**4. Discussion**

In the present study, OA-10, a newly synthesized oleanane-type triterpene, exhibits significant antiviral activity against highly pathogenic H5N1 IAV replication with an EC50 of 14.0 μM in A549 cell cultures. The in vitro cytotoxicity of OA-10 is quite low, with a CC50 of more than 640 μM and a selective index of more than 45. OA-10 also exhibits similar inhibitory effect on other IAV subtypes including PR8, H3N2 and H9N2, with EC50 values ranging from 6.7 to 19.6 μM. The time course inhibition study indicates that OA-10 exerts its antiviral effect during early stages of viral infection post binding to cell surface receptors. SPR analyses demonstrate that OA-10 interacts with HA protein strongly. The inhibition of OA-10 on H5N1 IAV induced hemolysis of chicken RBCs at low pH confirms interaction of OA-10 with HA subunit HA2. Furthermore, computer-aided molecular docking analysis suggested that OA-10 might bind to the interface of HA1 and HA2 in the HA stem region, which was known to undergo significant rearrangement during membrane fusion [19].

One IAV replication cycle is orderly composed of virus binding, internalization, RNA replication and viral protein synthesis, assembly, budding and release from infected cells [38]. Through the time course inhibition experiments, we found that OA-10 inhibited IAV replication in the co- and post-treatment modes (Figure 3), but was unable to block IAV binding to A549 cells (Figure S3). To identify the exact stage(s) of IAV replication cycle affected by OA-10, we investigated the time course inhibition within one IAV lifecycle and found that OA-10 exerted its effect during the early stages of IAV infection (Figure 4). During this stage, IAV attachment to target cells is mediated by HA1 via sialic acid-receptor binding, and subsequent virus-endosome membrane fusion is mediated by rearrangement of HA2 at low pH. Our results showed that OA-10 did not block IAV binding to cells at 4 ◦C (Figure S3). In addition, OA-10 did not inhibit IAV adsorption to chicken RBCs at concentrations

of 10 and 20 μM, in spite of the fact that non-classical hemagglutination inhibitions of OA-10 at 40 and 80 μM were observed. These results suggest that the HA1 receptor binding domain is likely not a main target. Given that OA-10 exhibited antiviral activity during early stage(s) of IAV infection cycle, we speculated that OA-10 might target the membrane fusion step mediated by the more conserved hemagglutinin transmembrane subunit HA2. This hypothesis was confirmed by the activity of blocking hemolysis of OA-10 in a low pH environment in a dose-dependent manner (Figure 5C), as hemolysis is mediated by HA2 rather than HA1 [39].

To investigate the binding intensity of OA-10 with HA, SPR analyses were conducted. SPR data showed that OA-10 interacted with HA strongly with KD of 2.98 <sup>×</sup> 10−<sup>12</sup> M (Figure 5B), which was consistent with its potent hemolysis inhibition at low pH. To study the possible binding site, docking simulation analyses were performed, by which a highly conserved hydrophobic cavity at the HA1-HA2 interface in the HA stem region was identified as the possible binding site of OA-10. The binding cavity is formed by residues of the two HA subunits, including Gly20, Val18, Trp21, Lys38 and Ile45 in HA2, and His38 in HA1, which was previously reported to be one of the critical regions responsible for the conformational changes in HA2 at low pH [19]. In addition, this region is recognized by an antibody with a broad-spectrum neutralizing ability to avian and human influenza A viruses [19]; thus, it could serve as a potential drug target for developing IAV entry inhibitors. Indeed, our data indicated that OA-10 might bind to this cavity through hydrogen bonds and hydrophobic interaction.

Influenza A viruses have been classified into 18 hemagglutinin subtypes (H1 to H18), which can be divided phylogenetically into two groups (1 and 2). H1, H5 and H9 belong to Group 1 HA of IAV, while H3 belongs to Group 2 HA of IAV [40]. It is known that the amino acid sequence identities of the HA2 portion between different HA subtypes are much higher than those of the full-length HA proteins [40], which makes HA2 an attractive target for developing broad-spectrum therapeutic antibodies and antiviral drugs. In fact, several broadly neutralizing antibodies (bnAbs) against IAV conserved HA stem have been developed, and their broad protection against IAV infection has been demonstrated in clinical trials. One such examples is the antibody CR6261 for most group 1 IAV subtypes [41,42]. Recently, guided by structural knowledge on the interactions of HA and anti-stem bnAb CR6261, a small molecule JNJ4796 that mimics the bnAb functionality with the ability to inhibit HA-mediated fusion was successfully developed. Importantly, this compound demonstrated potent antiviral activities against H1 and H5 strains in vitro and in vivo [19], but not group 2 IAVs such as H3 and H7 subtypes [19]. An OA derivative Y3 was reported to have significant antiviral activity in vitro against H3 and H1, suggesting broad-spectrum inhibitions of OA derivatives against both Group 1 and Group 2 IAVs.[13]. Further, HR2 in influenza HA2 was recently shown to be the target domain for Y3 [43]. In the present study, we show that OA-10 has promising antiviral activities against four IAV subtypes, including H1N1, H5N1, H9N2 (Group 1 IAVs) and H3N2 (Group 2 IAV), further demonstrating the potential of OA derivatives as broad-spectrum IAV entry inhibitors. To our knowledge, this is the first reported antiviral activity of OA derivatives against H5N1 IAV infection. Further in vivo studies will be carried out to clarify the efficacy of OA-10 as an IAV entry inhibitor in animals.

Another potential advantage of OA-10- s antiviral activity is its synergistic inhibition on viral replication when used together with ribavirin, a broad-spectrum virus RNA polymerase inhibitor (Figure 3). Such synergy was likely the result of the simultaneous disruption of HA-mediated viral entry by OA-10 (earlier step) and polymerase-mediated RNA replication by ribavirin (later step). It is expected that using an inhibitor like OA-10 together with ribavirin or NA inhibitors would not only enhance antiviral effects against IAV infection, but also prevent or significantly delay drug resistance.

In summary, we demonstrate that OA-10, a novel synthesized oleanane-type triterpenoid, inhibits four different IAV subtypes, including Group 1 H1N1 (PR8), H5N1 and H9N2, and Group 2 H3N2 infections with potent activity and negligible cytotoxicity in A549 cells. Mechanically, OA-10 blocks the conformational changes of the HA2 subunit at the low pH required for IAV to fuse with an endosomal membrane. These effects are attributed to a conserved hydrophobic cavity in the HA stem region as the likely binding site of OA-10. It could serve as a lead for optimization in order to design novel compounds with improved antiviral potency. OA-10 and its derivatives hold promise to be developed as broad-spectrum anti-influenza drugs.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1999-4915/12/2/225/s1: Figure S1: Chemical structures of oleanane acid (OA-0) and its 11 derivatives. OA-8, OA-9, OA-10 and OA-11 were newly synthesized derivatives. Figure S2: OA-10 inhibits IAV PR8 (H1N1), H9N2 and H3N2 replications in A549 cells. Figure S3: OA-10 does not block IAV binding to A549 cells. Figure S4: Amino acid sequence alignment of the full-length HA proteins between strains of A/Vietnam/1203/2004 (H5N1) and A/Duck/Guangdong/99(H5N1).

**Author Contributions:** J.C. and G.S. designed the study and wrote the paper. M.Y. and Y.L. (Yixian Liao) performed the experiments. L.W. and N.C. participated in the establishment of the methodology and data analysis. W.Q. and Y.L. (Yahong Liu) provided some critical reagents. W.C. thoroughly revised the manuscript and participated in study design. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Key Research and Development Program of China (grant number 2017YFD0501404) and the National Natural Science Foundation of China (grant numbers 31572565 and 31872521) to Jianxin Chen.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
