*3.6. Brevilin A Induces Influenza Viral RNP Aggregation in the Nucleus*

In contrast with most RNA viruses, the replication and transcription of IAV were carried out in the nucleus of the infected cells. After uncoating, the vRNP complex, which consists of the viral PB1-PB2-PA (3P) heterotrimeric RNA polymerase and NP protein, is imported into the nucleus for virus RNA replication and transcription and then exported to the cytoplasm for packaging into newly formed virions at the cellular plasma membranes [26]. The matrix protein M1, the viral nuclear export protein (NS2/NEP), and the M2 ion channel protein are essential proteins involved in viral trafficking, releasing into the cytoplasm, and budding [27]. To investigate the effects of brevilin A on the nucleocytoplasmic trafficking of vRNPs, MDCK cells were infected with PR8 virus for 1 h and then treated with DMSO, brevilin A, or LMB. The viral NP protein was detected by indirect immunofluorescence microscopy at 4, 8, and 11 hpi to determine the vRNP localization. As shown in Figure 6, in vehicle treated cells, the NP protein was detected exclusively in the nucleus at 4 hpi and shifted toward the cytoplasm at 8 hpi. By 11 hpi, vRNPs were mainly distributed in the cytoplasm. As a control, we used LMB, a potent and specific nuclear export inhibitor, which has been demonstrated to be able to cause the nuclear accumulation of newly generated vRNPs [28]. Consistent with the literature report, LMB prevented the export of vRNPs, even at the late stage of infection (11 hpi), whereas their nuclear import was not affected. Similar to LMB, upon treatment with brevilin A, the translocation of vRNPs to the nucleus was delayed. The aggregation of vRNPs was also observed at 11 hpi (Figure 6). These results suggest that brevilin A induces vRNPs aggregation in the nucleus.

**Figure 6.** The effects of brevilin A on the viral ribonucleoproteins (vRNP) localization. MDCK cells were infected with PR8 at a MOI of 3 and then treated with brevilin A, LMB, or vehicle as indicated. Samples were fixed at 4, 8, or 11 hpi, and then stained with anti-NP body (green) and DAPI (blue). Immunofluorescence was observed with confocal microscopy. The scale bar in the images is 20 μm.

### *3.7. Brevilin A Protects Mice from IAV Pathogenesis*

Since brevilin A showed antiviral activity against influenza A virus in vitro, we next examined whether brevilin A could also protect mice against influenza virus infection. Mice were infected intratracheally with PR8 at a dose of three 50% lethal doses (LD50). Brevilin A was given once every other day intraperitoneally immediately after infection at two concentrations (10 mg/kg or 25 mg/kg). Oseltamivir was used as the positive control that it is commonly used for treating influenza virus

infection in the clinic. The experiment was conducted following the scheme illustrated in Figure 7A. Morbidity and mortality were monitored daily by measuring the body weight and survival rate (Figure 7B,C). Animals falling below the threshold of 75% of their initial body weight were humanely euthanized. Vehicle-treated mice showed severe morbidity after infection with influenza virus, and 100% mortality at 10 days post-infection, while the uninfected group (normal) and oseltamivir-treated group (20 mg/kg/day) showed 100% survival during the entire experiment. Mice treated with brevilin A at 10 mg/kg did not show significant differences in terms of body weight loss or survival rate (Figure 7B,C). However, treatment of brevilin A at 25 mg/kg sustained the body weights of mice in comparison to vehicle-treated group (Figure 7B). Also, brevilin A-treated mice (25 mg/kg) showed a delayed time-to-death with 50% survival up to 14 days post-infection (Figure 7C). Thus, these results show that brevilin A protects mice from IAV pathogenesis.

**Figure 7.** Brevilin A decreases the lethality observed in IAV-infected mice. (**A**) BALB/c mice were infected intranasally with three 50% lethal doses (LD50) of PR8 virus. Brevilin A (25 mg/kg or 10 mg/kg) was intraperitoneally (i.p.) administered to mice 1 h after virus infection, and then once every other day for 6 days beginning on the day of infection (*n* = 8). Oseltamivir phosphate (20 mg/kg) used as a positive control was administered by oral gavage every day (*n* = 7). Vehicle (10% Lipovenos containing 0.2% DMSO, 10% PEG300 and 2.5% glycerol, *n* = 8) was used as a negative control. (**B**) The body weight of mice from each group was monitored daily from day 0 to day 14. The data represent means ± SD. (**C**) The survival rates of the mice were calculated. Animals falling below the threshold of 75% of their initial body weight were humanely euthanized. The *p*-value is shown (Log-rank (Mantel–Cox) test).

### **4. Discussion**

The present study shows that brevilin A at a noncytotoxic concentration has a broad-spectrum antiviral activity against IAV, including H1N1, H3N2, and H9N2. Mode of mechanism studies demonstrate that brevilin A exhibits its antiviral activity by regulating the replication and translation stages of IAV life cycle. Brevilin A strongly decreased viral RNA level, reduced the expression of viral proteins expressed from the smaller segments (M and NS), and impaired the nuclear export of vRNPs. Furthermore, we showed that brevilin A reduced influenza-associated morbidity and mortality in vivo.

In the current study, we determined that the anti-IAV activity of brevilin A is not viral subtype specific as brevilin A displayed a broad-spectrum antiviral activity against many IAV types/subtypes. These effects were assessed by plaque reduction assay and generation of virus growth curves (Figure 2). Using a time-of-addition assay, we deduced that brevilin A acts at the replication and translation stages of infection (Figure 3), which could explain why the virus titers at 4 hpi were not reduced by treatment with brevilin A (Figure 2B). Further examination reveals that brevilin A preferentially

regulates the synthesis of vRNA but not the complementary positive strand cRNA or the mRNA (Figure 4). Moreover, by analyzing the expression of viral protein, we determined that not all of the viral proteins are equally affected by brevilin A. Expression of M2, NS1, and NS2 proteins are more severely inhibited in comparison to other viral proteins (Figure 5A). It has been reported that some IAV genes (segments 1, 2, 3, 5, and unspliced 8) are preferentially expressed early and the others (segments 4, 6, unspliced 7, and two spliced transcripts) are expressed late during infection [23]. However, this could not explain our observation that at early time points post-infection, the proteins translated from the M and NS mRNAs were down-regulated by treatment with brevilin A, while the expression of HA and NA (supposed to be preferentially expressed early) were not affected. Considering that the M and NS segments are coincidentally the ones that produce spliced products, the mRNA levels of M1, M2, NS1, and NS2 were analyzed. Our results demonstrate that the production of M2 mRNA was strongly reduced following the treatment with brevilin A (Figure 5B–E), indicating that the alternative splicing of the M1 mRNA was affected. Splicing is a necessary step for influenza replication, while NS2 is required for nucleocytoplasmic transport of vRNPs [29], and M2 is an important factor in viral pathogenicity [30,31]. During the nuclear replication stage, numerous host-splicing factors, such as the spliceosome complex and host splicing regulators, are necessary to process the M and NS segments [32]. How brevilin A affects the alternative splicing of influenza A viruses needs to be further investigated.

Brevilin A's antiviral effects in vivo were also evaluated in a mouse model upon influenza virus infection. Often, a compound possessing potent inhibitory activity in vitro fails when tested in vivo. However, we found that treatment of IAV-infected mice with brevilin A markedly improved their survival, compared to vehicle control mice (Figure 7). Under our experimental settings, brevilin A (25 mg/kg) was delivered every other day for four times rather than every day for 6 days, since both delivery frequency showed similar protective effects in IAV-infected mice. Further exploration of the in vivo potential of brevilin A is warranted to assess its in vivo toxicity, and define the best conditions of treatment, including analysis of dosage and route(s) of inoculation. Besides, brevilin A's bioavailability and pharmacokinetics would also be considered in future research to get more detailed information about absorption, metabolism, and disposition.

In summary, the sesquiterpene lactone brevilin A is a promising candidate lead compound for development of antiviral agents that broadly inhibit IAV replication by impairing the vRNA synthesis and the viral protein translation. Further investigation is warranted of this and other similar inhibitors as potential therapeutic agents against influenza.

**Author Contributions:** Conceptualization, X.Z. and C.X.; methodology, X.Z. and Y.X.; investigation, X.Z., Y.X., L.Y., J.H., Y.L. and C.X.; data, X.Z., Y.X., L.Y. and C.X.; writing—original draft preparation, X.Z. and C.X.; writing—review and editing, X.Z. and C.X.; supervision, X.Z. and C.X.; funding acquisition, X.Z. and J.H.

**Funding:** This research was funded by National Natural Science Foundation of China (81703371, 81603165, 81672006), the Guangdong Innovative and Entrepreneurial Research Team Program (2014ZT05S136) and China Postdoctoral Science Foundation (2017M622924).

**Acknowledgments:** The authors thank Wenbao Qi from the Key Laboratory of Veterinary Vaccine Innovation of the Ministry of Agriculture, P. R. China for providing the influenza virus A/PR/8/34 (H1N1) and MDCK cells. We thank Fenyong Liu (University of California, Berkeley), Rong Hai (University of California, Riverside), Hua Zhu (Rutgers New Jersey Medical School), Jun Chen (Jinan University), and Ting Wang (Jinan University) for critical comments, reagents, and technical assistance.

**Conflicts of Interest:** The authors declare no conflicts of interests.

#### **References**


© 2019 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 (http://creativecommons.org/licenses/by/4.0/).

*Article*
