Correctable Landmark Discovery Via Large Models for Vision-Language Navigation.

Vision-Language Navigation (VLN) requires the agent to follow language instructions to reach a target position. A key factor for successful navigation is to align the landmarks implied in the instruction with diverse visual observations. However, previous VLN agents fail to perform accurate modality alignment especially in unexplored scenes, since they learn from limited navigation data and lack sufficient open-world alignment knowledge. In this work, we propose a new VLN paradigm, called COrrectable LaNdmark DiScOvery via Large ModEls (CONSOLE). In CONSOLE, we cast VLN as an open-world sequential landmark discovery problem, by introducing a novel correctable landmark discovery scheme based on two large models ChatGPT and CLIP. Specifically, we use ChatGPT to provide rich open-world landmark cooccurrence commonsense, and conduct CLIP-driven landmark discovery based on these commonsense priors. To mitigate the noise in the priors due to the lack of visual constraints, we introduce a learnable cooccurrence scoring module, which corrects the importance of each cooccurrence according to actual observations for accurate landmark discovery. We further design an observation enhancement strategy for an elegant combination of our framework with different VLN agents, where we utilize the corrected landmark features to obtain enhanced observation features for action decision. Extensive experimental results on multiple popular VLN benchmarks (R2R, REVERIE, R4R, RxR) show the significant superiority of CONSOLE over strong baselines. Especially, our CONSOLE establishes the new state-of-the-art results on R2R and R4R in unseen scenarios.

Optogenetic manipulation of lysosomal physiology and autophagy-dependent clearance of amyloid beta.

Lysosomes are degradation centers of cells and intracellular hubs of signal transduction, nutrient sensing, and autophagy regulation. Dysfunction of lysosomes contributes to a variety of diseases, such as lysosomal storage diseases (LSDs) and neurodegeneration, but the mechanisms are not well understood. Altering lysosomal activity and examining its impact on the occurrence and development of disease is an important strategy for studying lysosome-related diseases. However, methods to dynamically regulate lysosomal function in living cells or animals are still lacking. Here, we constructed lysosome-localized optogenetic actuators, named lyso-NpHR3.0, lyso-ArchT, and lyso-ChR2, to achieve optogenetic manipulation of lysosomes. These new actuators enable light-dependent control of lysosomal membrane potential, pH, hydrolase activity, degradation, and Ca2+ dynamics in living cells. Notably, lyso-ChR2 activation induces autophagy through the mTOR pathway, promotes Aß clearance in an autophagy-dependent manner in cellular models, and alleviates Aß-induced paralysis in the Caenorhabditis elegans model of Alzheimer's disease. Our lysosomal optogenetic actuators supplement the optogenetic toolbox and provide a method to dynamically regulate lysosomal physiology and function in living cells and animals.

Locus-specific detection of pseudouridine with CRISPR-Cas13a.

We combined the CRISPR-Cas13a system with CMC chemical labeling, developing an approach that enables precise identification of pseudouridine (Ψ) sites at specific loci within ribosomal RNA (rRNA), messenger RNA (mRNA) and small nuclear RNAs (snRNA). This method, with good efficiency and simplicity, detects Ψ sites through fluorescence measurement, providing a straightforward and fast validation for targeted Ψ sites of interest.

Engineered aptamers for molecular imaging.

Molecular imaging, including quantification and molecular interaction studies, plays a crucial role in visualizing and analysing molecular events occurring within cells or organisms, thus facilitating the understanding of biological processes. Moreover, molecular imaging offers promising applications for early disease diagnosis and therapeutic evaluation. Aptamers are oligonucleotides that can recognize targets with a high affinity and specificity by folding themselves into various three-dimensional structures, thus serving as ideal molecular recognition elements in molecular imaging. This review summarizes the commonly employed aptamers in molecular imaging and outlines the prevalent design approaches for their applications. Furthermore, it highlights the successful application of aptamers to a wide range of targets and imaging modalities. Finally, the review concludes with a forward-looking perspective on future advancements in aptamer-based molecular imaging.

Isolation of PD-L1 Extracellular Vesicle Subpopulations Using DNA Computation Mediated Microfluidic Tandem Separation.

Accurate isolation of targeted extracellular vesicle (EV) is challenging due to the antigenic heterogeneity of EV subpopulations which are from different cell origins. Most EV subpopulations lack a single marker whose expression cleanly distinguishes them from mixed populations of closely related EVs. Here, a modular platform capable of taking multiple binding events as input, performing logic computations, and producing two independent outputs for tandem microchips for EV subpopulation isolation, is developed. Taking advantages of the excellent selectivity of dual-aptamer recognition and the sensitivity of tandem microchips, this method achieves, for the first time, sequential isolation of tumor PD-L1 EVs and non-tumor PD-L1 EVs. As a result, the developed platform can not only effectively distinguish cancer patients from healthy donors but also provides new clues for assessing immune heterogeneity. Moreover, the captured EVs can be released through a DNA hydrolysis reaction with high efficiency, which is compatible with downstream mass spectrometry for EV proteome profiling. Overall, this strategy is expected to isolate different EV subpopulations, translate EVs into reliable clinical biomarkers, and accurately investigate the biological functions of different EV subsets.

Towards Deviation-Robust Agent Navigation via Perturbation-Aware Contrastive Learning.

Vision-and-language navigation (VLN) asks an agent to follow a given language instruction to navigate through a real 3D environment. Despite significant advances, conventional VLN agents are trained typically under disturbance-free environments and may easily fail in real-world navigation scenarios, since they are unaware of how to deal with various possible disturbances, such as sudden obstacles or human interruptions, which widely exist and may usually cause an unexpected route deviation. In this paper, we present a model-agnostic training paradigm, called Progressive Perturbation-aware Contrastive Learning (PROPER) to enhance the generalization ability of existing VLN agents to the real world, by requiring them to learn towards deviation-robust navigation. Specifically, a simple yet effective path perturbation scheme is introduced to implement the route deviation, with which the agent is required to still navigate successfully following the original instruction. Since directly enforcing the agent to learn perturbed trajectories may lead to insufficient and inefficient training, a progressively perturbed trajectory augmentation strategy is designed, where the agent can self-adaptively learn to navigate under perturbation with the improvement of its navigation performance for each specific trajectory. For encouraging the agent to well capture the difference brought by perturbation and adapt to both perturbation-free and perturbation-based environments, a perturbation-aware contrastive learning mechanism is further developed by contrasting perturbation-free trajectory encodings and perturbation-based counterparts. Extensive experiments on the standard Room-to-Room (R2R) benchmark show that PROPER can benefit multiple state-of-the-art VLN baselines in perturbation-free scenarios. We further collect the perturbed path data to construct an introspection subset based on the R2R, called Path-Perturbed R2R (PP-R2R). The results on PP-R2R show unsatisfying robustness of popular VLN agents and the capability of PROPER in improving the navigation robustness under deviation.

Aptamer-LYTACs for Targeted Degradation of Extracellular and Membrane Proteins.

Recently, lysosome targeting chimeras (LYTACs) have emerged as a promising technology that expands the scope of targeted protein degradation to extracellular targets. However, the preparation of chimeras by conjugation of the antibody and trivalent N-acetylgalactosamine (tri-GalNAc) is a complex and time-consuming process. The large uncertainty in number and position and the large molecular weights of the chimeras result in low internalization efficiency. To circumvent these problems, we developed the first aptamer-based LYTAC (Apt-LYTAC) to realize liver-cell-specific degradation of extracellular and membrane proteins by conjugating aptamers to tri-GalNAc. Taking advantage of the facile synthesis and low molecular weight of the aptamer, the Apt-LYTACs can efficiently and quickly degrade the extracellular protein PDGF and the membrane protein PTK7 through a lysosomal degradation pathway. We anticipate that the novel Apt-LYTACs will expand the usage of aptamers and provide a new dimension for targeted protein degradation.

Antibody responses to COVID-19 vaccination in people with obesity: A systematic review and meta-analysis.

COVID-19 vaccine is critical in preventing SARS-CoV-2 infection and transmission. However, obesity's effect on immune responses to COVID-19 vaccines is still unknown. We performed a meta-analysis of the literature and compared antibody responses with COVID-19 vaccines among persons with and without obesity. We used Pubmed, Embase, Web of Science, and Cochrane Library to identify all related studies up to April 2022. The Stata.14 software was used to analyze the selected data. Eleven studies were included in the present meta-analysis. Five of them provided absolute values of antibody titers in the obese group and non-obese group. Overall, we found that the obese population was significantly associated with lower antibody titers (standardized mean difference [SMD] = -0.228, 95% CI [-0.437, -0.019], P < 0.001) after COVID-19 vaccination. Significant heterogeneity was present in most pooled analyses but was reduced after subgroup analyses. No publication bias was observed in the present analysis. The Trim and Fill method did not change the results in the primary analysis. The present meta-analysis suggested that obesity was significantly associated with decreased antibody responses to SARS-CoV-2 vaccines. Future studies should be performed to unravel the mechanism of response to the COVID-19 vaccine in obese individuals.

Recent Advances in Exosomal miRNA Biosensing for Liquid Biopsy.

As a noninvasive detection technique, liquid biopsy plays a valuable role in cancer diagnosis, disease monitoring, and prognostic assessment. In liquid biopsies, exosomes are considered among the potential biomarkers because they are important bioinformation carriers for intercellular communication. Exosomes transport miRNAs and, thus, play an important role in the regulation of cell growth and function; therefore, detection of cancer cell-derived exosomal miRNAs (exo-miRNAs) gives effective information in liquid biopsy. The development of sensitive, convenient, and reliable exo-miRNA assays will provide new perspectives for medical diagnosis. This review presents different designs and detection strategies of recent exo-miRNA assays in terms of signal transduction and amplification, as well as signal detection. In addition, this review outlines the current attempts at bioassay methods in liquid biopsies. Lastly, the challenges and prospects of exosome bioassays are also considered.

Lysosomal K+ channel TMEM175 promotes apoptosis and aggravates symptoms of Parkinson's disease.

Lysosomes are degradative organelles and play vital roles in a variety of cellular processes. Ion channels on the lysosomal membrane are key regulators of lysosomal function. TMEM175 has been identified as a lysosomal potassium channel, but its modulation and physiological functions remain unclear. Here, we show that the apoptotic regulator Bcl-2 binds to and inhibits TMEM175 activity. Accordingly, Bcl-2 inhibitors activate the channel in a caspase-independent way. Increased TMEM175 function inhibits mitophagy, disrupts mitochondrial homeostasis, and increases production of reactive oxygen species (ROS). ROS further activates TMEM175 and thus forms a positive feedback loop to augment apoptosis. In a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of Parkinson's disease (PD), knockout (KO) of TMEM175 mitigated motor impairment and dopaminergic (DA) neuron loss, suggesting that TMEM175-mediated apoptosis plays an important role in Parkinson's disease (PD). Overall, our study reveals that TMEM175 is an important regulatory site in the apoptotic signaling pathway and a potential therapeutic target for Parkinson's disease (PD).

Quantification-Promoted Discovery of Glycosylated Exosomal PD-L1 as a Potential Tumor Biomarker.

Exosomal programmed cell death ligand 1 (exoPD-L1) has emerged as a promising biomarker for cancer diagnosis and immunotherapy outcome prediction. However, the existing quantitation methods are incapable of addressing the heterogeneity of exoPD-L1 glycosylation, which has been demonstrated to be the institutional basis for PD-L1/PD-1 interaction and the crucial participant in inhibiting the activity of CD8+ T cells. Herein, an aptamer- and lectin-induced proximity ligation assay combined with quantitative real-time polymerase chain reaction for precise quantitation of glycosylated exoPD-L1 is developed. Leveraging the metabolism-free lectin labeling of glycosylation, the glycosylation-independent aptamer tagging of PD-L1, and excellent selectivity of dual-recognition, this method enables glycosylated exoPD-L1 quantitation with high sensitivity and selectivity in a wash-free manner. As a result, this method is able to distinguish the levels of glycosylated exoPD-L1 between healthy donors and cancer patients with sensitivity and specificity of 100%. Compared with the total circulating exoPD-L1 level, glycosylated exoPD-L1 is for the first time identified to be a more reliable biomarker for tumor diagnosis. Overall, this strategy holds a great potential for revealing the significance of exoPD-L1 glycosylation and converting glycosylated exoPD-L1 into a reliable clinical indicator.

Configurable Graph Reasoning for Visual Relationship Detection.

Visual commonsense knowledge has received growing attention in the reasoning of long-tailed visual relationships biased in terms of object and relation labels. Most current methods typically collect and utilize external knowledge for visual relationships by following the fixed reasoning path of {subject, object → predicate} to facilitate the recognition of infrequent relationships. However, the knowledge incorporation for such fixed multidependent path suffers from the data set biased and exponentially grown combinations of object and relation labels and ignores the semantic gap between commonsense knowledge and real scenes. To alleviate this, we propose configurable graph reasoning (CGR) to decompose the reasoning path of visual relationships and the incorporation of external knowledge, achieving configurable knowledge selection and personalized graph reasoning for each relation type in each image. Given a commonsense knowledge graph, CGR learns to match and retrieve knowledge for different subpaths and selectively compose the knowledge routed path. CGR adaptively configures the reasoning path based on the knowledge graph, bridges the semantic gap between the commonsense knowledge, and the real-world scenes and achieves better knowledge generalization. Extensive experiments show that CGR consistently outperforms previous state-of-the-art methods on several popular benchmarks and works well with different knowledge graphs. Detailed analyses demonstrated that CGR learned explainable and compelling configurations of reasoning paths.

Adversarial Reinforced Instruction Attacker for Robust Vision-Language Navigation.

Language instruction plays an essential role in the natural language grounded navigation tasks. However, navigators trained with limited human-annotated instructions may have difficulties in accurately capturing key information from the complicated instruction at different timesteps, leading to poor navigation performance. In this paper, we exploit to train a more robust navigator which is capable of dynamically extracting crucial factors from the long instruction, by using an adversarial attacking paradigm. Specifically, we propose a Dynamic Reinforced Instruction Attacker (DR-Attacker), which learns to mislead the navigator to move to the wrong target by destroying the most instructive information in instructions at different timesteps. By formulating the perturbation generation as a Markov Decision Process, DR-Attacker is optimized by the reinforcement learning algorithm to generate perturbed instructions sequentially during the navigation, according to a learnable attack score. Then, the perturbed instructions, which serve as hard samples, are used for improving the robustness of the navigator with an effective adversarial training strategy and an auxiliary self-supervised reasoning task. Experimental results on both Vision-and-Language Navigation (VLN) and Navigation from Dialog History (NDH) tasks show the superiority of our proposed method over state-of-the-art methods. Moreover, the visualization analysis shows the effectiveness of the proposed DR-Attacker, which can successfully attack crucial information in the instructions at different timesteps. Code is available at https://github.com/expectorlin/DR-Attacker.

In situ Raman enhancement strategy for highly sensitive and quantitative lateral flow assay.

As a paper-based analytical platform, lateral flow assay (LFA) gets benefit from the rapid analysis, low cost, high selectivity, good stability, and user-friendliness, and thus has been widely used in rapid screening or assisted diagnosis. Nevertheless, LFA still suffers from low detection sensitivity via the naked eye, limiting its applications to qualitative and semi-quantitative tests. To enhance the signal readout, various nanoparticle signal tags have been employed to replace traditional colloidal gold nanoparticles (AuNPs), such as fluorescent nanoparticles (FNPs), magnetic nanoparticles (MNPs), and Raman reporter-labeled nanoparticles. In particular, Raman reporter-labeled nanoparticles are extremely sensitive due to remarkable signal enhancement effect on metal surface. However, the application of LFA is still hampered by the poor stability of Raman reporter-labeled nanoparticles. Herein, we developed an in situ Raman enhancement strategy to create a surface-enhanced Raman scattering (SERS) signal on the AuNPs, shortened as "i-SERS," which not only preserves the original advantages of the colloidal gold strip (AuNPs-LFA), but also realizes highly sensitive and quantitative detection. We applied the i-SERS for procalcitonin (PCT) detection. The experimental process takes only 16 min, and the limit of detection (LOD) is 0.03 ng mL-1, far below the value using AuNPs-LFA. These results indicate that i-SERS assay was highly sensitive and suitable for the rapid detection of PCT.

Microfluidic-Based Exosome Analysis for Liquid Biopsy.

Liquid biopsy offers non-invasive and real-time molecular profiling of individual patients, and is thus considered a revolutionary technology in precision medicine. Exosomes have been acknowledged as significant biomarkers in liquid biopsy, as they play a central role in cell-cell communication and are closely related to the pathogenesis of most human malignancies. Nevertheless, in biofluids exosomes always co-exist with other particles, and the cargo components of exosomes are highly heterogeneous. Thus, the isolation and molecular characterization of exosomes are still technically challenging. Microfluidics technology effectively addresses this challenge by virtue of its inherent advantages, such as precise manipulation of fluids, low consumption of samples and reagents, and a high level of integration. Recent advances in microfluidics allow in situ exosome capture and molecular detection with unprecedented selectivity and sensitivity. In this review, the state-of-the-art developments in microfluidics-based exosome research, including exosome isolation approaches and molecular detection strategies, with highlights of the characterization of exosomal biomarkers in cancer liquid biopsy is summarized. The major challenges are also discussed and some perspectives for the future directions of exosome-based liquid biopsy in microfluidic systems are presented.

CLN7 is an organellar chloride channel regulating lysosomal function.

Neuronal ceroid lipofuscinoses (NCLs) are a group of autosomal recessive lysosomal storage diseases. One variant form of late-infantile NCL (vLINCL) is caused by mutations of a lysosomal membrane protein CLN7, the function of which has remained unknown. Here, we identified CLN7 as a novel endolysosomal chloride channel. Overexpression of CLN7 increases endolysosomal chloride currents and enlarges endolysosomes through a Ca2+/calmodulin-dependent way. Human CLN7 and its yeast homolog exhibit characteristics of chloride channels and are sensitive to chloride channel blockers. Moreover, CLN7 regulates lysosomal chloride conductance, luminal pH, and lysosomal membrane potential and promotes the release of lysosomal Ca2+ through transient receptor potential mucolipin 1 (TRPML1). Knocking out CLN7 causes pathological features that are similar to those of patients with vLINCL, including retinal degeneration and autofluorescent lipofuscin. The pathogenic mutations in CLN7 lead to a decrease in chloride permeability, suggesting that reconstitution of lysosomal Cl− homeostasis may be an effective strategy for the treatment of vLINCL.

Reversible Immunoaffinity Interface Enables Dynamic Manipulation of Trapping Force for Accumulated Capture and Efficient Release of Circulating Rare Cells.

Controllable assembly and disassembly of recognition interface are vital for bioanalysis. Herein, a strategy of dynamic manipulation of trapping force by engineering a dynamic and reversible immunoaffinity microinterface (DynarFace) in a herringbone chip (DynarFace-Chip) for liquid biopsy is proposed. The DynarFace is assembled by magnetically attracting immunomagnetic beads (IMBs) on chip substrate, with merits of convenient operation and reversible assembly. The DynarFace allows accumulating attachment of IMBs on circulating rare cell (CRC) surfaces during hydrodynamically enhanced interface collision, where accumulatively enhanced magnetic trapping force improves capture efficiency toward CRCs with medium expression of biomarkers from blood samples by 134.81% compared with traditional non-dynamic interfaces. Moreover, magnet withdrawing-induced disappearance of trapping force affords DynarFace disassembly and CRC release with high efficiency (>98%) and high viability (≈98%), compatible with downstream in vitro culture and gene analysis of CRCs. This DynarFace strategy opens a new avenue to accumulated capture and reversible release of CRCs, holding great potential for liquid biopsy-based precision medicine.

HUNTER-Chip: Bioinspired Hierarchically Aptamer Structure-Based Circulating Fetal Cell Isolation for Non-Invasive Prenatal Testing.

Isolation and genetic analysis of circulating fetal cells from billions of maternal cells in peripheral blood are the cornerstone of fetal cell-based non-invasive prenatal testing. Inspired by the hierarchically multivalent architecture for enhanced capture of nature, an aptamer-based Hierarchically mUltivalent aNTibody mimic intERface (HUNTER) was designed with a tremendous avidity effect for highly efficient capture and non-destructive release of fetal cells. It was engineered by grafting Y-shaped DNA nanostructures to a linear polymer chain, creating a flexible polymer chain with bivalent aptamer side chains. This hierarchical arrangement of the aptamer ensures morphological complementarity, collective multiple-site interaction, and multivalent recognition between the aptamer and target cells. In combination with a deterministic lateral displacement (DLD)-patterned microdevice named as HUNTER-Chip, it achieves a binding affinity over 65-fold and a capture efficiency over 260%-fold due to the combination of hierarchically designed aptamers and frequent cell-ligand collision created by DLD. Moreover, a nuclease-assisted cell release strategy facilitates the release of fetal cells for gene analysis, such as fluorescence in situ hybridization. With the advantages of high affinity, excellent capture efficiency, and compatible downstream analysis, the HUNTER-Chip holds great potential for non-invasive prenatal diagnosis.

Coupling Aptamer-based Protein Tagging with Metabolic Glycan Labeling for In Situ Visualization and Biological Function Study of Exosomal Protein-Specific Glycosylation.

Exosomal glycoproteins play important roles in many physiological and pathological functions. Herein, we developed a dual labeling strategy based on a protein-specific aptamer tagging and metabolic glycan labeling for visualizing glycosylation of specific proteins on exosomes. The glycosylation of exosomal PD-L1 (exoPD-L1) was imaged in situ using intramolecular fluorescence resonance energy transfer (FRET) between fluorescent PD-L1 aptamers bound on exoPD-L1 and fluorescent tags on glycans introduced via metabolic glycan labeling. This method enables in situ visualization and biological function study of exosomal protein glycosylation. Exosomal PD-L1 glycosylation was confirmed to be required in interaction with PD-1 and participated in inhibiting of CD8+ T cell proliferation. This is an efficient and non-destructive method to study the presence and function of exosomal protein-specific glycosylation in situ, which provides a powerful tool for exosomal glycoproteomics research.

Aptamer-Based Detection of Circulating Targets for Precision Medicine.

The past decade has witnessed ongoing progress in precision medicine to improve human health. As an emerging diagnostic technique, liquid biopsy can provide real-time, comprehensive, dynamic physiological and pathological information in a noninvasive manner, opening a new window for precision medicine. Liquid biopsy depends on the sensitive and reliable detection of circulating targets (e.g., cells, extracellular vesicles, proteins, microRNAs) from body fluids, the performance of which is largely governed by recognition ligands. Aptamers are single-stranded functional oligonucleotides, capable of folding into unique tertiary structures to bind to their targets with superior specificity and affinity. Their mature evolution procedure, facile modification, and affinity regulation, as well as versatile structural design and engineering, make aptamers ideal recognition ligands for liquid biopsy. In this review, we present a broad overview of aptamer-based liquid biopsy techniques for precision medicine. We begin with recent advances in aptamer selection, followed by a summary of state-of-the-art strategies for multivalent aptamer assembly and aptamer interface modification. We will further describe aptamer-based micro-/nanoisolation platforms, aptamer-enabled release methods, and aptamer-assisted signal amplification and detection strategies. Finally, we present our perspectives regarding the opportunities and challenges of aptamer-based liquid biopsy for precision medicine.