Search

Gap junction intercellular channels (GJIChs) facilitate direct communication between adjacent cells. Here, authors provide high-resolution information on dynamic structural changes in Cx43 GJICh that are necessary to understand the gating mechanism.

YgjG is a putrescine aminotransferase enzyme that transfers amino groups from compounds with terminal primary amines to compounds with an aldehyde group using pyridoxal-5'-phosphate (PLP) as a cofactor. Previous biochemical data show that the enzyme prefers primary diamines, such as putrescine, over ornithine as a substrate. To better understand the enzyme's substrate specificity, crystal structures of YgjG from Escherichia coli were determined at 2.3 and 2.1 Å resolutions for the free and putrescine-bound enzymes, respectively. Sequence and structural analyses revealed that YgjG forms a dimer that adopts a class III PLP-dependent aminotransferase fold. A structural comparison between YgjG and other class III aminotransferases revealed that their structures are similar. However, YgjG has an additional N-terminal helical structure that partially contributes to a dimeric interaction with the other subunit via a helix-helix interaction. Interestingly, the YgjG substrate-binding site entrance size and charge distribution are smaller and more hydrophobic than other class III aminotransferases, which suggest that YgjG has a unique substrate binding site that could accommodate primary aliphatic diamine substrates, including putrescine. The YgjG crystal structures provide structural clues to putrescine aminotransferase substrate specificity and binding.

Coproduction of multiple proteins at high levels in a single human cell line would be extremely useful for basic research and medical applications. Here, a novel strategy for the stable expression of multiple proteins by integrating the genes into defined transcriptional hotspots in the human genome is presented. As a proof-of-concept, it is shown that EYFP is expressed at similar levels from hotspots and that the EYFP expression increases proportionally with the copy number. It is confirmed that three different fluorescent proteins, encoded by genes integrated at different loci, can be coexpressed at high levels. Further, a stable cell line is generated, producing antigens from different human coronaviruses: MERS-CoV and HCoV-OC43. Antibodies raised against these antigens, which contain human N-glycosylation, show neutralizing activities against both viruses, suggesting that the coexpression system provides a quick and predictable way to produce multiple coronavirus antigens, such as the recent 2019 novel human coronavirus. [ABSTRACT FROM AUTHOR]

Quantization is an effective technique to reduce the memory and computational complexity of CNNs. Recent advances utilize additive powers-of-two to perform non-uniform quantization, which resembles a normal distribution and shows better performance than uniform quantization. With powers-of-two quantization, the computational complexity is also largely reduced because the slow multiplication operations are replaced with lightweight shift operations. However, there are serious problems in the previously proposed grid formulation for 2-bit quantization. In particular, these powers-of-two schemes produce zero values, generating significant training error and causing low accuracy. In addition, due to improper grid formulation, they also fallback to uniform quantization when the quantization level reaches 2-bit. Due to these reasons, on large CNN like ResNet-110, these powers-of-two schemes may not even train properly. To resolve these issues, we propose a new non-zero grid formulation that enables 2-bit non-uniform quantization and allow the CNN to be trained successfully in every attempt, even for a large network. The proposed technique quantizes weight as power-of-two values and projects it close to the mean area through a simple constant product on the exponential part. This allows our quantization scheme to closely resemble a non-uniform quantization at 2-bit, enabling successful training at 2-bit quantization, which is not found in the previous work. The proposed technique achieves 70.57% accuracy on the CIFAR-100 dataset trained with ResNet-110. This result is 6.24% higher than the additive powers-of-two scheme which only achieves 64.33% accuracy. Beside achieving higher accuracy, our work also maintains the same memory and computational efficiency with the original additive powers-of-two scheme.

We propose a self-supervised visual learning method by predicting the variable playback speeds of a video. Without semantic labels, we learn the spatio-temporal visual representation of the video by leveraging the variations in the visual appearance according to different playback speeds under the assumption of temporal coherence. To learn the spatio-temporal visual variations in the entire video, we have not only predicted a single playback speed but also generated clips of various playback speeds and directions with randomized starting points. Hence the visual representation can be successfully learned from the meta information (playback speeds and directions) of the video. We also propose a new layer-dependable temporal group normalization method that can be applied to 3D convolutional networks to improve the representation learning performance where we divide the temporal features into several groups and normalize each one using the different corresponding parameters. We validate the effectiveness of our method by fine-tuning it to the action recognition and video retrieval tasks on UCF-101 and HMDB-51. All the source code is released in https://github.com/hyeon-jo/PSPNet.

Assistive robots in home environments are steadily increasing in popularity. Due to significant variabilities in human behaviour, as well as physical characteristics and individual preferences, personalising assistance poses a challenging problem. In this paper, we focus on an assistive dressing task that involves physical contact with a human's upper body, in which the goal is to improve the comfort level of the individual. Two aspects are considered to be significant in improving a user's comfort level: having more natural postures and exerting less effort. However, a dressing path that fulfils these two criteria may not be found at one time. Therefore, we propose a user modelling method that combines vision and force data to enable the robot to search for an optimised dressing path for each user and improve as the human-robot interaction progresses. We compare the proposed method against two single-modality state-of-the-art user modelling methods designed for personalised assistive dressing by user studies (31 subjects). Experimental results show that the proposed method provides personalised assistance that results in more natural postures and less effort for human users.

Low-barrier hydrogen bonds (LBHBs) have been proposed to have important influences on the enormous reaction rate increases achieved by many enzymes. Δ5-3-ketosteroid isomerase (KSI) catalyzes the allylic isomerization of Δ5-3-ketosteroid to its conjugated Δ4-isomers at a rate that approaches the diffusion limit. Tyr14, a catalytic residue of KSI, has been hypothesized to form an LBHB with the oxyanion of a dienolate steroid intermediate generated during the catalysis. The unusual chemical shift of a proton at 16.8 ppm in the nuclear magnetic resonance spectrum has been attributed to an LBHB between Tyr14 Oη and C3-O of equilenin, an intermediate analogue, in the active site of D38N KSI. This shift in the spectrum was not observed in Y30F/Y55F/D38N and Y30F/Y55F/Y115F/D38N mutant KSIs when each mutant was complexed with equilenin, suggesting that Tyr14 could not form LBHB with the intermediate analogue in these mutant KSIs. The crystal structure of Y30F/Y55F/Y115F/D38N-equilenin complex revealed that the distance between Tyr14 Oη and C3-O of the bound steroid was within a direct hydrogen bond. The conversion of LBHB to an ordinary hydrogen bond in the mutant KSI reduced the binding affinity for the steroid inhibitors by a factor of 8.1-11. In addition, the absence of LBHB reduced the catalytic activity by only a factor of 1.7-2. These results suggest that the amount of stabilization energy of the reaction intermediate provided by LBHB is small compared with that provided by an ordinary hydrogen bond in KSI. [ABSTRACT FROM AUTHOR]

MutL is required to assist the mismatch repair protein MutS during initiation of the methyl-directed mismatch repair (MMR) response in various organisms ranging from prokaryotes to eukaryotes. Despite this necessity, the inherent propensity of MutL to aggregate has led to significant difficulties in determining its biological relationship with other MMR-related proteins. Here, we perform analysis on the thermostable MutL protein found in Thermotoga maritima MSB8 (TmL). Size exclusion chromatographic analysis indicates the lack of aggregated forms with the exception of a dimeric TmL. Small-angle X-ray scattering (SAXS) analysis reveals that the solution structures of the full-length TmL and its corresponding complexes with nucleotides and ssDNA undergo conformational changes. The elucidated TmL SAXS model is superimposed to the crystal structure of the C-terminal domain of Escherichia coli MutL. In addition, the N-terminal SAXS model of TmL exists as monomeric form, indicating that TmL has a structurally flexible N-terminal domain. TmL SAXS analysis can suggest a considerable possibility on a new 3D view of the previously unresolved full-length MutL molecule. [ABSTRACT FROM PUBLISHER]

The backbone dynamics of Y14F mutant of Δ5-3-ketosteroid isomerase (KSI) from Comamonas testosteroni has been studied in free enzyme and its complex with a steroid analogue, 19-nortestosterone hemisuccinate (19-NTHS), by 15N NMR relaxation measurements. Model-free analysis of the relaxation data showed that the single-point mutation induced a substantial decrease in the order parameters (S2) in free Y14F KSI, indicating that the backbone structures of Y14F KSI became significantly mobile by mutation, while the chemical shift analysis indicated that the structural perturbations of Y14F KSI were more profound than those of wild-type (WT) KSI upon 19-NTHS binding. In the 19-NTHS complexed Y14F KSI, however, the key active site residues including Tyr14, Asp38 and Asp99 or the regions around them remained flexible with significantly reduced S2 values, whereas the S2 values for many of the residues in Y14F KSI became even greater than those of WT KSI upon 19-NTHS binding. The results thus suggest that the hydrogen bond network in the active site might be disrupted by the Y14F mutation, resulting in a loss of the direct interactions between the catalytic residues and 19-NTHS. [ABSTRACT FROM PUBLISHER]

Multidimensional NMR was employed to investigate the structural changes in the urea-induced equilibrium unfolding of the dimeric ketosteroid isomerase (KSI) from Pseudomonas putida biotype B. Sequence specific backbone assignments for the native KSI and the protein with 3.5 M urea were carried out using various 3D NMR experiments. Hydrogen exchange measurements indicated that the secondary structures of KSI were not affected significantly by urea up to 3.5 M. However, the chemical shift analysis of 1H-15N HSQC spectra at various urea concentrations revealed that the residues in the dimeric interface region, particularly around the β5-strand, were significantly perturbed by urea at low concentrations, while the line-width analysis indicated the possibility of conformational exchange at the interface region around the β6-strand. The results thus suggest that the interface region primarily around the β5- and β6-strands could play an important role as the starting positions in the unfolding process of KSI. [ABSTRACT FROM PUBLISHER]

One of important issues in protein science is to increase the protein stability. Utilizing a ketosteroid isomerase (KSI) as a model system, we constructed a mutant KSI which has a higher stability than its WT. We focused on the role of aromatic moieties of three tyrosine residues, Tyr14, Tyr55, and Tyr30, which constitute a hydrogen bond network in KSI. Therefore, we made a series of mutants which could substitute tyrosine with phenylalanine to remove the hydroxyl group in the tyrosine residue. We measured the free-energy change for unfolding using urea as a denaturant. We investigated the interaction among these aromatic residues for stability using the double mutant cycle analysis. We found that additional mutations on Y14F, Y55F, and Y14F/Y55F, respectively, by replacement of tyrosine at the position 30 with phenylalanine increased the stability ca. 4.0 kcal/mol. Especially the triple mutant KSI (Y14F/Y30F/Y55F) was more stable 2 kcal/mol than wild type KSI. Our structural analyses suggests that the enhanced stability of the triple mutant should be due to increased both hydrophobic interaction and aromatic-aromatic interaction upon the mutation [ABSTRACT FROM AUTHOR]

The article informs about research on discovery of direct-acting antiviral agents with a graphene-based fluorescent nanosensor. Topics discussed include direct-acting drugs exist for the treatment against dengue virus (DV) infection, which can develop into life-threatening diseases; and development of an RNA-based graphene biosensor system [RNA nano-graphene oxide system (RANGO)] to enable the fluorescence-based quantitative analysis of the RdRp enzyme activity.