A Variational Autoencoder Cascade Generative Adversarial Network for Scalable 3D Object Generation and Reconstruction.

Generative Adversarial Networks (GANs) for 3D volume generation and reconstruction, such as shape generation, visualization, automated design, real-time simulation, and research applications, are receiving increased amounts of attention in various fields. However, challenges such as limited training data, high computational costs, and mode collapse issues persist. We propose combining a Variational Autoencoder (VAE) and a GAN to uncover enhanced 3D structures and introduce a stable and scalable progressive growth approach for generating and reconstructing intricate voxel-based 3D shapes. The cascade-structured network involves a generator and discriminator, starting with small voxel sizes and incrementally adding layers, while subsequently supervising the discriminator with ground-truth labels in each newly added layer to model a broader voxel space. Our method enhances the convergence speed and improves the quality of the generated 3D models through stable growth, thereby facilitating an accurate representation of intricate voxel-level details. Through comparative experiments with existing methods, we demonstrate the effectiveness of our approach in evaluating voxel quality, variations, and diversity. The generated models exhibit improved accuracy in 3D evaluation metrics and visual quality, making them valuable across various fields, including virtual reality, the metaverse, and gaming.

Monitoring the interactions between N-degrons and N-recognins of the Arg/N-degron pathway.

As defined by the N-degron pathway, single N-terminal (Nt) amino acids can function as N-degrons that induce the degradation of proteins and other biological materials. Central to this pathway is the selective recognition of N-degrons by cognate N-recognins that direct the substrates to either the ubiquitin (Ub)-proteasome system (UPS) or autophagy-lysosome pathway (ALP). Eukaryotic cells have developed diverse pathways to utilize all 20 amino acids in the genetic code as pro-N-degrons or N-degrons which can be generated through endoproteolytic cleavage or post-translational modifications. Amongst these, the arginine (Arg) N-degron plays a key role in both cis- and trans-degradation of a large spectrum of cellular materials by the proteasome or lysosome. In mammals, Arg/N-degrons can be generated through endoproteolytic cleavage or post-translational conjugation of the amino acid L-Arg by ATE1-encoded R-transferases (EC 2.3.2.8), which requires Arg-tRNAArg as a cofactor. Arg/N-degrons of short-lived substrates are recognized by a family of N-recognins characterized by the UBR box for polyubiquitination and proteasomal degradation. Under stresses, however, the same degrons can be recognized for autophagic degradation by the ZZ domain of the N-recognin p62/SQSTSM-1/Sequestosome-1 or KCMF1. Biochemical tools were developed to monitor the interaction of Arg/N-degrons with its cognate N-recognins. These assays were employed to identify new N-recognins and to characterize their biochemical properties and physiological functions. The principles of these assays may be applied for other types of N-degron pathways. Below, we describe the methods that analyze the interaction of Arg/N-degrons and their chemical mimics to N-recognins.

Targeted degradation of ⍺-synuclein aggregates in Parkinson's disease using the AUTOTAC technology.

BACKGROUND: There are currently no disease-modifying therapeutics for Parkinson's disease (PD). Although extensive efforts were undertaken to develop therapeutic approaches to delay the symptoms of PD, untreated α-synuclein (α-syn) aggregates cause cellular toxicity and stimulate further disease progression. PROTAC (Proteolysis-Targeting Chimera) has drawn attention as a therapeutic modality to target α-syn. However, no PROTACs have yet shown to selectively degrade α-syn aggregates mainly owing to the limited capacity of the proteasome to degrade aggregates, necessitating the development of novel approaches to fundamentally eliminate α-syn aggregates. METHODS: We employed AUTOTAC (Autophagy-Targeting Chimera), a macroautophagy-based targeted protein degradation (TPD) platform developed in our earlier studies. A series of AUTOTAC chemicals was synthesized as chimeras that bind both α-syn aggregates and p62/SQSTM1/Sequestosome-1, an autophagic receptor. The efficacy of Autotacs was evaluated to target α-syn aggregates to phagophores and subsequently lysosomes for hydrolysis via p62-dependent macroautophagy. The target engagement was monitored by oligomerization and localization of p62 and autophagic markers. The therapeutic efficacy to rescue PD symptoms was characterized in cultured cells and mice. The PK/PD (pharmacokinetics/pharmacodynamics) profiles were investigated to develop an oral drug for PD. RESULTS: ATC161 induced selective degradation of α-syn aggregates at DC50 of ~ 100 nM. No apparent degradation was observed with monomeric α-syn. ATC161 mediated the targeting of α-syn aggregates to p62 by binding the ZZ domain and accelerating p62 self-polymerization. These p62-cargo complexes were delivered to autophagic membranes for lysosomal degradation. In PD cellular models, ATC161 exhibited therapeutic efficacy to reduce cell-to-cell transmission of α-syn and to rescue cells from the damages in DNA and mitochondria. In PD mice established by injecting α-syn preformed fibrils (PFFs) into brain striata via stereotaxic surgery, oral administration of ATC161 at 10 mg/kg induced the degradation of α-syn aggregates and reduced their propagation. ATC161 also mitigated the associated glial inflammatory response and improved muscle strength and locomotive activity. CONCLUSION: AUTOTAC provides a platform to develop drugs for PD. ATC161, an oral drug with excellent PK/PD profiles, induces selective degradation of α-syn aggregates in vitro and in vivo. We suggest that ATC161 is a disease-modifying drug that degrades the pathogenic cause of PD.

The Cys-N-degron pathway modulates pexophagy through the N-terminal oxidation and arginylation of ACAD10.

In the N-degron pathway, N-recognins recognize cognate substrates for degradation via the ubiquitin (Ub)-proteasome system (UPS) or the autophagy-lysosome system (hereafter autophagy). We have recently shown that the autophagy receptor SQSTM1/p62 (sequestosome 1) is an N-recognin that binds the N-terminal arginine (Nt-Arg) as an N-degron to modulate autophagic proteolysis. Here, we show that the N-degron pathway mediates pexophagy, in which damaged peroxisomal fragments are degraded by autophagy under normal and oxidative stress conditions. This degradative process initiates when the Nt-Cys of ACAD10 (acyl-CoA dehydrogenase family, member 10), a receptor in pexophagy, is oxidized into Cys sulfinic (CysO2) or sulfonic acid (CysO3) by ADO (2-aminoethanethiol (cysteamine) dioxygenase). Under oxidative stress, the Nt-Cys of ACAD10 is chemically oxidized by reactive oxygen species (ROS). The oxidized Nt-Cys2 is arginylated by ATE1-encoded R-transferases, generating the RCOX N-degron. RCOX-ACAD10 marks the site of pexophagy via the interaction with PEX5 and binds the ZZ domain of SQSTM1/p62, recruiting LC3+-autophagic membranes. In mice, knockout of either Ate1 responsible for Nt-arginylation or Sqstm1/p62 leads to increased levels of peroxisomes. In the cells from patients with peroxisome biogenesis disorders (PBDs), characterized by peroxisomal loss due to uncontrolled pexophagy, inhibition of either ATE1 or SQSTM1/p62 was sufficient to recover the level of peroxisomes. Our results demonstrate that the Cys-N-degron pathway generates an N-degron that regulates the removal of damaged peroxisomal membranes along with their contents. We suggest that tannic acid, a commercially available drug on the market, has a potential to treat PBDs through its activity to inhibit ATE1 R-transferases.Abbreviations: ACAA1, acetyl-Coenzyme A acyltransferase 1; ACAD, acyl-Coenzyme A dehydrogenase; ADO, 2-aminoethanethiol (cysteamine) dioxygenase; ATE1, arginyltransferase 1; CDO1, cysteine dioxygenase type 1; ER, endoplasmic reticulum; LIR, LC3-interacting region; MOXD1, monooxygenase, DBH-like 1; NAC, N-acetyl-cysteine; Nt-Arg, N-terminal arginine; Nt-Cys, N-terminal cysteine; PB1, Phox and Bem1p; PBD, peroxisome biogenesis disorder; PCO, plant cysteine oxidase; PDI, protein disulfide isomerase; PTS, peroxisomal targeting signal; R-COX, Nt-Arg-CysOX; RNS, reactive nitrogen species; ROS, reactive oxygen species; SNP, sodium nitroprusside; UBA, ubiquitin-associated; UPS, ubiquitinproteasome system.

Graph Convolutional Network for 3D Object Pose Estimation in a Point Cloud.

Graph Neural Networks (GNNs) are neural networks that learn the representation of nodes and associated edges that connect it to every other node while maintaining graph representation. Graph Convolutional Neural Networks (GCNs), as a representative method in GNNs, in the context of computer vision, utilize conventional Convolutional Neural Networks (CNNs) to process data supported by graphs. This paper proposes a one-stage GCN approach for 3D object detection and poses estimation by structuring non-linearly distributed points of a graph. Our network provides the required details to analyze, generate and estimate bounding boxes by spatially structuring the input data into graphs. Our method proposes a keypoint attention mechanism that aggregates the relative features between each point to estimate the category and pose of the object to which the vertices of the graph belong, and also designs nine degrees of freedom of multi-object pose estimation. In addition, to avoid gimbal lock in 3D space, we use quaternion rotation, instead of Euler angle. Experimental results showed that memory usage and efficiency could be improved by aggregating point features from the point cloud and their neighbors in a graph structure. Overall, the system achieved comparable performance against state-of-the-art systems.

VR/AR Head-mounted Display System-based Measurement and Evaluation of Dynamic Visual Acuity.

This study evaluated the dynamic visual acuity of candidates by implementing a King-Devick (K-D) test chart in a virtual reality head-mounted display (VR HMD) and an augmented reality head-mounted display (AR HMD). Hard-copy KD (HCKD), VR HMD KD (VHKD), and AR HMD KD (AHKD) tests were conducted in 30 male and female candidates in the age of 10S and 20S and subjective symptom surveys were conducted. In the subjective symptom surveys, all except one of the VHKD questionnaire items showed subjective symptoms of less than 1 point. In the comparison between HCKD and VHKD, HCKD was measured more rapidly than VHKD in all tests. In the comparison between HCKD and AHKD, HCKD was measured more rapidly than AHKD in Tests 1, 2, and 3. In the comparison between VHKD and AHKD, AHKD was measured more rapidly than VHKD in Tests 1, 2, and 3. In the correlation analyses of test platforms, all platforms were correlated with each other, except for the correlation between HCKD and VHKD in Tests 1 and 2. There was no significant difference in the frequency of errors among Tests 1, 2, and 3 across test platforms. VHKD and AHKD, which require the body to be moved to read the chart, required longer measurement time than HCKD. In the measurements of each platform, AHKD was measured closer to HCKD than VHKD, which may be because the AHKD environment is closer to the actual environment than the VHKD environment. The effectiveness of VHKD and AHKD proposed in this research was evaluated experimentally. The results suggest that treatment and training could be performed concurrently through the use of clinical test and content development of VHKD and AHKD.