Human infections with avian influenza H7N9 or H10N8 viruses have been reported in China, raising concerns that they might cause human epidemics and pandemics. However, how these viruses adapt to mammalian hosts is unclear. Here we show that besides the commonly recognized viral polymerase subunit PB2 residue 627 K, other residues including 87E, 292 V, 340 K, 588 V, 648 V, and 676 M in PB2 also play critical roles in mammalian adaptation of the H10N8 virus. The avian-origin H10N8, H7N9, and H9N2 viruses harboring PB2-588 V exhibited higher polymerase activity, more efficient replication in mammalian and avian cells, and higher virulence in mice when compared to viruses with PB2-588 A. Analyses of available PB2 sequences showed that the proportion of avian H9N2 or human H7N9 influenza isolates bearing PB2-588 V has increased significantly since 2013. Taken together, our results suggest that the substitution PB2-A588V may be a new strategy for an avian influenza virus to adapt mammalian hosts.

Session-based target behavior prediction aims to predict the next item to be interacted with specific behavior types (e.g., clicking). Although existing methods for session-based behavior prediction leverage powerful representation learning approaches to encode items' sequential relevance in a low-dimensional space, they suffer from several limitations. Firstly, they focus on only utilizing the same type of user behavior for prediction, but ignore the potential of taking other behavior data as auxiliary information. This is particularly crucial when the target behavior is sparse but important (e.g., buying or sharing an item). Secondly, item-to-item relations are modeled separately and locally in one behavior sequence, and they lack a principled way to globally encode these relations more effectively. To overcome these limitations, we propose a novel Multirelational Graph Neural Network model for Session-based target behavior Prediction, namely MGNN-SPred for short. Specifically, we build a Multi-Relational Item Graph (MRIG) based on all behavior sequences from all sessions, involving target and auxiliary behavior types. Based on MRIG, MGNN-SPred learns global itemto-item relations and further obtains user preferences w.r.t. current target and auxiliary behavior sequences, respectively. In the end, MGNN-SPred leverages a gating mechanism to adaptively fuse user representations for predicting next item interacted with target behavior. The extensive experiments on two real-world datasets demonstrate the superiority of MGNN-SPred by comparing with state-of-the-art session-based prediction methods, validating the benefits of leveraging auxiliary behavior and learning item-to-item relations over MRIG.

The H5 subtype virus of Highly Pathogenic Avian Influenza Virus has caused huge economic losses to the poultry industry and is a threat to human health. Until 2010, H5N1 subtype virus was the major genotype in China. Since 2011, reassortant H5N2, H5N6, and H5N8 viruses were identified in domestic poultry in China. The clade 2.3.4.4 H5N6 and H5N8 AIV has now spread to most of China. Clade 2.3.4.4 H5N6 virus has caused 17 human deaths. However, the prevalence, pathogenicity, and transmissibility of the distinct NA reassortment with H5 subtypes viruses (H5Nx) is unknown. We constructed five clade 2.3.4.4 reassortant H5Nx viruses that shared the same HA and six internal gene segments. The NA gene segment was replaced with N1, N2, N6, ΔN6 (with an 11 amino acid deletion at the 58th to 68th of NA stalk region), and N8 strains, respectively. The reassortant viruses with distinct NAs of clade 2.3.4.4 H5 subtype had different degrees of fitness. All reassortant H5Nx viruses formed plaques on MDCK cell monolayers, but the ΔH5N6 grew more efficiently in mammalian and avian cells. The reassortant H5Nx viruses were more virulent in mice as compared to the H5N2 virus. The H5N6 and H5N8 reassortant viruses exhibited enhanced pathogenicity and transmissibility in chickens as compared to the H5N1 reassortant virus. We suggest that comprehensive surveillance work should be undertaken to monitor the H5Nx viruses.

Gap junctions (GJs), collections of multiple intercellular channels between neighboring cells, are specialized channels facilitating intercellular electrical and chemical communication. GJs are important for synchronizing coupling and coordinated contraction in the heart, and are crucial regulators of heart gene transcription, cardiac development, and protection of ischemic cardiomyocytes through second messenger communication. Identification of proteins that interact with Connexin43 (Cx43), the predominant protein in cardiac GJs, may contribute to the understanding of GJ functional regulation. Using a yeast two-hybrid system, we identified Caveolin-3 (Cav3) as a new Cx43-interacting protein. This interaction was confirmed by co-immunoprecipitation and co-localization experiments. CX43 interacts with Cav3, suggesting that Cav3 may participate in the functional regulation of GJs.