We present a neural network-based joint approach for emotion classification and emotion cause detection, which attempts to capture mutual benefits across the two sub-tasks of emotion analysis. Considering that emotion classification and emotion cause detection need different kinds of features (affective and event-based separately), we propose a joint encoder which uses a unified framework to extract features for both sub-tasks and a joint model trainer which simultaneously learns two models for the two sub-tasks separately. Our experiments on Chinese microblogs show that the joint approach is very promising.

A notably challenging problem in emotion analysis is recognizing the cause of an emotion. Although there have been a few studies on emotion cause detection, most of them work on news reports or a few of them focus on microblogs using a single-user structure (i.e., all texts in a microblog are written by the same user). In this article, we focus on emotion cause detection for Chinese microblogs using a multiple-user structure (i.e., texts in a microblog are successively written by several users). First, based on the fact that the causes of an emotion of a focused user may be provided by other users in a microblog with the multiple-user structure, we design an emotion cause annotation scheme which can deal with such a complicated case, and then provide an emotion cause corpus using the annotation scheme. Second, based on the analysis of the emotion cause corpus, we formalize two emotion cause detection tasks for microblogs (current-subtweet-based emotion cause detection and original-subtweet-based emotion cause detection). Furthermore, in order to examine the difficulty of the two emotion cause detection tasks and the contributions of texts written by different users in a microblog with the multiple-user structure, we choose two popular classification methods (SVM and LSTM) to do emotion cause detection. Our experiments show that the current-subtweet-based emotion cause detection is much more difficult than the original-subtweet-based emotion cause detection, and texts written by different users are very helpful for both emotion cause detection tasks. This study presents a pilot study of emotion cause detection which deals with Chinese microblogs using a complicated structure.

Cyclization of the polypeptide backbone has proven to be a powerful strategy for enhancing protein stability for fundamental research and pharmaceutical application. The use of such an approach is restricted by how well a targeted polypeptide can be efficiently ligated. Recently, an Asx-specific peptide ligase identified from a tropical cyclotide-producing plant and named butelase 1 exhibited excellent cyclization kinetics that cannot be matched by other known ligases, including intein, PATG, PCY1, and sortase A. In this work, we aimed to examine whether butelase 1 facilitated protein conformational stability for structural investigation. First, we successfully expressed recombinant butelase 1 (rBTase) in the yeast Pichia pastoris. Next, rBTase was shown to be highly efficient in the cyclization of the p53-binding domain (N-terminal domain) of murine double minute X (N-MdmX), an important target for designing anticancer drugs. The cyclized N-MdmX (cMdmX) exhibited increased conformational stability and improved interaction with the ligand compared with those of noncyclized N-MdmX. Importantly, the thermal melting process was completely reversible, contrary to noncyclized N-MdmX, and the melting temperature (T m ) of cMdmX was increased to 47 from 43 °C. This stable conformation of cMdmX was further confirmed by 15 N− 1 H heteronuclear single-quantum coherence nuclear magnetic resonance (NMR) spectroscopy. The complex of cMdmX and the ligand was tested for protein crystallization, and several promising findings were revealed. Therefore, our work not only provides a recombinant version of butelase 1 but also suggests a conventional approach for preparing stable protein samples for both protein crystallization and NMR structural investigation.

The type 3 secretion system (T3SS) found as cell-surface appendages of many pathogenic Gram-negative bacteria, although nonessential for bacterial survival, is an important therapeutic target for drug discovery and development aimed at inhibiting bacterial virulence without inducing antibiotic resistance. We designed a fluorescence-polarization-based assay for high-throughput screening as a mechanistically well-defined general strategy for antibiotic discovery targeting the T3SS and made a serendipitous discovery of a subset of tanshinones—natural herbal compounds in traditional Chinese medicine widely used for the treatment of cardiovascular and cerebrovascular diseases—as effective inhibitors of the biogenesis of the T3SS needle of multi-drug-resistant Pseudomonas aeruginosa . By inhibiting the T3SS needle assembly and, thus, cytotoxicity and pathogenicity, selected tanshinones reduced the secretion of bacterial virulence factors toxic to macrophages in vitro , and rescued experimental animals challenged with lethal doses of Pseudomonas aeruginosa in a murine model of acute pneumonia. As first-in-class inhibitors with a demonstrable safety profile in humans, tanshinones may be used directly to alleviate Pseudomonas-aeruginosa -associated pulmonary infections without inducing antibiotic resistance. Since the T3SS is highly conserved among Gram-negative bacteria, this antivirulence strategy may be applicable to the discovery and development of novel classes of antibiotics refractory to existing resistance mechanisms for the treatment of many bacterial infections.

The oncoprotein MdmX (mouse double minute X) is highly homologous to Mdm2 (mouse double minute 2) in terms of their amino acid sequences and three-dimensional conformations, but Mdm2 inhibitors exhibit very weak affinity for MdmX, providing an excellent model for exploring how protein conformation distinguishes and alters inhibitor binding. The intrinsic conformation flexibility of proteins plays pivotal roles in determining and predicting the binding properties and the design of inhibitors. Although the molecular dynamics simulation approach enables us to understand protein-ligand interactions, the mechanism underlying how a flexible binding pocket adapts an inhibitor has been less explored experimentally. In this work, we have investigated how the intrinsic flexible regions of the N-terminal domain of MdmX (N-MdmX) affect the affinity of the Mdm2 inhibitor nutlin-3a using protein engineering. Guided by heteronuclear nuclear Overhauser effect measurements, we identified the flexible regions that affect inhibitor binding affinity around the ligand-binding pocket on N-MdmX. A disulfide engineering mutant, N-MdmX, which incorporated two staples to rigidify the ligand-binding pocket, allowed an affinity for nutlin-3a higher than that of wild-type N-MdmX (K ∼ 0.48 vs K ∼ 20.3 μM). Therefore, this mutant provides not only an effective protein model for screening and designing of MdmX inhibitors but also a valuable clue for enhancing the intermolecular interactions of the pharmacophores of a ligand with pronounced flexible regions. In addition, our results revealed an allosteric ligand-binding mechanism of N-MdmX in which the ligand initially interacts with a compact core, followed by augmenting intermolecular interactions with intrinsic flexible regions. This strategy should also be applicable to many other protein targets to accelerate drug discovery.