TMT was added to label proteins for 2 h at RT, followed by hydroxylamine treatment for another 30 min.

### *2.4. Immunofluorescence Assay (IFA)*

The IFA was performed according to the protocol mentioned previously [21]. In brief, Vero cells seeded on coverslips in a six-well plate over 90% confluence were inoculated with 0.1 MOI PRV HB1201; then, the inoculated cells were fixed with 3.7% paraformaldehyde at indicated time points for 10 min and permeabilized with 2% bovine serum albumin (BSA) containing 0.1% Triton X-100 for 10 min; 2% BSA was used to block cells for 30 min and primary monoclonal antibody specific for gB with 1:1000 dilution incubated cells for 1 h at RT and then washed with PBS three times; secondary antibodies were added at RT for 1 h in a humid chamber; after one wash, nucleus were stained with DAPI (Molecular Probes) for 10 min and washed with PBS five times for 5 min each; finally, the coverslips were observed with a Nikon A1 microscope or laser confocal microscope.

### *2.5. RNA Extraction and Real-Time PCR Analysis*

Total RNAs of mock- or PRV-infected Vero cells were extracted by TRIzol reagent (Biomed, Beijing, China). The culture medium was removed and the cells in six-well plates were lysed with 750 µL TRIzol for 5 min, then 250 µL chloroform was added to separate RNA. After centrifugation at 12,000 rpm at 4 ◦C for 10 min, the RNA fraction was transferred into a new tube and precipitated by 0.8 volumes of isopropanol. After centrifugation for 15 min at 12,000 rpm, RNA pellets were washed twice with 75% iced ethanol and resuspended in 20 µL RNase-free H2O. The synthesis of cDNA was performed using Fast Quant RT Kit (With gDNase) (Tian Gen Biotech, Beijing, China) according to the manufacturer's instructions. The cDNA samples were quantified by SYBR Green RT-qPCR Master Mix (Vazyme, Nanjing, China) and repeated three times. All reactions were carried out by the Bio-Rad PCR system. All primers used in this study are listed in Table 1. The mRNA abundance of GAPDH, XRCC5, and XRCC6 were detected by RT-qPCR assay using specific primer sets GAPDHF/GAPDHR, XRCC5F/XRCC5R, and XRCC6F/XRCC6R respectively.


**Table 1.** Primers used in this study.

### *2.6. Western Blot Analysis*

PRV HB1201-infected Vero, CRL-2843, and PK-15 cells were all harvested at 24 h p.i. The cells were lysed with radioimmunoprecipitation (RIPA) lysis buffer (Beyotime, Shanghai, China) containing protease inhibitor (1 mM PMSF) for 30 min, and the supernatant was transferred to a new tube after centrifugation. The protein concentration was determined with Pierce BCA Protein Assay Kit (Thermo Fisher, Waltham, MA, USA) and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Separated protein (10 µg each channel) was transferred onto polyvinylidene difluoride (PVDF) membrane (Millipore). PVDF membranes were blocked in 5% skimmed-milk-PBST at RT for 2 h, followed by incubation with primary antibodies at 4 ◦C overnight. Then, the PVDF membrane were washed three times with 0.05% PBST for 5 min each at a rotator and incubated with the HRP-conjugated secondary antibodies at 1:3000 dilution. After three washes, the membranes were incubated with ECL chemiluminescence detection kit (Pierce) for 2 min, and finally exposed to a chemiluminescence apparatus (Bio-Rad, Hercules, CA, USA).

### *2.7. Virus Titration*

Viruses were serially diluted 10-fold with DMEM containing 2% FBS and inoculated into Vero cells at 90% confluence in 96-well culture plates. 72 h p.i. or later, the virus titers were calculated based on the cytopathic effects (CPE) according to the Reed-Muench method. Virus titers were determined from at least three independent experiments.

### *2.8. Data Analysis*

All data were processed with GraphPad Prism 6 (GraphPad Software Inc., San Diego, CA, USA). The student's *t*-test or non-parametric test was used to analyze the difference between the values of two groups. A value of *p* < 0.05 was considered statistically significant.

### **3. Results**

### *3.1. Kinetics of PRV HB1201 Replication in Vero Cells*

Efficient viral infection and relatively mild cell collapse are critical factors for optimal sampling. PRV HB1201 could cause severe CPE and subsequently cell collapse on Vero cells, thus relative lower MOI (MOI = 0.1) was applied to infect Vero cells. To screen the optimal time points of sampling, the kinetics of PRV replication in Vero cells were determined at various time points by TCID50. As shown in Figure 1B, the virus titers were up to 10<sup>8</sup> TCID50/mL at 24 h p.i., similar to that at 30 to 48 h p.i., indicating PRV could propagate in Vero cells efficiently, and the virus titers reached a plateau at 24 h p.i. (Figure 1B). Furthermore, IFA results showed that gB positive cells increased as the infection progressed. Notably, most cells were infected at 24 h p.i., and the gB positive cells decreased after 30 h p.i. due to excessive cell collapse (Figure 1A). Meanwhile, the CPE was observed microscopically at various time points. Compared with mock-infected cells, PRV-Infected cells developed slightly visible CPE at 12 h p.i. and CPE were fairly apparent at 24 h p.i. (Figure 1A). Cell collapse soars from 30 h p.i., and many of the cells were detached and floated in the medium. In addition, the expression of viral capsid protein VP5 was detected by western blot. The level of VP5 increased gradually as infection progressed (Figure 1C). However, VP5 expression level decreased slightly at 30 h p.i. compared to that at 24 and 18 h p.i. This may result from cell detachment and virus release into the medium (Figure 1A).

Based on the results above, Vero cells infected with 0.1 MOI PRV for 24 h p.i. were regarded as optimal sampling time points and subjected to the following proteomic analysis. were regarded as optimal sampling time points and subjected to the following proteomic analysis.

cells developed slightly visible CPE at 12 h p.i. and CPE were fairly apparent at 24 h p.i. (Figure 1A). Cell collapse soars from 30 h p.i., and many of the cells were detached and floated in the medium. In addition, the expression of viral capsid protein VP5 was detected by western blot. The level of VP5 increased gradually as infection progressed (Figure 1C). However, VP5 expression level decreased slightly at 30 h p.i. compared to that at 24 and 18 h p.i. This may result from cell detachment and virus release into the medium (Figure 1A). Based on the results above, Vero cells infected with 0.1 MOI PRV for 24 h p.i.

*Viruses* **2022**, *14*, x FOR PEER REVIEW 5 of 15

**Figure 1.** The replication of PRV variant HB1201 in Vero cells. (**A**) PRV at 0.1 MOI infects Vero cells and CPE was observed in microscopy at various time points. Meanwhile, IFA was also applied to view the efficiency of virus infection with antibody against gB protein. (**B**) The whole cells infected with 0.1 MOI virus were collected at indicated time points and tittered by TCID50. (**C**) Vero cells infected with 0.1 MOI PRV were collected and the whole cell lysis was subjected to western blot analysis to detect the expression of VP5. **Figure 1.** The replication of PRV variant HB1201 in Vero cells. (**A**) PRV at 0.1 MOI infects Vero cells and CPE was observed in microscopy at various time points. Meanwhile, IFA was also applied to view the efficiency of virus infection with antibody against gB protein. (**B**) The whole cells infected with 0.1 MOI virus were collected at indicated time points and tittered by TCID50. (**C**) Vero cells infected with 0.1 MOI PRV were collected and the whole cell lysis was subjected to western blot analysis to detect the expression of VP5.

#### *3.2. Protein Profiles Determined by TMT/MS Analysis* Proteomics is a systematic approach to study the virus-host interactions. To identify *3.2. Protein Profiles Determined by TMT/MS Analysis*

the differentially expressed proteins (DEPs) between mock- and PRV-infected cells, TMTbased quantitative proteomic analysis was performed, and the workflow is shown as Figure 2A. A total of 7072 cellular proteins were identified and quantified at 24 h p.i., among which 91 proteins were significantly upregulated and 879 proteins were downregulated compared to those in mock-infected Vero cells (Figure 2B) according to the criteria (*p*value < 0.05 and fold change >1.5 or fold change <0.67). In addition, the top 20 upregulated and top 20 downregulated proteins are listed in Tables 2 and 3, respectively. Three technical replicates were carried out to improve the reliability of our data. Proteomics is a systematic approach to study the virus-host interactions. To identify the differentially expressed proteins (DEPs) between mock- and PRV-infected cells, TMT-based quantitative proteomic analysis was performed, and the workflow is shown as Figure 2A. A total of 7072 cellular proteins were identified and quantified at 24 h p.i., among which 91 proteins were significantly upregulated and 879 proteins were downregulated compared to those in mock-infected Vero cells (Figure 2B) according to the criteria (*p*-value < 0.05 and fold change >1.5 or fold change <0.67). In addition, the top 20 upregulated and top 20 downregulated proteins are listed in Tables 2 and 3, respectively. Three technical replicates were carried out to improve the reliability of our data.

data.

the criteria (*p*‐value < 0.05 and fold change >1.5 or fold change <0.67). In addition, the top 20 upregulated and top 20 downregulated proteins are listed in Tables 2 and 3, respectively. Three technical replicates were carried out to improve the reliability of our

**Figure 2.** Overview of proteomic analysis procedure and DEPs. (**A**) The workflow of proteomic analysis. Vero cells were infected with 0.1 MOI PRV for 24 h and the whole cells lysis was collected after centrifugation at 12,000 rpm for 5 min. After reductive alkylation, the protein was labeled with TMT. Finally, the samples were subjected to liquid chromatography tandem mass spectrometry and bioinformatics analysis. (**B**) A total of 7022 proteins were identified, among which 89 proteins were markedly upregulated (red dots), 879 proteins were downregulated (green dots), and the remaining 6102 proteins stayed constant (gray dots). Proteins were considered significantly differently expressed when *p* value was less than 0.05 and fold change was less than 0.67 or more than 1.5 in this study. **Figure 2.** Overview of proteomic analysis procedure and DEPs. (**A**) The workflow of proteomic analysis. Vero cells were infected with 0.1 MOI PRV for 24 h and the whole cells lysis was collected after centrifugation at 12,000 rpm for 5 min. After reductive alkylation, the protein was labeled with TMT. Finally, the samples were subjected to liquid chromatography tandem mass spectrometry and bioinformatics analysis. (**B**) A total of 7022 proteins were identified, among which 89 proteins were markedly upregulated (red dots), 879 proteins were downregulated (green dots), and the remaining 6102 proteins stayed constant (gray dots). Proteins were considered significantly differently expressed when *p* value was less than 0.05 and fold change was less than 0.67 or more than 1.5 in this study.

#### **Accession Description FC** *p* **Value** *3.3. Validation of TMT/MS Data by Western Blot and RT-qPCR*

XP\_007959284.1 bromodomain‐containing protein 9 isoform X1 3.054434 0.006955 Yes

**Table 2.** Top 20 up‐regulated proteins.

**(P\_24h/M\_24h) (P\_24h/M\_24h) Significant** XP\_007997295.1 <sup>4</sup>‐hydroxybenzoate polyprenyltransferase, mitochondrial 16.440879 0.0111 Yes XP\_008000412.1 ATP synthase subunit gamma, mitochondrial isoform X2 5.82243 0.03774 Yes XP\_007964526.1 non‐homologous end‐joining factor 1 5.638689 0.002289 Yes XP\_007966188.1 transmembrane <sup>7</sup> superfamily member <sup>3</sup> isoform X1 4.570457 0.009848 Yes XP\_007985885.1 DNA‐directed RNA polymerase III subunit RPC5 isoform X1 4.318305 0.01897 Yes XP\_008001057.1 proton myo‐inositol cotransporter 4.10077 0.04733 Yes XP\_007978034.1 hemoglobin subunit alpha 3.092117 0.03053 Yes To verify TMT/MS data, X-ray repair cross-complementing protein 5 (XRCC5) and X-ray repair cross-complementing protein 6 (XRCC6) were analyzed by western blot in both mock- and PRV-infected Vero cells. The two proteins were selected for validation for the following reasons: they were downregulated significantly; they were closely related and involved in DNA repair process, which was a general cellular response during herpes virus infection [26]; and antibodies against them were commercially available. Western blot results showed that the protein level of XRCC5 and XRCC6 both decreased in PRV-infected Vero cells (Figure 3A). Then, Image J software was applied to quantify protein levels, and the ratios of the XRCC5 and XRCC6 between mock-and PRV-infected Vero cells coincided with proteomic data (Figure 3B). Moreover, the levels of XRCC5 and XRCC6 in PK-15 and CRL-2843 cells were also reduced (Figure 3E,F), indicating PRV-mediated XRCC5 and XRCC6 reduction was in a cell type-independent manner. *Viruses* **2022**, *14*, x FOR PEER REVIEW 8 of 15

**Figure 3.** Validation of proteomics data by western blot and RT-qPCR. (**A**) Vero cells infected with PRV for 24 h were collected and western blot was performed to detect the expression of XRCC5 and XRCC6 with corresponding antibodies. (**B**) The western blot and proteomics ratio of XRCC5 and XRCC6. (**C**) Relative XRCC6 transcription in Vero cells. (**D**) Relative XRCC5 transcription in Vero cells. (**E**) The expression of XRCC5 and XRCC6 in PK-15 infected with PRV. (**F**) The expression of XRCC5 and XRCC6 in CRL-2843 infected with PRV. \*\* indicates significance at a 99% confidence interval (*p* < 0.01) *3.4. GO Analysis of The DEPs* GO annotation analysis could classify the tested proteins in three aspects: biological **Figure 3.** Validation of proteomics data by western blot and RT-qPCR. (**A**) Vero cells infected with PRV for 24 h were collected and western blot was performed to detect the expression of XRCC5 and XRCC6 with corresponding antibodies. (**B**) The western blot and proteomics ratio of XRCC5 and XRCC6. (**C**) Relative XRCC6 transcription in Vero cells. (**D**) Relative XRCC5 transcription in Vero cells. (**E**) The expression of XRCC5 and XRCC6 in PK-15 infected with PRV. (**F**) The expression of XRCC5 and XRCC6 in CRL-2843 infected with PRV. \*\* indicates significance at a 99% confidence interval (*p* < 0.01).

and so on; in the MF category, the DEPs were involved in binding function.

process (BP), cellular component (CC), and molecular function (MF). To dissect the function of DEPs, GO functional analysis revealed that 89 upregulated proteins and 871 downregulated proteins were involved in 12 biological processes (Figure 3A), including cellular

tributed in different cell components, including cell parts, cells, organelle, organelle parts,

In addition, GO enrichment analysis demonstrated that DEPs were mostly enriched in chaperonin-containing T-complex within the CC category. Furthermore, the majority of DEPs were enriched in the BP category, such as NIK/NF-κB signaling, Fc-epsilon receptor signaling pathway, negative regulation of G2/M transition of mitotic cell cycle, and innate immune response activating cell surface receptor signaling pathway (Figure 4B).


**Table 2.** Top 20 up-regulated proteins.

In addition, the transcription level of XRCC5 and XRCC6 were also analyzed by RT-qPCR. Consistently, the mRNA level of XRCC5 and XRCC6 markedly decreased in virus-infected Vero cells compared with that in mock-infected cells (Figure 3C,D), suggesting that PRVmediated XRCC5 and XRCC6 downregulation might result from transcription inhibition.

### *3.4. GO Analysis of The DEPs*

GO annotation analysis could classify the tested proteins in three aspects: biological process (BP), cellular component (CC), and molecular function (MF). To dissect the function of DEPs, GO functional analysis revealed that 89 upregulated proteins and 871 downregulated proteins were involved in 12 biological processes (Figure 3A), including cellular processes, single-organism processes, metabolic processes, biological processes, regulation of biological processes, and so on; within the CC category, the DEPs were well distributed in different cell components, including cell parts, cells, organelle, organelle parts, and so on; in the MF category, the DEPs were involved in binding function.


**Table 3.** Top 20 down-regulated proteins.

In addition, GO enrichment analysis demonstrated that DEPs were mostly enriched in chaperonin-containing T-complex within the CC category. Furthermore, the majority of DEPs were enriched in the BP category, such as NIK/NF-κB signaling, Fc-epsilon receptor signaling pathway, negative regulation of G2/M transition of mitotic cell cycle, and innate immune response activating cell surface receptor signaling pathway (Figure 4B).

### *3.5. KEGG Functional Annotation of DEPs*

KEGG pathway analysis was performed to further explore the underlying signaling pathways or functions among DEPs. As shown in Figure 5A, the 89 upregulated proteins participated in 32 pathways, and the top three were related to the immune system, signal transduction, and cancer. Meanwhile, the 871 downregulated proteins were involved in 44 pathways, and the top three were the "folding, sorting, and degradation of protein", signal transduction, and translation (Figure 5B).

KEGG enrichment analysis were also conducted to analyze the enriched signaling pathways in DEPs. Among all 970 DEPs, 20 pathways were significantly enriched, and the top three were proteasome, amino sugar and nucleotide sugar metabolism, and RNA polymerase (Figure 5C).

9 of 15

**Figure 4.** GO annotation and GO enrichment analysis of DEPs between mock- and PRV-infected Vero cells. (**A**) Up- and downregulated proteins are classified into three categories, respectively, by GO analysis: biological process (BP), cellular component (CC), and molecular function (MF). The xaxis represents the specific categories in BP, CC, and MF. The numbers on the y-axis indicate proteins in the category. (**B**) DEPs were subjected to GO enrichment analysis and the top 20 GO terms are listed on the x-axis. The y-axis indicates the enrichment ratio of DEPs and different colors represent different *p* values. \*\*\* indicates significance at a 99.9% confidence interval (*p* < 0.001). **Figure 4.** GO annotation and GO enrichment analysis of DEPs between mock- and PRV-infected Vero cells. (**A**) Up- and downregulated proteins are classified into three categories, respectively, by GO analysis: biological process (BP), cellular component (CC), and molecular function (MF). The x-axis represents the specific categories in BP, CC, and MF. The numbers on the y-axis indicate proteins in the category. (**B**) DEPs were subjected to GO enrichment analysis and the top 20 GO terms are listed on the x-axis. The y-axis indicates the enrichment ratio of DEPs and different colors represent different *p* values. \*\*\* indicates significance at a 99.9% confidence interval (*p* < 0.001).

*Viruses* **2022**, *14*, x FOR PEER REVIEW 10 of 15

KEGG pathway analysis was performed to further explore the underlying signaling pathways or functions among DEPs. As shown in Figure 5A, the 89 upregulated proteins participated in 32 pathways, and the top three were related to the immune system, signal transduction, and cancer. Meanwhile, the 871 downregulated proteins were involved in 44 pathways, and the top three were the "folding, sorting, and degradation of protein",

KEGG enrichment analysis were also conducted to analyze the enriched signaling pathways in DEPs. Among all 970 DEPs, 20 pathways were significantly enriched, and the top three were proteasome, amino sugar and nucleotide sugar metabolism, and RNA

*3.5. KEGG Functional Annotation of DEPs*

polymerase (Figure 5C).

signal transduction, and translation (Figure 5B).

**Figure 5.** (**A**) The 89 upregulated proteins are classified into six main categories by KEGG analysis: metabolism, genetic information processing, environmental information processing, cellular processes, organismal systems, and human diseases. The x-axis indicates the numbers of proteins within particular categories. The y-axis indicates the specific pathways within six main categories. (**B**) The 871 downregulated proteins are classified in the same manner as Figure A: (**C**) KEGG pathway enrichment analysis of DEPs. The x-axis indicates the name of the KEGG pathways, the y-axis indicates the enrichment ratio (there were no apparent differences between 0.01 and 0.05, so the colors look similar and can't be distinguished by the naked eye). \* indicates significance at a 95% confidence interval (*p* < 0.05), \*\* indicates significance at a 99% confidence interval (*p* < 0.01), \*\*\* indicates significance at a 99.9% confidence interval (*p* < 0.001).

### *3.6. COG Annotation of DEPs*

The COG database is able to predicate the function of proteins based on protein sequence. To categorize the functions of DEPs, COG analysis was performed. As shown in

Figure 6 (left panel) 10 categories were involved in upregulated proteins. In particular, seven proteins were related to posttranslational modification, protein turnover, and chaperones; four proteins were classified into general function prediction only; three proteins were related to replication, recombination, and repair; two proteins were relevant to energy production and conversion, transcription, intracellular trafficking, secretion, vesicular transport, and so on. In addition, 22 categories were involved in 879 downregulated proteins: 61 proteins were related to posttranslational modification, protein turnover, and chaperones; 46 proteins were relevant to translation, ribosomal structure, and biogenesis; 39 proteins were classified into general function prediction only shown in Figure 6 (right panel). Further research is imperative to characterize the involvement of these categories during PRV infection. sequence. To categorize the functions of DEPs, COG analysis was performed. As shown in Figure 6 (left panel) 10 categories were involved in upregulated proteins. In particular, seven proteins were related to posttranslational modification, protein turnover, and chaperones; four proteins were classified into general function prediction only; three proteins were related to replication,recombination, and repair; two proteins were relevant to energy production and conversion, transcription, intracellular trafficking, secretion, vesicular transport, and so on. In addition, 22 categories were involved in 879 downregulated proteins: 61 proteins were related to posttranslational modification, protein turnover, and chaperones; 46 proteins were relevant to translation, ribosomal structure, and biogenesis; 39 proteins were classified into general function prediction only shown in Figure 6 (right panel). Further research is imperative to characterize the involvement of these categories during PRV infection.

The COG database is able to predicate the function of proteins based on protein

**Figure 5.** (**A**) The 89 upregulated proteins are classified into six main categories by KEGG analysis: metabolism, genetic information processing, environmental information processing, cellular processes, organismal systems, and human diseases. The x‐axis indicates the numbers of proteins within particular categories. The y‐axis indicates the specific pathways within six main categories. (**B**) The 871 downregulated proteins are classified in the same manner as Figure A: (**C**) KEGG pathway enrichment analysis of DEPs. The x‐axis indicates the name of the KEGG pathways, the y‐ axis indicates the enrichment ratio (there were no apparent differences between 0.01 and 0.05, so the colors look similar and can't be distinguished by the naked eye). \* indicates significance at a 95% confidence interval (*p* < 0.05), \*\* indicates significance at a 99% confidence interval (*p* < 0.01), \*\*\*

*Viruses* **2022**, *14*, x FOR PEER REVIEW 11 of 15

indicates significance at a 99.9% confidence interval (*p* < 0.001)

*3.6. COG Annotation of DEPs*

**Figure 6.** The COG annotation of significantly upregulated proteins (**left panel**) and downregulated proteins (**right panel**). The x‐axis indicates the COG functional classification (presented with A to Z): (A) RNA processing and modification; (B) Chromatin structure and dynamics; (C) Energy production and conversion; (D) Cell cycle control, cell division, and chromosome partitioning; (E) Amino acid transport and metabolism; (F) Nucleotide transport and metabolism; (G) Carbohydrate transport and metabolism; (H) Coenzyme transport and metabolism; (I) Lipid transport and metabolism; (J) Translation, ribosomal structure, and biogenesis; (K) Transcription; (L) Replication, recombination, and repair; (M) Cell wall/membrane/envelope biogenesis; (N) Cell motility; (O) Posttranslational modification, protein turnover, and chaperones; (P) Inorganic ion transport and metabolism; (Q) Secondary metabolites biosynthesis, transport, and catabolism; (R) General function prediction only; (S) Function unknown; (T) Signal transduction mechanisms; (U) Intracellular trafficking, secretion, and vesicular transport; (V) Defense mechanisms; (W) Extracellular structures; (Y) Nuclear structure; (Z) Cytoskeleton. The y‐axis indicates the protein number of particular functional classifications. **Figure 6.** The COG annotation of significantly upregulated proteins (**left panel**) and downregulated proteins (**right panel**). The x-axis indicates the COG functional classification (presented with A to Z): (A) RNA processing and modification; (B) Chromatin structure and dynamics; (C) Energy production and conversion; (D) Cell cycle control, cell division, and chromosome partitioning; (E) Amino acid transport and metabolism; (F) Nucleotide transport and metabolism; (G) Carbohydrate transport and metabolism; (H) Coenzyme transport and metabolism; (I) Lipid transport and metabolism; (J) Translation, ribosomal structure, and biogenesis; (K) Transcription; (L) Replication, recombination, and repair; (M) Cell wall/membrane/envelope biogenesis; (O) Posttranslational modification, protein turnover, and chaperones; (P) Inorganic ion transport and metabolism; (Q) Secondary metabolites biosynthesis, transport, and catabolism; (R) General function prediction only; (S) Function unknown; (T) Signal transduction mechanisms; (U) Intracellular trafficking, secretion, and vesicular transport; (V) Defense mechanisms; (Z) Cytoskeleton. The y-axis indicates the protein number of particular functional classifications.

#### **4. Discussion 4. Discussion**

PRV variant HB1201 exhibits higher pathogenicity, and its pathogenesis remains poorly defined. Nowadays, proteomics has been broadly used in profiling cellular protein expression patterns in virus‐infected cells. In this paper, a TMT‐based quantitative PRV variant HB1201 exhibits higher pathogenicity, and its pathogenesis remains poorly defined. Nowadays, proteomics has been broadly used in profiling cellular protein expression patterns in virus-infected cells. In this paper, a TMT-based quantitative proteomics approach was applied, and we revealed striking protein profile shifts in PRV-infected Vero cells compared with those in mock-infected cells.

In the present study, a total of 7072 proteins were identified in whole Vero cells, among which 980 proteins were differentially expressed at 24 h p.i. Among the top 20 upregulated proteins, non-homologous end joining-1 (NHEJ-1) markedly induced an over five-fold change (Table 2), which is reported to be involved in DNA repair [27]. DNA viruses replicates their genomes in the nuclei of cells, and the mass accumulation of viral DNA genome in the nucleus may trigger host cell DNA damage responses. For example, intensive studies showed that herpes virus may engage components of DNA damage response to enhance its replication, while some of the DNA repair components are antiviral [28–30]. In our analysis, the expression of two DNA repair-related proteins, XRCC5 (Ku80) and XRCC6 (Ku70), were both shown to be reduced in PRV-infected cells. Moreover, western blot and RT-qPCR results supported the proteomic data at both protein and transcription level. XRCC5 and XRCC6 comprise the heterodimer, which recognizes and binds to double

strand DNA break ends, and then promotes non-homologous end joining [30] or induces innate immune defenses against DNA virus infection [31,32]. Previous reports showed that XRCC6 not only modulated human T lymphotropic virus type 1 (HTLV-1) replication [33], but also regulated DNA virus-mediated innate immune response [34]. However, the expression of XRCC6 was significantly upregulated in HTLV-1-infected cells compared with PRV. We hold that PRV, with its larger genome, encodes more proteins and evolves more sophisticated strategies to evade host immune clearance by targeting XRCC5 and XRCC6. As compensation, many other tricks have been developed instead by HTLV-1. For example, HTLV-1 Tax could impair K63-linked ubiquitination of STING to evade host innate immunity [35]; HTLV-1 Tax blocks IRF3 phosphorylation through the interaction with and inhibition of TBK1 kinase [36]. These results above indicated that the DNA damage repair signaling pathway might be closely related to virus infection.

Our previous study showed that PRV infection induced the phosphorylation of PERK; however, the expression of GRP78 stayed unaltered [21], indicating that other host factors might alleviate the intensity of unfolded protein responses (UPR). The endoplasmic reticulum (ER) is a major factor of glycoprotein synthesis, and the excessive expression of glycoprotein may activate UPR [37]. According to proteomic data, the expression of seven-transmembrane superfamily member 3 (TM7SF3), engaged in the attenuation of cellular stress and the subsequent UPR [38], was significantly induced. TM7SF3 is a downstream target of p53 [38], which is involved in innate immune response regulation, cell cycling, DNA repair, and apoptosis [39,40]. Although Xun Li et al. reported that overexpression of p53 positively regulated PRV replication both in vivo and in vitro [41], many questions were still hanging in the air: for example, the expression level and activation status of p53 during PRV infection and its contributions to TM7SF3 overexpression. Most importantly, the biological significance of TM7SF3 on UPR and virus replication were imperative to be elucidated.

Innate immune response, particularly type I interferon production and inflammatory cytokines secretion, is the first line to fight against pathogen invasion. PRRs recognize pathogen-associated molecular patterns (PAMPs) and then trigger innate immune responses. PRV is a common pathogen in multiple animal species and has even been isolated from human patients [42], thus attention should also be paid to the protein profile shifts in Vero lines. Vero cells are type I interferon-deficient, so inflammatory responses are emphasized in this paper. Our results showed that the NIK/NF-κB signaling pathway was markedly enriched by GO enrichment analysis. It was reported that the virulent PRV variant induced substantial lethal inflammatory response by TRL2, while the attenuated live vaccine of PRV lost the ability to activate an inflammatory response [13,43]. The abnormal inflammatory responses mediated by PRV variants might contribute to its pathogenicity. KEGG enrichment analysis showed that DEPs were significantly enriched in proteasome (Figure 5C); in particular, 28 DEPs were relevant to proteasome. Proteasome was reported to shape innate immune response and regulate the production of inflammatory cytokines [44]. Therefore, we proposed that PRV could modulate inflammatory response via regulating 26s proteasome non-ATPase regulatory subunits and other proteasome-related proteins expression. During PRV propagation, there are amounts of viral proteins synthesized in cells. We hold that the proteasome-related proteins may also be involved in useless or damaged protein degradation to maintain cellular homeostasis. In addition, cells may recruit proteasome to degrade viral proteins by ubiquitinating them and achieve an antiviral during PRV infection. Furthermore, the PRRs regulator tripartite motif-containing protein 40 (TRIM40) was also significantly down-regulated, indicating PRV might subvert innate immune responses by inhibiting PRRs activation. Additionally, the enzymatic activity of cGAS is tightly regulated by XRCC5 and XRCC6 to maintain immune homeostasis [45,46]. PRV is enveloped, and multiple processes all require lipids, such as virus-cell membrane fusion and virus budding. Our results showed that the elongation of very long chain fatty acids protein 1 (ELOVL1), involved in unsaturated fatty acid biosynthetic process [47], was significantly upregulated during PRV infection. This suggests that cellular lipids metabolism may take part in PRV propagation and pathogenesis. In addition, E3 ubiquitinprotein ligase NEDD4 was also significantly upregulated in PRV-infected cells. NEDD4 is essential for neural development and homeostasis of neural circuit excitability during neuronal ER stress [48], indicating this may be a protective mechanism to maintain cell homeostasis and normal biological functions during PRV infection. Taken together, the interactions between PRV infection and innate immune responses are complex and need further investigation.

This study systematically analyzed the protein profiles of PRV-infected Vero cells using a TMT-based proteomic analysis method. Eighty-nine upregulated and 871 downregulated proteins were identified, and biological analysis demonstrated that various cellular processes were involved in PRV-infected cells, including cellular processes, single-organism processes, metabolic processes, biological processes, regulation of biological processes, and so on. Unfortunately, our analysis of DEPs remains only instructional, and the elucidation of their biological functions is required. This research will help to deepen the understanding of the virus pathogenesis and host immune responses.

**Author Contributions:** Conceptualization, X.Y., S.X. and X.G. (Xin Guo); methodology, X.Y., S.X.; software, S.X.; validation, X.Y., S.X. and X.G. (Xin Guo); formal analysis, S.X. and X.Y.; investigation, X.Y.; resources, X.G. (Xin Guo); data curation, S.X.; writing—original draft preparation, S.X., and X.Y.; writing—review and editing, X.Y., S.X., D.C., R.J., H.K., X.G. (Xin Guo), L.Z., J.H., Y.Z., H.Y. and X.G. (Xinna Ge); funding acquisition, X.G. (Xin Guo) All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Natural Science Foundation of Beijing Municipal (Number: 6192014).

**Institutional Review Board Statement:** Not applicable for studies not involving humans or animals.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The proteomic data is available with the link: https://pan.baidu.com/ s/12D\_7pNG1VP79E\_aNfgi-LQ (Password: 39ht, accessed on 1 March 2022).

**Acknowledgments:** This research was supported by Natural Science Foundation of Beijing Municipal (Number: 6192014).

**Conflicts of Interest:** The authors declare that there are no conflict of interest.

### **References**

