Protein S-nitrosylation mediated by cellular nitric oxide (NO) plays a primary role in executing biological functions in cGMPindependent NO signaling. Although S-nitrosylation appears similar to Cys oxidation induced by reactive oxygen species, the molecular mechanism and biological consequence remain unclear. We investigated the structural process of S-nitrosylation of protein-tyrosine phosphatase 1B (PTP1B). We treated PTP1B with various NO donors, including S-nitrosothiol reagents and compound-releasing NO radicals, to produce sitespecific Cys S-nitrosylation identified using advanced mass spectrometry (MS) techniques. Quantitative MS showed that the active site Cys-215 was the primary residue susceptible to S-nitrosylation. The crystal structure of NO donor-reacted PTP1B at 2.6 Å resolution revealed that the S-NO state at Cys-215 had no discernible irreversibly oxidized forms, whereas other Cys residues remained in their free thiol states. We further demonstrated that S-nitrosylation of the Cys-215 residue protected PTP1B from subsequent H 2 O 2 -induced irreversible oxidation. Increasing the level of cellular NO by pretreating cells with an NO donor or by activating ectopically expressed NO synthase inhibited reactive oxygen species-induced irreversible oxidation of endogenous PTP1B. These findings suggest that S-nitrosylation might prevent PTPs from permanent inactivation caused by oxidative stress.

Various inositide phosphatases participate in the regulation of inositol polyphosphate signaling molecules. Plant phytases are phosphatases that hydrolyze phytate to less-phosphorylated myo-inositol derivatives and phosphate. The phytase from Selenomonas ruminantium shares no sequence homology with other microbial phytases. Its crystal structure revealed a phytase fold of the dual-specificity phosphatase type. The active site is located near a conserved cysteine-containing (Cys241) P loop. We also solved two other crystal forms in which an inhibitor, myo-inositol hexasulfate, is cocrystallized with the enzyme. In the "standby" and the "inhibited" crystal forms, the inhibitor is bound, respectively, in a pocket slightly away from Cys241 and at the substrate binding site where the phosphate group to be hydrolyzed is held close to the -SH group of Cys241. Our structural and mutagenesis studies allow us to visualize the way in which the P loop-containing phytase attracts and hydrolyzes the substrate (phytate) sequentially.

Omalizumab, an anti-IgE antibody, used to treat severe allergic asthma and chronic idiopathic urticaria, binds to IgE in blood or membrane-bound on B lymphocytes but not to IgE bound to its high (FcεRI) or low (CD23) affinity receptor. Mutagenesis studies indicate overlapping FcεRI and omalizumab-binding sites in the Cε3 domain, but crystallographic studies show FcεRI and CD23-binding sites that are far apart, so how can omalizumab block IgE from binding both receptors? We report a 2.42-Å omalizumab-Fab structure, a docked IgE-Fc/omalizumab-Fab structure consistent with available experimental data, and the free energy contributions of IgE residues to binding omalizumab, CD23, and FcεRI. These results provide a structural and physical basis as to why omalizumab cannot bind receptor-bound IgE and why omalizumab-bound IgE cannot bind to CD23/FcεRI. They reveal the key IgE residues and their roles in binding omalizumab, CD23, and FcεRI.

The P-loop-containing protein phos-phatases are important regulators in signal transduction. These enzymes have structural and functional similarity with a conserved sequence of Dx(25-41)HCxxGxxR(T/S) essential for catalysis. The singular protein tyrosine phosphatase (PTP) from archaeal Sulfolobus solfataricus is one of the smallest known PTPs with extreme thermostability. Here, we report the crystal structure of this phosphatase and its complexes with two tyrosyl phosphopeptides A-(p)Y-R and N-K-(p)Y-G-N. The structure suggests the minimal structural motif of the PTP family, having two variable sequences inserted between the beta2-beta3 and beta3-beta4 strands, respectively. The phosphate of both phosphopeptide substrates is bound to the P-loop through several hydrogen bonds. Comparison of several phosphatase-substrate complexes revealed that Gln135 on the Q-loop has different modes of recognition toward phosphopeptides. The substrate specificity of SsoPTP is primarily localized at the phosphotyrosine, suggesting that this phosphatase may be a prototypical PTP.

The protein Sac7d belongs to a class of small chromosomal proteins from the hyperthermophilic archaeon Sulfolobus acidocaldarius. Two new crystal forms of Sac7d in complex with the DNA decamers CCTATATAGG and CCTACGTAGG were obtained and their structures were determined by molecular replacement. The refined models yielded R/Rfree values of 0.221/0.257 and 0.248/0.290 at 1.9 and 2.2 A resolution, respectively. The protein structures are similar to the previously determined structure of Sac7d-GCGATCGC (PDB code 1azp), but the DNA molecules are more bent overall, by 14-20 degrees. The relative positions of the Sac7d protein and the bound DNA also differ by rotations of 6-10 degrees and translations of 1.0-2.4 A. In addition to the water molecules in the central cavity, three additional conserved water molecules are found that mediate the protein-DNA interactions. The decamer DNA fragments form virtual double helices in the crystal, with a unit length of eight base pairs. The molecular packing of the new crystal forms differs from that of 1azp. The terminal nucleotides are opened up and form triple base pairs with other DNA molecules. Through lattice contacts, the Sac7d molecule also makes additional interactions with DNA, whereas only limited protein-protein interactions are seen.

Active learning is an important machine learning problem in reducing the human labeling effort. Current active learning strategies are designed from human knowledge, and are applied on each dataset in an immutable manner. In other words, experience about the usefulness of strategies cannot be updated and transferred to improve active learning on other datasets. This paper initiates a pioneering study on whether active learning experience can be transferred. We first propose a novel active learning model that linearly aggregates existing strategies. The linear weights can then be used to represent the active learning experience. We equip the model with the popular linear upperconfidence-bound (LinUCB) algorithm for contextual bandit to update the weights. Finally, we extend our model to transfer the experience across datasets with the technique of biased regularization. Empirical studies demonstrate that the learned experience not only is competitive with existing strategies on most single datasets, but also can be transferred across datasets to improve the performance on future learning tasks.

Cytokinins are important plant hormones, and their biosynthesis most begins with the transfer of isopentenyl group from dimethylallyl diphosphate (DMAPP) to the N6-amino group of adenine by either adenylate isopentenyltransferase (AIPT) or tRNA–IPT. Plant AIPTs use ATP/ADP as an isopentenyl acceptor and bacterial AIPTs prefer AMP, whereas tRNA–IPTs act on specific sites of tRNA. Here, we present the crystal structure of an AIPT–ATP complex from Humulus lupulus (HlAIPT), which is similar to the previous structures of Agrobacterium AIPT and yeast tRNA–IPT. The enzyme is structurally homologous to the NTP-binding kinase family of proteins but forms a solvent-accessible channel that binds to the donor substrate DMAPP, which is directed toward the acceptor substrate ATP/ADP. When measured with isothermal titration calorimetry, some nucleotides displayed different binding affinities to HlAIPT with an order of ATP > dATP ∼ ADP > GTP > CTP > UTP. Two basic residues Lys275 and Lys220 in HlAIPT interact with the β and γ-phosphate of ATP. By contrast, the interactions are absent in Agrobacterium AIPT because they are replaced by the acidic residues Asp221 and Asp171. Despite its structural similarity to the yeast tRNA–IPT, HlAIPT has evolved with a different binding strategy for adenylate.

IgE mediates hypersensitivity reactions responsible for most allergic diseases, which affect 20-40% of the population in developed countries. A 52-residue domain of membrane-bound IgE (mIgE) called CemX is currently a target for developing therapeutic antibodies; however, its structure is unknown. Here we show that two antibodies with therapeutic potential in IgE-mediated allergic diseases, which can cause cytolytic effects on mIgE-expressing B lymphocytes and downregulate IgE production, target different conformations of an intrinsically disordered region (IDR) in the extracellular CemX domain. We provide an important example of antibodies targeting an extracellular IDR of a receptor on the surface of intended target cells. We also provide fundamental structural characteristics unique to human mIgE, which may stimulate further studies to investigate whether other monoclonal antibodies (mAbs) targeting intrinsically disordered peptide segments or vaccine-like products targeting IDRs of a membrane protein can be developed.