Jiongjiong Wang1 | Jing Sun1,2,3,4 | Jiao Hu5 | Chengming Wang6 | Richard A. Prinz7 | Daxin Peng2,3,4,5 | Xiufan Liu3,4,5 | Xiulong Xu1,2,3,4
Abstract
The newly reassorted IAV subtypes from zoonotic reservoirs respond poorly to cur- rent vaccines and antiviral therapy. There is an unmet need in developing novel anti- viral drugs for better control of IAV infection. The cellular factors that are crucial for virus replication have been sought as novel molecular targets for antiviral therapy. Recent studies have shown that Janus kinases (JAK), JAK1, and JAK2, play an im- portant role in IAV replication. Leflunomide is an anti-inflammatory drug primarily used for treating rheumatoid arthritis (RA). Prior studies suggest that A77 1726, Microscopes and Cell Imaging Systems the active metabolite of leflunomide, inhibits the activity of JAK1 and JAK3. Our cur- rent study aims to determine if A77 1726 can function as a JAK inhibitor to control IAV infection. Here, we report that A77 1726 inhibited the replication of three IAV subtypes（H5N1, H1N1, H9N2）in three cell types (chicken embryonic fibroblasts, A549, and MDCK). A77 1726 inhibited JAK1, JAK2, and STAT3 tyrosine phos- phorylation. Similar observations were made with Ruxolitinib (Rux), a JAK-specific inhibitor. JAK2 overexpression enhanced H5N1 virus replication and compromised the antiviral activity of A77 1726. Leflunomide inhibited virus replication in the lungs of IAV-infected mice, alleviated their body weight loss, and prolonged their survival. Our study demonstrates for the first time the ability of A77 1726 to inhibit JAK2 activity and suggests that inhibition of JAK activity contributes to its antiviral activity.
KEYWORDS:A77 1726, dihydroorotate dehydrogenase, influenza A virus, leflunomide, p70 S6 kinase, protein tyrosine kinases
1 | INTRODUCTION
Influenza is a highly contagious acute respiratory disease that often presents as local outbreaks or seasonal epidemics and causes substantial morbidity and mortality.1 Approximately 3 to 5 million patients worldwide get severe influenza infec- tions annually, and 250 000-500 000 patients will die of this illness each year. Since the 2009 swine flu pandemic, strains of that outbreak continue to cause serious illness and high mortality, particularly among young adults and children. The influenza virus belongs to the Orthomyxoviridae family and is divided into A, B, C, and D types, which differ in host ranges and pathogenicity.2 Influenza A virus (IAV) infects a wide range of avian and mammalian hosts, whereas influ- enza B virus infects humans and seals only. Influenza C virus causes a mild respiratory infection and is not considered as a cause of epidemics.2 Influenza D viruses primarily affect cattle and are not known to infect or cause illness in people.2 IAV can be further classified into various subtypes based on the combination of two viral surface glycoproteins, for exam- ple, hemagglutinin (HA) and neuraminidase (NA). A total of 16 HA and 9 NA subtypes from birds as well as additional H17 and H18 plus N10 and N11 from bats have been iden- tified so far.1 Several reassortant IAV genotypes, including H5N1, H7N9, H5N6, and H10N8, cause sporadic fatal infec- tions in humans.3-5 H7N9 is a highly pathogenic avian subtype that emerged in China in late 2013 and early 2014 during a second wave of influenza. Severe infections of H10N8 and H5N1 subtypes continue to be reported.
Vaccination against predicted IAV strains has been used as a prophylaxis measure to prevent influenza. However, pre- dicting the annual circulating IAV strains and preparing suffi- cient quantities of vaccines are challenging. Several antiviral drugs including NA inhibitors and M2 ion channel blockers provide limited protection from the potentially fatal infection of these newly emerged IAV strains. There is an unmet need for more effective antiviral drugs. An alternate approach to designing these drugs is to target the cellular factors that are crucial for virus replication.6 Among them, JAK tyrosine kinases,7-9 mTOR,10,11 and dihydroorotate dehydrogenase (DHO-DHase),12 a rate-limiting enzyme in pyrimidine nu- cleotide synthesis, have been implicated in playing import- ant roles in IAV replication. Although the inhibitors of these enzymes are capable of inhibiting IAV replication in vitro, many do not have a robust antiviral activity in vivo.Leflunomide is an anti-inflammatory drug primarily used for treating RA. A77 1726, the active metabolite of lefluno- mide, inhibits the activity of protein tyrosine kinases (PTKs) such as PDGF receptor, the members of the Src, and JAK family tyrosine kinases.13-20 Subsequent studies demon- strated that A77 1726 is also an inhibitor of DHO-DHase.
Inhibition of DHO-DHase activity decreases intracellular pyrimidine nucleotide levels in vitro in cell culture but does not significantly decrease tissue pyrimidine nucleotide levels in vivo due to the presence of serum uridine derived from degraded nucleic acids.13,14 We recently reported that in an in vitro kinase assay and in cell culture, A77 1726 is able to inhibit the activity of p70 S6 kinase 1 (S6K1), a serine/ threonine kinase downstream of mTOR, with an IC50 value of approximately 50-75 µM.21 A77 1726 has a wide spectrum of antiviral activity. It inhibits the replication of cytomegalo- virus (CMV),22-24 BK virus,25,26 Epstein-Barr virus (EB),27 and respiratory syncytial virus.28 Whether A77 1726 inhibits virus replication by inhibiting pyrimidine nucleotide syn- thesis remains controversial. Here, we report that A77 1726 inhibited IAV replication in vitro in three cell types and in vivo in the lungs of IAV-infected mice, and that the antivi- ral activity of A77 1726 was largely mediated by inhibiting JAK2 activity.
2 | MATERIALS AND METHODS
2.1 | Reagents
A77 1726 and Ruxolitinib (Rux) were purchased from Selleck Inc (Shanghai, China). Leflunomide was purchased as tablets from Cinkate Pharmaceutical Inc (Suzhou, Jiangsu Prinvince, China) and dissolved in water for in vivo experiments. Brequinar sodium (BQR), a potent inhibitor of DHO-DHase, was kindly provided by DuPont Corporation (Wilmington, DE). Antibodies for phospho-tyrosine (P-Y-1000 #8954), STAT3 (#9139),pSTAT3Y705(#9145), S6 (#2217), and phosphor-S6S235/236(#4858) were purchased from Cell Signaling Technology (Danvers, MA). Anti-phospho-JAK1 antibody was purchased from American Research Products, Inc (Waltham, MA). Anti-JAK1 antibody was purchased from Sigma Aldrich (St. Louis, MO). PF-4708671 and an- tibodies against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and β-actin were purchased from Santa Cruz Biotechnology, Inc (Santa Cruz, CA). Monoclonal or poly- clonal antibodies against NP, M1, and HA proteins were prepared by immunizing mice with purified recombinant proteins. Anti-PB2 antibody (#GTX125926) was purchased from GeneTex Inc (Irvine, CA).
2.2 | Cell culture and viruses
MDCK (Madin-Darby canine kidney) and A549 (a human lung cancer cell line of alveolar epithelial cell origin) cells were purchased from the American Tissue Culture Collection (Manassas, VA). Cells were grown in DMEM containing 10% of fetal bovine serum (FBS). Chicken embryo fibro- blasts (CEF) were prepared and grown in M199 containing 4% of fetal bovine serum. The A/mallard/Huadong/S/2005 H5N1 virus was isolated from poultry.29 Viruses were plaque purified three times in MDCK cells. A human influenzavirus A/California/ 04/09 (CA/09, H1N1) was kindly provided by Dr Jinhua Liu (China Agricultural University, Beijing, China). The source of IAV H9N2 virus (AIV Ck/SH/F/98) has been reported previously.30 All IAV strains were prop- agated in 10-day-old specific-pathogen-free embryonic chicken eggs. The virus titers were measured by a 10-fold serial dilution (101 to 109, and each dilution (105-109) was inoculated into MDCK cells. The 50% tissue culture infec- tion dose (TCID50/ml) was calculated according to the Reed and Muench method.
2.3 | Virus growth
To determine the effect of A77 1726 on virus growth, CEF infected with 0.01 multiplicity of infection (MOI) H5N1 virus were incubated with various concentrations of A77 1726 minus or plus uridine (200 µM). Conditioned media were collected at 12, 20, and 28 hours post inoculation (hpi).The TCID50 values were determined according to our pre- vious publications.29,30 Data represent the mean ± standard deviation (SD) of four independent experiments. The effects of PF-4708671, BQR, and Ruxolitinib on virus growth were similarly done. The inhibitory effects of A77 1726 on IAV growth in MDCK and A549 cells were similarly carried out as above.
2.4 | RT-qPCR analyses
CEF cells seeded in 6-well plates were infected with 0.01 MOI H5N1 virus, and then, incubated in the absence or pres- ence of A77 1726 (200 µM) plus or minus uridine (200 µM) for 16 hours. Total RNA was extracted by using a Qiagen RNA extraction kit and quantified for RNA concentrations. Viral RNA levels were measured by using a one-step reverse- transcription FRET-PCR to amplify the M gene according to our recent publication.31
2.5 | JAK2 transfection
MDCK cells were transiently transfected with the empty vec- tororpBABE-JAK2,a retroviral vector encoding JAK2 tagged with a fragment of yellow fluorescence protein (YFP) (kindly provided by Dr Eric Chang, Baylor College of Medicine, Houston).32 After incubation for 36 hours, the cells were left uninfected or infected with the indicated MOI of H5N1 virus, and then, incubated for 12 hours. Alternatively, the cells were infected with H5N1 virus (0.1 MOI), and then, incubated in the absence or presence of various concentrations of A77 1726 (0, 25, 50, 100 µM) for 12 hours. Cell lysates were pre- pared and analyzed for JAK2 and STAT3 phosphorylation and for the levels of viral NP and M1 proteins. The virus titers in the conditioned media were collected and analyzed for the TCID50 values. Relative virus titers were calculated as the percent of controls. The results represent the mean ± SD of three independent experiments.
2.6 | Western blotting
CEF, MDCK, and A549 cells seeded in 6-well plates were infected with H5N1 (0.01 MOI), H1N1, or H9N2 (0.1 MOI each), and then, incubated in the absence or presence of vari- ous concentrations of A77 1726 and/or uridine (200 µM) for 16 hours (CEF) or 24 hours (A549 and MDCK). Cells were harvested and lysed in NP-40 lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 5 mM EDTA, 5 mm EGTA, 1 mM NaF, 2 mM sodium vanadate, the cocktail of protease inhibitors (1X) (Pierce Chemical Co., Rockford, IL), and 2 mM sodium pervandate). Cell lysates were prepared and analyzed for the expression of viral proteins (PB2, HA, NP, and M1) or other indicated proteins (P-Y-1000, pJAK2, JAK2, pSTAT3, STAT3, and β-actin) with their specific antibodies, followed by horseradish peroxidase-conjugated goat anti-rabbit IgG or goat-anti-mouse IgG and SuperSignal Western Pico enhanced chemiluminescence substrate (Pierce Chemical Co., Rockford, IL).
2.7 | Cell proliferation assay
To exclude the possibility that the inhibitory effect of A77 1726 on virus replication was due to inhibition of cell prolif- eration, cells were incubated with various concentrations of A77 1726 minus or plus uridine (200 µM) for 24 hours, and then, analyzed for cell proliferation by using a CellTiter-Glo kit (Promega, WI) as reported.33,34
2.8 | Animal study
Use of animals was approved by the Institutional Animal Care and Use Committee of College of Veterinary Medicine, Yangzhou University. Female C57BL/6 mice were pur- chased from the Center of Experimental Animals of College of Veterinary Medicine, Yangzhou University. All mice were maintained on a 12-hours light/dark cycle and housed in ven- tilated cages at an ambient temperature of 23ºC. Mice were fed ad libitum on a normal chow diet (NCD). Leflunomide was dissolved in distilled water. Mice (8-weeks-old) were given water or leflunomide by gavage. Twelve hours after the first treatment, mice under diethyl ether anesthesia were mock-infected intranasally with PBS or infected with H5N1 virus (1 × 105 pfu in 50 µL PBS per mouse) by instillation. To determine the antiviral effect of leflunomide on virus replication, mice were treated with water or leflunomide (20 mg/kg/day, gavage) (5 mice/group) daily for 3 or 5 days. Mice were euth- anized. Lung tissues were homogenized in NP-40 lysis buffer for Western blot or in PBS for analyzing virus loads. To de- termine the anti-inflammatory effect of leflunomide, the lung tissues from untreated mice or from mice treated with leflu- nomide (5 mice/group) for 3 or 5 days were fixed in 4% of paraformaldehyde and embedded within 48 hours. The sec- tions of paraffin-embedded blocks were stained with hema- toxylin and eosin. The numbers of infiltrates in 5 randomly selected fields (40X) in different areas of the lung (top, bot- tom, left, right, and central region) were counted. The infil- tration was scored as 1, 0-10 cells/field; 2, 11-30 cells/field; 3, 31-60 cells/field; 4, 61-100 cells/field; 5, >100 cells/field. The mean scores in control and leflunomide-treated groups (10 mice per group) were calculated, statistically analyzed by using an unpaired student t test and plotted in a bar graph.To determine the therapeutic effect of leflunomide, mice in- fected with the H5N1 virus were treated with water or leflu- nomide (35 mg/kg/day) (10 mice/group) daily for 7 days by gavage. Body weights were recoded daily.
2.9 | Statistical analysis
Differences in the virus titers, viral mRNA levels, cell pro- liferation, and the scores of inflammatory cell infiltration were statistically analyzed by using an unpaired Student t test. Differences in the body weight of untreated and lefluno- mide-treated mice were analyzed using a repeated measures ANOVA test. Differences in the survival of untreated and
leflunomide-treated mice were statistically analyzed by using a Log-Rank test. A P value of <.05 was considered statisti- cally significant. All statistics were done with SigmaPlot 11 software (Systat Software, Inc, San Jose, CA).
3 | RESULTS
3.1| A77 1726 inhibits the replication of three IAV subtypes
We first determined the ability of A77 1726 to inhibit virus growth. As shown in Figure 1A, the virus tiers were very low in the conditioned media of H5N1 virus-infected CEF at 12 hours post-infection (hpi) in the absence or presence of uridine (200 µM). A77 1726 lowered the titers of H5N1 virus in the conditioned media at 20 and 28 hpi in a dose- dependent manner (Figure 1A). The virus titers of H5N1 virus in the conditioned media of A77 1726-treated cells were not affected by addition of exogenous uridine (200 µM) (Figure 1B). A77 1726 dramatically reduced the levels of viral M1, NP, and HA proteinsinCEF infected with H5N1 virus at 16 hpi (Figure 1C). Uridine (200 µM) slightly increased the levels of HA, M1, and NP proteins, compared to the coun- terparts in the cells treated with A77 1726 at 50 or 100 µM but not 200 µM. Inhibition of viral protein synthesis by A77 1726 (200 µM) were not affected at all by even much higher concentrations of uridine up to 1 mM (Figure S1). RT-PCR revealed that A77 1726 reduced viral RNA levels of H5N1 virus in CEF by 91%,whereas A77 1726 in combination with uridine reduced RNA replication by 67% (Figure 1D). We next determined the inhibitory effect of A77 1726 on H5N1 virus replication in two mammalian cell lines, MDCK and A549 cells. Confluent monolayers of MDCK and A549 cells plated at approximately 90% were infected with H5N1 virus (0.1 MOI), and then, incubated for 24 hours in the absence or presence of the indicated concentrations of A77 1726 in serum-free medium. A77 1726 lowered the titers in the conditioned media of H5N1 virus-infected MDCK (Figure 1E) and A549 (Figure 1G) cells and decreased the levels of viral PB2, NP, and M1 proteins in a dose-dependent man- ner (Figure 1F,H). The antiviral effect of A77 1726 against H5N1 virus in MDCK andA549 cells were not compromised by addition of exogenous uridine (200 µM) (Figure 1E-H).We then tested if A77 1726 inhibited the replication of two other IAV subtypes, H1N1 and H9N2 viruses. A77 1726 low- ered the titers of H1N1 (Figure 2A) and H9N2 (Figure 2B) viruses in CEF, MDCK, and A549 cells and reduced the levels of PB2, NP, and M1 proteins in these three cell types (Figure 2C,D) in a dose-dependent manner. The antiviral ac- tivity of A77 1726 is not due to its antiproliferative activity since A77 1726 minimally inhibited the proliferation of CEF, MDCK, and A549 cells in the absence or presence of uridine (Figure 2E).
3.2 | BQR and PF-4708671 do not inhibit H5N1 virus replication
To investigate if A77 1726 exerted its antiviral effect by inhibiting S6K1 or DHO-DHase activity, we tested if PF-4708671 and BQR,a potent and specific inhibi- tor of S6K1 and DHO-DHase,respectively, could sup- press IAV replication. As shown in Figure 3A,neither PF-4708671 (5 µM) nor BQR (5 µM) alone or in combi- nation blocked IAV-induced cytopathic effect (CPE) in CEF 24 hpi. Consistently, PF-4708671 and BQR alone or in combination did not substantially reduce the levels of HA, NP, and M1 proteins in CEF at 24 hpi (Figure 3B) and did not lower the virus titers in the conditioned media of H5N1 virus-infected CEF (Figure 3C); whereas A77 1726 alone or in combination with uridine, included as positive controls, dramatically reduced the levels of these proteins (Figure 3B). BQR and PF-4708671 at very high concentrations up to Hospice and palliative medicine 20 µM each were unable to inhibit H5N1 virus replication in CEF, A549, and MDCK cells (Figure 3D,E). PF-4708671 inhibited S6 phosphorylation in a dose-dependent manner in three cell types (Figure 3E).
FIGURE 1 A77 1726 inhibits H5N1 virus replication. A and B, CEF cells infected with H5N1 virus (0.01 MOI) were incubated in the
absence or presence of the indicated concentrations of A77 1726 (A) or A77 1726 plus uridine (200 µM). B, Conditioned media were collected at the indicated time and analyzed for the TCID50 values. The results represent the mean ± SD of four independent experiments. *P < .05; **P < .01, compared to the untreated or uridine-treated controls. C, CEF seeded in 6-well plates were infected with H5N1 virus, and then, incubated in the absence or presence of indicated concentrations of A77 1726 minus or plus uridine (200 µM) for 16 hours. Cell lysates were prepared and analyzed for HA, M1, and NP expression by Western blot with their specific antibodies. Actin was detected as a loading control. D, A77 1726 inhibits viral mRNA transcription. CEF infected with H5N1 virus were incubated in the absence or presence of A77 1726 (200 µM) and/or uridine (200 µM)for 16 hours. Total RAN was extracted and quantified, followed by real-time RT-PCR analysis of the M gene of IAV. **P < .01, compared to the untreated controls. *P < .05, compared to A77 1726-treated cells. E-H, MDCK (E and F) and A549 (G and H) cells infected with H5N1 virus
(0.1 MOI) were incubated in the absence or presence of the indicated concentrations of A77 1726 minus or plus uridine (200 µM). Conditioned
media were collected 24 hpi and analyzed for TCID50 values (E and G). The results represent the mean ± SD from three independent experiments.
Cell lysates were prepared and analyzed for PB2, NP, and M1 expression by Western blot (F and H). Actin was included as a loading control
3.3 | Evidence that A77 1726 inhibits IAV replication by inhibiting JAK activity
Recent studies have identified multiple cellular factors includ- ing JAK1 and JAK2 as critical cellular components involved in IAV replication.7,35,36 Our previous studies have shown that A77 1726 inhibits the activity of JAK1 and JAK3 in an in vitro kinase assay.18,19 Here, we examined the ability of A77 1726 to inhibit JAK activity in CEF and MDCK cells. A77 1726 in- hibited tyrosine phosphorylation of JAK1, JAK2, and STAT3 phosphorylation in H5N1 virus-infected CEF (Figure 4A) and JAK1 and STAT3 in uninfected CEF (Figure 4B). A77 1726 (200 µM) inhibited total protein tyrosine phosphorylation in H5N1-infected CEF (Figure 4C) as well as in uninfected CEF (data not shown) in a dose-dependent manner (Figure 4C). A77 1726 inhibited JAK1, JAK2, and STAT3 tyrosine phosphoryl- ation in H5N1-infected (Figure 4D) or uninfected (Figure 4E) MDCK cells in a dose-dependent manner. H5N1 virus did not significantly induce JAK2 and STAT3 tyrosine phosphoryla- tion in A549 and MDCK cells (Figure S2). Ruxolitinib (Rux), a JAK inhibitor,7 was included as a positive control. It low- ered the levels of M1 and NP proteins and inhibited JAK2 and STAT3 tyrosine phosphorylation in CEF infected with H5N1 virus in a dose-dependent manner (Figure 4F). Rux also dose- dependently inhibited total protein tyrosine phosphorylation (Figure 4G). Of note, A77 1726 and Rux inhibited JAK2 au- tophosphorylation less effectively than the phosphorylation of its substrates such as STAT3, a phenomenon also reported in our prior studies (18, 19). Rux dose-dependently decreased the titers of H5N1 virus in the conditioned media of CEF (Figure 4H) but did not cause significant cytotoxicity in CEF (Figure 4I).
3.4 | JAK2 overexpression enhances IAV replication and attenuates A77 1726-mediated antiviral activity
To further test our hypothesis that inhibition of JAK ty- rosine kinase activities contributed to A77 1726-mediated antiviral effects, we first determined if overexpressing JAK2 would accelerate IAV replication. As shown in Figure 5A, JAK2 overexpression significantly increased JAK2 autophosphorylation and the levels of viral NP and M1 pro- teins (Figure 5A). JAK2 overexpression also significantly increased H5N1 virus titers in the conditioned media of MDCK cells by >100% (Figure 5B). We then investigated if JAK2 overexpression would attenuate the antiviral effect of A77 1726. As shown in Figure 5C, A77 1726 dose-de- pendently lowered the levels of viral NP and M1 proteins in H5N1 virus-infected MDCK cells transfected with JAK2 at a significantly lower magnitude than in the cells transfected with pcDNA3.1 (Figure 5C). The virus titers in the conditioned media collected from untreated or A77 1726-treated, JAK2-overexpressed cells were significantly higher than in their corresponding counterparts in pcDNA-transfected cells (Figure find more 5D).
3.5 | Leflunomide controls IAV infection in vivo
Finally, we determined if the antiviral effect of A77 1726 observed in vitro could be translated in an in vivo setting in a mouse model. As shown in Figure 6A, leflunomide treatment significantly lowered the levels of three viral proteins, PB2, NP, and M1 of H5N1 virus in the lung tissue of leflunomide- treated mice, compared to that in the untreated controls on day 3 (Figure 6A) and 5 (Figure 6B) post-infection. Consistently, the titers of H5N1 virus were significantly lower in the lung tissue of leflunomide-treated mice than in the untreated con-
trols on day 3 (Figure 6C) and 5 (Figure 6D) post-infection. Leflunomide treatment also dramatically reduced inflam- matory cell infiltration on day 3 and 5 after virus infection in the lung tissue of H5N1 virus-infected mice (Figure 6E). The mean body weight of leflunomide-treated mice was sig- nificantly higher than that of untreated mice (Figure 6F). The survival time of leflunomide-treated mice was significantly longer than that of untreated mice (Figure 6G). Leflunomide treatment significantly decreased the levels of viral proteins
FIGURE 2 A77 1726 inhibits H1N1 and H9N2 virus replication. CEF, MDCK, and A549 cells infected with H1N1 (A and C) or H9N2
(B and D) viruses (0.01 MOI for CEF, 0.1 MOI for A549 or MDCK cells) were incubated in the absence or presence of the indicated concentrations of A77 1726 for 24 h. TCID50 values of H1N1 (A) and H9N2 (B) were determined in MDCK cells. The results represent the mean ± SD of three
independent experiments. *P < .05; **P <.01, compared to the untreated control. C and D, Cells were harvested and analyzed for M1 and NP
expression by Western blot. The results represent one of three experiments with similar results. E, A77 1726 has little effect on cell proliferation.CEF, MDCK, and A549 cells seeded in 96-well plates were incubated in the absence or presence of the indicated concentrations of A77 1726 minus or plus uridine (200 µM) for 24 hours. Cell proliferation was analyzed as described in the Materials and Methods. Data are the mean ± SD of a representative experiment in triplicate. The experiments were repeated twice with similar results. *P < .05; **P < .01, compared to the untreated controls and virus loads (Figure S3A,B) in the lungs of H1N1 virus- infected mice (FigureS3C,D), reduced inflammatory infiltra- tion and edema in the lung tissue of H1N1 virus-infected mice (Figure S3E), and significantly alleviated the body weight loss (Figure S3F), prolonged their survival (Figure S3G).
4 | DISCUSSION
Prior studies suggests that A77 1726 inhibits the activ- ity of JAK1 and JAK3.18,19 Our present study provides further evidence that A77 1726 was able to inhibit JAK2
FIGURE 3 Inability of PF-4708671 and BQR to inhibit H5N1 virus replication. CEF were left uninfected or infected with H5N1 virus,
and then, incubated in the absence or presence of PF-4708671 (5 µM) or BQR (5 µM), A77 1726 (200 µM), or A77 1726 (200 µM) plus uridine
(200 µM) for 24 hours. Images of cytopathic changes were taken (A). Cell lysates were analyzed for HA, NP, and M1 expression by Western blot (B). Conditioned media were collected and analyzed for the TCID50 values (C). D and E, A549, CEF, and MDCK cells infected with H5N1 virus (0.01 MOI for CEF, 0.1 MOI for A549 and MDCK) were incubated in the absence or presence of indicated concentrations of BQR (D) or PF-
4708671 (E) for 24 hours. Cell lysates were prepared and analyzed for M1 and NP expression and S6 phosphorylation by Western blot. The results represent one of three experiments with similar results
FIGURE 4 Evidence that A77 1726 inhibits IAV replication by inhibiting JAK2 activity. A-C, CEF cells were infected with H5N1 virus
(A and C) or uninfected (B), and then, incubated for 16 hours in the absence or presence of the indicated concentration of A77 1726. Cell lysateswere prepared and analyzed for JAK2 and STAT3 tyrosine phosphorylation (A and B) and for the expression of M1 and NP proteins (A) or for
total protein tyrosine phosphorylation with a phospho-tyrosine antibody (P-Y-1000) (C). D and E, MDCK cells were infected with H5N1 virus
(D) or left uninfected (E). The cells were incubated with the indicated concentrations of A77 1726 for 16 hours. The cell lysates were prepared
and analyzed for JAK1, JAK2, and STAT3 tyrosine phosphorylation and for the expression of PB2 and NP proteins. F and G, CEF cells infected
with H5N1 virus were incubated for 16 hours in the absence or presence of the indicated concentration of Ruxolitinib (Rux). The cell lysates were prepared and analyzed for JAK2 and STAT3 tyrosine phosphorylation and for the expression of M1 and NP proteins (F) or for total protein tyrosine phosphorylation with a phospho-tyrosine antibody (P-Y-1000) (G). H, CEF cells were infected with H5N1 virus, and then, incubated for 16 h in the absence or presence of the indicated concentration of Rux. Conditioned media were collected at the indicated time and analyzed for the TCID50 values. The results represent the mean ± SD of four independent experiments. *P < .05; **P < .01, compared to the untreated controls. I, Rux does not inhibit CEF proliferation. CEF cells seeded in 96-well plates were incubated in the absence or presence of the indicated concentrations of Rux for 16 hours. Cell proliferation was analyzed as described in the Materials and Methods. Data are the mean ± SD of a representative experiment in triplicate. The experiments were repeated twice with similar results. *P < .05; **P < .01, compared to the untreated controls
FIGURE 5 JAK2 overexpression attenuates the antiviral activity of A77 1726. A and B, JAK2 overexpression enhances IAV replication.
MDCK cells were transiently transfected with the empty vector or the vector encoding JAK2. After incubation for 36 hours, the cells were left
uninfected or infected with the indicated MOI of H5N1 virus, and then, incubated for 12 hours. Cell lysates were prepared and analyzed for
JAK2 phosphorylation and for the levels of viral NP and M1 proteins (A). The virus titers in the conditioned media were collected and analyzed for the TCID50 values (B). Relative virus titers were calculated as the percent of controls. C and D, MDCK cells were transiently transfected
with the empty vector or the vector encoding JAK2. After incubation for 36 hours, the cells were infected with H5N1 virus (0.1 MOI), and then, incubated for 12 hours in the absence or presence of the indicated concentrations of A77 1726. Cell lysates were prepared and analyzed for JAK2 phosphorylation and for the levels of viral NP and M1 proteins (C). The virus titers in the conditioned media were collected and analyzed for the TCID50 values (D). The results represent the mean ± SD of three independent experiments. *P <.05; **P < .01 autophosphorylation and the phosphorylation of its substrate, STAT3. JAK1 and JAK2 have been identified as two crucial cellular factors that play an important role in IAV replica- tion.7,8 JAKs could be potentially targeted for antiviral ther- apy. In support of this notion, our present study showed that A77 1726 and Rux, a JAK-specific inhibitor, were able to inhibit IAV replication. JAK2 overexpression enhanced IAV replication but compromised the antiviral effect of A77 1726. These data collectively suggest that inhibition of JAK activ- ity plays an important role in A77 1726-mediated antiviral effects.
In addition to PTKs, A77 1726 also inhibits the activity of S6K1 and DHO-DHase, both of which have been im- plicated in supporting virus replication.10-12 Whether A77 1726 inhibits virus replication by inhibiting DHO-DHase activity remains controversial. A77 1726 inhibits the rep- lication of CMV,22-24 herpes simplex type 1 virus,37 BK,26 and Epstein-Barr (EB) virus27 largely independent of its inhibitory effect on pyrimidine nucleotide synthesis. In the present study, we found that uridine at 200 µM (Figure 1A) or higher concentrations (Figure S1) did not block the in- hibitory effect of A77 1726 on IAV replication in CEF, MDCK, and A549 cells. BQR is a potent inhibitor of DHO- DHase with an IC50 value approximately 10-fold lower than A77 1726.38,39 BQR is often used at nanomolar concen- trations,40 with the highest concentration of 5 µM.41,42 We found that BQR up to 20 µM did not inhibit the replication of H5N1 virus in three cell types. These observations col- lectively suggest that inhibition of pyrimidine nucleotide synthesis by A77 1726 does not contribute to its antiviral activity.
In contrast to these observations, several groups re- ported that A77 1726-mediated antiviral activity against BK virus,43 hepatitis E virus,44 and rotavirus45 is almost completely blocked by uridine. These seemingly contra- dicting results could be due to the differences in experimen- tal settings and due to the differences in virus replication and cell proliferation rates. Theoretically, inhibition of pyrimidine nucleotide synthesis and depletion of intracel- lular nucleotide pools by DHO-DHase inhibitors should lead to the inhibition of virus replication. However, intra- cellular pyrimidine nucleotide depletion relies on several factors, including the concentrations of fetal bovine serum in the media (serum at high concentrations stimulates cell proliferate), the length of an experiment, the confluence of cell monolayers used to infect virus, primary cells or cell lines, the rate of cell proliferation and virus replication, the efficiency of DHO-DHase inhibitors to inhibit pyrim- idine nucleotide synthesis, and the levels of DHO-DHase expression in host cells. Our unpublished results showed that uridine weakly but significantly reversed the antiviral activity of A77 1726 on H5N1 virus replication in CEF after prolonged incubation (36 hpi). In addition, BQR significantly inhibited H5N1 virus replication in MDCK
FIGURE 6 Leflunomide controls H5N1 virus infection in vivo. Female C57BL/6 (6-8-weeks-old) were pretreated with leflunomide 12 hours prior to infection. Mice were intranasally infected with H5N1 virus (1×105 pfu/mouse) and treated with water or leflunomide (20 mg/kg) (five
mice/group) daily for 3 or 5 days. Lung tissues were collected and analyzed for the levels of viral proteins (A and B) by Western blot and virus
titers (C and D) by measuring the TCID50 values. PB2, NP, and M1 protein levels were determined by analyzing the band density of these proteins, and then, normalized by β-actin protein band density with an NIH Image-J software. Data represent the mean ± SD of the lung tissues from fiveanimals. E, Hematoxylin and Eosin staining of the lung tissues. The anti-inflammatory effect of leflunomide was determined by counting infiltrates in five randomly selected fields. A bar graph representing the scores of infiltrates from five mice per group is shown. F and G, Mice (10 per group) were similarly treated with water or leflunomide (35 mg/kg/day) and infected with H5N1 virus. Mice were weighed and monitored for survival for 2 weeks. Body weights (F) and percent survival (G) were plotted. **P < .01, compared to the untreated controls cells when the sub-confluent monolayers were infected with H5N1 virus (data not shown). We postulate that under these conditions, the fast proliferating MDCK cells com- pete with IAV replication and quickly deplete of the py- rimidine nucleotide pools, leading to the inhibition of virus replication.
The notion that A77 1726 inhibits IAV replication largely independent of its inhibitory effect on pyrimidine nucleotide synthesis is strengthened by the observations that DHO- DHase inhibitors except leflunomide do not have any in vivo antiviral activity.12,46,47 Serum uridine derived from the sal- vage pathway through the digestion and degradation of di- etary nucleic acids can be used to replenish the intracellular pyrimidine nucleotide pools, thus nullifying the anticipated antiviral effect of DHO-DHase inhibitors. In addition, the re- spiratory and intestinal epithelial cells do not proliferate as fast as tumor cell lines used in the in vitro study, further ex- plaining why the DHO-DHase-specific inhibitors are unable to suppress virus replication in vivo. Leflunomide has been tested in clinical trials for the
control of CMV and BK virus infection in transplant pa- tients. Leflunomide is particularly effective as an anti-CMV prophylaxis treatment and for clearing low-grade viremia.48 Moreover, leflunomide in combination with a short course of Ganciclovir or intravitreal foscarn proves very effective in some patients.48 Leflunomide treatment promotes blood virus clearance in 76% of 55 kidney transplant patients in- fected with BK virus49 and leads to the eradication of acute hepatitis E virus in a rheumatoid arthritis patient.50 Although A77 1726 inhibits the activity of multiple intracellular en- zymes that are potentially involved in IAV replication, our study for the first time demonstrates that A77 1726 is an effective inhibitor of JAK1 and JAK2 and suggests that A77 1726 exerts its antiviral activity by inhibiting JAK activity.
Notably, A77 1726 exhibited strong antiviral activity in vitro but only moderate therapeutic efficacy in two mouse models(Figures 6 and S3). Consistent with this observation,Aeffner et al51 reported that post-infection lefluno- mide administration by aerosol nebulization modestly but significantly prolongs the survival of A/WSN/33 H1N1 virus-infected mice and decreases virus loads in the lung. A77 1726 also improves cardiopulmonary function in this model.51 Although a positive control with a clinically approved anti-IAV drug was not included in the studies of our own and theirs,51 leflunomide appears to be much less effec- tive at controlling IAV infections than currently approved an- ti-influenza drugs such as oseltamivir, which can provide up to 100% protection on the survival of IAV-infected mice.52-55 It is well received that the antimicrobial immunity in patients treated with an immunosuppressant is often compromised. Although leflunomide exhibits a broad spectrum of antivi- ral activity, it is an immunosuppressive drug. Patients treated with leflunomide should not be considered less susceptible to IAV infection before it is proven otherwise. In support of this cautious note, Dirven et al56 reported that RA patients treated with leflunomidetend to have a higher flu incidence rate than those who did not receive any treatment.
There has been great interest in searching for JAK inhibitors as novel anti-RA drugs. Validation of A77 1726 as an effective JAK inhibitor further supports the notion that le- flunomide functions as a PTK inhibitor to exert its immuno- modulatoryeffect. In addition, demonstration of the ability of A77 1726 to effectively inhibit JAK1 and JAK2 activity also suggests that leflunomide could be potentially repurposed for treating leukemia.
