Cyclophilin A protects mice against infection by influenza A virus

Our previous studies indicate that Cyclophilin A (CypA) impairs the replication of influenza A virus in vitro. To further evaluate the antiviral functions of CypA and explore its mechanism, transgenic mice with overexpression of CypA by two specific promoters with SPC (CypA-SPC) or CMV (CypA-CMV) were developed. After challenge with the A/WSN/33(H1N1) influenza virus, CypA-SPC and CypA-CMV transgenic mice displayed nearly 2.5- and 3.8-fold stronger disease resistance to virus infection, respectively, compared to wild-type animals. Virus replication, pathological lesions and inflammatory cytokines were substantially reduced in both lines of transgenic mice. In addition, after infection there was an upregulation of genes associated with cell migration, immune function, and organ development; and a downregulation of genes associated with the positive regulation of immune cells and apoptosis in the peritoneal macrophages of CypA-overexpressing transgenic mice (CypA+). These results indicate that CypA is a key modulator of influenza virus resistance in mice, and that CypA+ mice constitutes an important model to study the roles of CypA in the regulation of immune responses and infections.

. Construction and identification of transgenic mice over-expressing CypA. The founders of transgenic mice lines over-expressing CypA were identified at the protein level in lung tissue using western blots (A,B) and at the DNA level using PCR for the specific CypA-expression promoter in the genome (C,D). CMV and SPC represent the body and lung over-expression mice, respectively. The first filial generations were identified with PCR for the specific CypA-expression CMV (E) and SPC (F) promoters in the genome. M represents the DNA marker (from bottom to top: 100, 250, 500, 750, 1000 and 2000 bp) in panels C-F. Scientific RepoRts | 6:28978 | DOI: 10.1038/srep28978 or 8000 plaque forming units (PFUs) (n = 5 per dilution, 20 mice per group). At each challenge dose, the wild type C57BL/6 mice showed more severe clinical symptoms after infection, including weight loss, ruffled fur, and inactivity, compared to the CypA+ mice. Interestingly, the CypA-SPC mice showed slightly more weight loss than the CypA-CMV animals (Fig. 2). The MLD 50 values of A/WSN/33(H1N1) against wild type, CypA-SPC, and CypA-CMV mice were calculated to be 1561, 3892, and 5907 PFU, respectively. Therefore, the CypA-SPC and CypA-CMV mice were 2.5-and 3.8-fold more resistant, respectively, against infection with influenza A virus (Table 1). These results show that overexpression of CypA in the mouse contributed to disease resistance in the context of infection with influenza A virus.

Clinical symptoms of mice infected by A/WSN/33(H1N1).
To compare the clinical symptoms of mice after infection with a single challenge dose, wild type, CypA-SPC, and CypA-CMV mice were infected with 3000 PFU of A/WSN/33(H1N1). This infection dose was chosen on the basis that difference in clinical symptoms is the most easily distinguishable amongst the different mouse genotypes. Clinical signs were monitored over a 14-day period. After challenge, the wild type mice displayed more severe clinical signs with apparent inactivity, ruffled fur, decreased food intake, and higher weight loss than the CypA+ mice. The weights of the wild type mice begin to decrease at 3 d.p.i., losing up to 28-29% of initial weight, before gradually recovering. In contrast, the CypA+ The mice in the control group were i.n. inoculated with PBS for mock infection. The body weights of five mice in each group were monitored daily for 14 days and are expressed as a percentage of the initial value. The data represents the mean of five mice in each group. mice began to lose weight around 5 d.p.i., and the weight loss was significantly lower than that of the wild type mice (Fig. 3A). The CypA-SPC and CypA-CMV mice lost approximately 20% and 14% of their initial weights, respectively.

Virus replication in CypA+ and wild type mice. The level of virus replication observed after infection is
correlated to the susceptibility of the host to the pathogen. Therefore, the lung virus titers (LVTs) of the infected mice groups were tested after A/WSN/33(H1N1) infection at a challenge dose of 3000 PFU. We found that the virus was detected in the wild type, CypA-SPC, and CypA-CMV groups at 3 d.p.i. with LVTs of ~12000, ~8700, and ~6000 PFU/ml, respectively. At 5 d.p.i., the LVTs increased to ~100000, ~14000, and ~5000 PFU/mL in the wild type, CypA-SPC, and CypA-CMV mice, respectively ( Fig. 3B). At 7 d.p.i., the LVTs dropped to ~10000 and ~2000 PFU/mL in the wild type and CypA-SPC mice, respectively, and were undetectable in the CypA-CMV animals. The virus was also eliminated in the wild type and CypA-SPC mice by 9 d.p.i (Fig. 3B). These results show that the virus was suppressed in CypA+ mice compared to wild type. Moreover, the CypA-CMV mice displayed better levels of influenza virus inhibition than the CypA-SPC mice.
Pathological changes in CypA+ and wild type mice. The pathologic changes in the lungs of each experimental group were examined after a challenge dose of 3000 PFU. Three mice from each group were euthanized at time points of 3, 5, 7, and 9 d.p.i. and the lungs were harvested. Gross pathology with engorgement was observed at 3 d.p.i. Significant engorgement emerged at 5 d.p.i. and was most severe at 7-9 d.p.i. before gradually recovering. Notably, the wild type mice displayed more severe gross pathology than their CypA+ counterparts during the course of the infection (Fig. 4A). Consistent with the gross pathologic changes, the lung indices of the infected wild type mice were consistently higher than that of the CypA+ mice, but the difference was not always statistically significant (Fig. 4B). H&E staining was performed to evaluate the histopathological changes in mouse lungs after challenge. Severe bronchopneumonia was observed 5 d.p.i. (Fig. 5A-H), and severe interstitial pneumonia occurred at ~7 d.p.i. in the wild-type mice (Fig. S1). In contrast, the CypA+ mice showed comparatively less histopathological changes after infection, from mild to moderate/severe bronchopneumonia. The inflammatory cells also infiltrated in higher numbers in the lungs of the infected wild type mice compared to CypA+ mice ( Fig. 5A-H and Fig. S1). In accordance to the histopathological changes, the virus infected cells stained by M1 protein in the lungs of CypA+ mice was significant less than that of wild type ( Fig. 5I-P).
Cytokine responses in CypA+ and wild type mice. Cytokine levels in the sera of the infected mice were then tested to investigate the antiviral mechanisms of CypA. After infection with A/WSN/33(H1N1), IL-6 and MCP-1 levels began to rise at 3 d.p.i. and peaked around 7 d.p.i. before dropping at 9 d.p.i. in both the CypA+ and wild type mice. IL-6 and MCP-1 levels in the wild type mice were consistently higher than in CypA+ mice throughout the infection (Fig. 6A,B). In contrast, IFN-γ levels were consistently higher for CypA+ mice than their wild type counterparts (Fig. 6C).
Transcriptomics of peritoneal macrophages from infected CypA+ and wild type mice. Macrophages perform critical roles in both innate and adaptive immunity and are distributed throughout the host in every organ type. Thus, macrophages act as an important immunologic barrier for the host defense system against infection and injury 37 . A recent study shows that peritoneal macrophages appear to be The wild type, CypA-CMV, and CypA-SPC mice were inoculated i.n. with A/WSN/33(H1N1) virus at a dose of 3000 PFU. The body weights of five mice in each group were monitored daily for a 14-day observation period and are expressed as a percentage of the initial value (A). Three mice per group were euthanatized at 3, 5, 7, 9, 11, and 14 d.p.i., and the LVTs were titrated in MDCK cells by standard plaque assay to determine viral titers (B). The LVTs were compared with a paired-sample t-test and two-way ANOVA, respectively, * P < 0.05, * * P < 0.01, and * * * P < 0.001. The black, blue, and red asterisks represent the results compared to wild type, CypA-SPC, and CypA-CMV mice, respectively. more mature than bone marrow-derived macrophages and splenic macrophages 38 . Based on these considerations, peritoneal macrophages were used to study the potential immune mechanism of CypA. The transcriptomes of the peritoneal macrophages of wild type and CypA+ mice were investigated ex vivo. Eight hours after infection with the A/WSN/33(H1N1) virus at a multiplicity of infection (MOI) of 2, 197 specific up-and 91 down-regulated genes were identified in the infected peritoneal macrophages of CypA-CMV mice compared to wild type (Fig. 7A,B). The functions of these differently expressed genes (DEGs) were then analyzed by DAVID Bioinformatics Resources with the Gene Functional Classification method. The upregulated genes were associated with cell migration, immune function, and organ development (Fig. 7C). In contrast, the downregulated genes were associated with the positive regulation of immune cells and apoptosis (Fig. 7C). Interestingly, there were also 96 up-and 83 down-regulated DEGs in the uninfected peritoneal macrophages of CypA-CMV mice compared to wild type (Fig. 7A,B), and phenotypic differences between the two mouse lines as a result of these DEGs (ie. immune responses, organ development, etc.) will be an interesting topic of study in the future.

Discussion
CypA is the most abundant intracellular protein in the cyclophilin family and plays important roles in regulating immune responses, involving T-cell activation, pro-inflammatory responses, and the innate immune system 7,9,10,11,15,16 . CypA expression is correlated with apoptosis and the development of some diseases 14,[18][19][20] . In addition, CypA contributes to the infection and infectivity of some pathogens, such as HIV-1, HCV, HBV, VSV, and L. major 5,[25][26][27][28][29][30] , though it also suppresses the replication of rotavirus 31 , infectious bursal disease virus, TBSV, mouse cytomegalovirus, and influenza virus 1,26,[32][33][34] . However, nearly all of these data come from in vitro experiments and lack in vivo support from animal models. Hence, it is necessary to confirm these conclusions in vivo, which would help to further understand the functions of CypA. CypA+ transgenic mice would undoubtedly help to resolve these discrepancies.
In the present study, we found that CypA+ mice exhibit antiviral activity against influenza virus challenge, and that overexpression of CypA resulted in the upregulation of antiviral genes in the host, likely contributing to disease resistance. The lung indices, gross pathology, and histopathological changes we observed demonstrated that inflammatory responses were significantly inhibited in CypA+ mice after infection (Figs 4 and 5). Cytokine responses, including IL-6, MCP-1, IFN-γ , and TNF-α levels, are closely related to the host damage caused by the influenza virus infection in the host [39][40][41][42][43][44][45] . Interestingly, the serum IL-6 and MCP-1 levels were obviously repressed in the CypA+ mice, whereas changes in IFN-γ expression were milder and more prolonged than that of wild type mice (Fig. 6). However, there was no statistical significance for the cytokine concentrations between the CypA+ and wild type groups as determined by two-way ANOVA, which may be due to the limited numbers of animals (n = 3) at each detection timepoint. Our previous studies show that CypA inhibits influenza virus replication in vitro through the direct interaction between CypA and the viral M1 protein 22,35 . Hence, the antiviral function of CypA was thought to occur via two routes, involving the virus itself and the host.
IFN-γ secretion from CypA+ mice was found to be higher than their wild type counterparts after influenza virus infection. IFN-γ is produced by NK cells and specific subsets of T-lymphocytes (including CD4 Th1 and CD8 cytotoxic T-lymphocyte effector cells), is known to be critical for innate and adaptive immunity against viral, some bacterial, and protozoan infections 43,44 . IFN-γ acts as an immunostimulatory and immunomodulatory factor regulating immune responses in the CypA+ mice. Furthermore, IFN-γ is an important activator of macrophages and an inducer of Class I major histocompatibility complex (MHC) molecule expression 46 .
The host requires a robust immune response to suppress influenza virus infection 47 . Transcriptomic analyses of mouse peritoneal macrophages revealed that the upregulated genes associated with cell migration and leukocyte chemotaxis, which may increase the efficiency of immune cell transport between infected sites and other organs. Several leukocyte-mediated immunity-associated genes were also up-regulated (Fig. 7), which may increase the utilization rate for the immune cells. The up-regulation of genes in immune system development would be important in helping the host mount a robust inflammatory response, whereas up-regulation of genes related to respiratory tube development helps the host to repair tissue damage during the infection. In contrast, a down-regulation of genes associated with positive regulation of immune cell proliferation and activation decrease the chance of aberrant cytokine responses. Suppressed IL-6 and MCP-1, and prolonged IFN-γ levels may be related to changes in the levels/activities of host factors that are influenced by CypA in CypA+ mice. Furthermore, apoptosis plays an important role in tissue damage during influenza virus infections 46 . Therefore, the up-regulation of genes that down-regulate apoptosis delays the formation of pathological lesions. The up-regulation of other genes related to immune signaling, such as ubiquitination and phosphorylation, may also play important regulatory and antiviral functions, topics that should be further investigated.
On one hand, CypA increases the efficiency of transporting and utilizing immune cells between infected sites and other organs through the regulation of leukocyte chemotaxis, and decreases immunologic injury by reducing apoptosis. On the other hand, CypA enhances immune system development to repair impairments caused by the infection over time, while decreasing excessive immune provocations by down-regulating immune cell proliferation and activation. In addition, CypA promotes the recovery of damaged respiratory tissues by regulating the up-DEGs of respiratory tube development. These systematic antiviral functions might be accurately regulated by the immune signal pathways, including ubiquitination, phosphorylation, and other signaling pathways in hosts overexpressing CypA. Interestingly, the CypA-CMV mice were more resistant to disease than the CypA-SPC mice, and this was reflected in the MLD 50 calculations during infection with A/WSN/33(H1N1). The over-expression of CypA from the CMV promoter all over the body of the host likely increased its disease resistance threshold against influenza virus infection.
This study is the first to investigate CypA antiviral effects in vivo, using influenza virus as an example. Due to the wide range of known roles for CypA in the regulation of immune responses as well as infectious diseases, the transgenic mouse model opens the door in confirming previous in vitro observations, as well as determining the exact roles and mechanisms of CypA in host immunity and pathogenesis of infectious diseases. In addition, the CypA+ transgenetic animal is undoubtedly paving the way for animal breeding with the design that simultaneously regulate immune responses and limit pathogen replication, which might be a better strategy than only target on virus itself as previous study 48 .

Materials and Methods
Construction and identification of transgenic mice over-expressing CypA. The cyclophilin A (CypA, IA) gene (GenBank No. AB451307) was cloned into the expression vectors pCMV-Myc and pCD-NA3.1-SPC under control of the CMV and SPC promoters, respectively. The pDNA3.1-SPC construct was generated by our laboratory, in which the CMV promotor of pCDNA3.1 vector was replaced by the SPC promotor. The transgenic C57BL/6 mice were generated by the microinjection method as previously described 49 . The CMV or SPC promoters were designed to express the target gene over the entire body and specifically in the lungs of the transgenic mice, respectively. Genomic DNA from the transgenic mice was extracted from tail tissues  Mock-infected control animals were inoculated i.n. with 50 μ L PBS. The mice were monitored daily for a 14-day observation period, and the body weights were measured and expressed as a percentage of the initial value, calculated at day 0. Mice that lost ≥ 25% of its preinoculation body weight was pronounced dead and lose ≥ 30% of its preinoculation body weight were euthanized according to animal ethics guidelines. MLD 50 values were calculated by the Reed and Muench method 50 and expressed as the PFU value corresponding to 1 MLD 50 .

Pathogenesis comparison in influenza-infected transgenic and wild type mice. Twenty-three
wild type, CypA-CMV, and CypA-SPC mice were inoculated i.n. with A/WSN/33(H1N1) virus at a dose of 3000 PFU. The mice were monitored daily for general behavior and clinical signs, including food intake, body weight, and inactivity. Five stable mice in each group were segregated for weighing. Three mice from each group were euthanized at 3, 5, 7, 9, 11, and 14 days post-infection (d.p.i.), and their lungs were collected, weighed, and homogenized using a QIAGEN TissueLyser II machine (30 cycles/s, 4 min) in 1 mL of cold PBS under sterile conditions. Solid debris was pelleted by centrifugation at 5,000 × g for 10 min, and the homogenates were used for virus titrations in MDCK cells as previously published 51 . Their sera were used for cytokine detection by the cytometric bead array method using a Mouse Inflammation Kit (BD, USA). The lung index was calculated as the lung wet weight/body weight × 100. Additionally, a portion of each left lung lobe was fixed in 10% buffered formalin, embedded in paraffin, and 5-μ m sections were stained with hematoxylin-eosin (H&E) and the monoclonal antibody against M1 protein for histopathological and immunohistochemical (IHC) analyses, respectively.
RNA-seq analysis. The peritoneal macrophages of wild type and CypA-CMV mice were isolated from the peritoneal cavities and maintained in RPMI 1640 containing 10% fetal bovine serum at 37 °C until use, as described previously 52  To analyze the Illumina reads, TopHat and Cufflinks were used to investigate the differential gene expression (DGE) profiles and changes in transcript abundance in the CypA over-expressing and wild type peritoneal macrophages with or without A/WSN/1933(H1N1) infection. Four files of transcriptome data from the IA-WSN, WT-WSN, IA-Mock (Mk), and WT-Mk groups were aligned to the UCSC Rhesus Macaque genome build in preparation for differential expression analysis. The four files were processed through Cufflinks to assemble the aligned RNA-seq reads into transcripts and estimate the abundances in FPKM of the paired-end reads. The Cufflinks q-value was the false discovery rate (FDR)-adjusted p-value of the uncorrected test statistic; the q-value used in this study was 0.05. The significance status was "yes" when p > q after Benjamini-Hochberg correction for multiple-testing. Cuffmerge was then used to create a single transcript dataset from the multiple reconstructions. Two runs were conducted using the IA-WSN vs. WT-WSN and the IA-Mk vs. WT-Mk datasets and the Cuffdiff program to test for differential expression and regulation among the two compared datasets. All significant hits were used to analyze the specific up-and down-regulation of gene expression of the A/WSN/1933(H1N1)-infected peritoneal macrophages from the CypA transgenic mice by Venn diagrams 53 . The gene functional classification of the associated up-and down-regulated genes were analyzed by DAVID Bioinformatics Resources 54 , and the Heat-map was analyzed using Cluster 3.0 and TreeView 1.6.0 software (Michael Eisen, Berkeley, CA).

Ethics statement. The animal research was approved by the Chinese Academy of Sciences of Research Ethics
Committee, under approval number PZIMCAS2013001, and complied with the Beijing Laboratory Animal Welfare and Ethical Guidelines of the Beijing Administration Committee of Laboratory Animals.