DOVE: Doodled vessel enhancement for photoacoustic angiography super resolution.

Deep-learning-based super-resolution photoacoustic angiography (PAA) has emerged as a valuable tool for enhancing the resolution of blood vessel images and aiding in disease diagnosis. However, due to the scarcity of training samples, PAA super-resolution models do not generalize well, especially in the challenging in-vivo imaging of organs with deep tissue penetration. Furthermore, prolonged exposure to high laser intensity during the image acquisition process can lead to tissue damage and secondary infections. To address these challenges, we propose an approach doodled vessel enhancement (DOVE) that utilizes hand-drawn doodles to train a PAA super-resolution model. With a training dataset consisting of only 32 real PAA images, we construct a diffusion model that interprets hand-drawn doodles as low-resolution images. DOVE enables us to generate a large number of realistic PAA images, achieving a 49.375% fool rate, even among experts in photoacoustic imaging. Subsequently, we employ these generated images to train a self-similarity-based model for super-resolution. During cross-domain tests, our method, trained solely on generated images, achieves a structural similarity value of 0.8591, surpassing the scores of all other models trained with real high-resolution images. DOVE successfully overcomes the limitation of insufficient training samples and unlocks the clinic application potential of super-resolution-based biomedical imaging.

Thrombospondin 1 and Reelin act through Vldlr to regulate cardiac growth and repair.

Adult mammalian cardiomyocytes have minimal cell cycle capacity, which leads to poor regeneration after cardiac injury such as myocardial infarction. Many positive regulators of cardiomyocyte cell cycle and cardioprotective signals have been identified, but extracellular signals that suppress cardiomyocyte proliferation are poorly understood. We profiled receptors enriched in postnatal cardiomyocytes, and found that very-low-density-lipoprotein receptor (Vldlr) inhibits neonatal cardiomyocyte cell cycle. Paradoxically, Reelin, the well-known Vldlr ligand, expressed in cardiac Schwann cells and lymphatic endothelial cells, promotes neonatal cardiomyocyte proliferation. Thrombospondin1 (TSP-1), another ligand of Vldlr highly expressed in adult heart, was then found to inhibit cardiomyocyte proliferation through Vldlr, and may contribute to Vldlr's overall repression on proliferation. Mechanistically, Rac1 and subsequent Yap phosphorylation and nucleus translocation mediate the regulation of the cardiomyocyte cell cycle by TSP-1/Reelin-Vldlr signaling. Importantly, Reln mutant neonatal mice displayed impaired cardiomyocyte proliferation and cardiac regeneration after apical resection, while cardiac-specific Thbs1 deletion and cardiomyocyte-specific Vldlr deletion promote cardiomyocyte proliferation and are cardioprotective after myocardial infarction. Our results identified a novel role of Vldlr in consolidating extracellular signals to regulate cardiomyocyte cell cycle activity and survival, and the overall suppressive TSP-1-Vldlr signal may contribute to the poor cardiac repair capacity of adult mammals.

Impaired end joining induces cardiac atrophy in a Hutchinson-Gilford progeria mouse model.

Patients with Hutchinson-Gilford progeria syndrome (HGPS) present with a number of premature aging phenotypes, including DNA damage accumulation, and many of them die of cardiovascular complications. Although vascular pathologies have been reported, whether HGPS patients exhibit cardiac dysfunction and its underlying mechanism is unclear, rendering limited options for treating HGPS-related cardiomyopathy. In this study, we reported a cardiac atrophy phenotype in the LmnaG609G/G609G mice (hereafter, HGPS mice). Using a GFP-based reporter system, we demonstrated that the efficiency of nonhomologous end joining (NHEJ) declined by 50% in HGPS cardiomyocytes in vivo, due to the attenuated interaction between γH2AX and Progerin, the causative factor of HGPS. As a result, genomic instability in cardiomyocytes led to an increase of CHK2 protein level, promoting the LKB1-AMPKα interaction and AMPKα phosphorylation, which further led to the activation of FOXO3A-mediated transcription of atrophy-related genes. Moreover, inhibiting AMPK enlarged cardiomyocyte sizes both in vitro and in vivo. Most importantly, our proof-of-concept study indicated that isoproterenol treatment significantly reduced AMPKα and FOXO3A phosphorylation in the heart, attenuated the atrophy phenotype, and extended the mean lifespan of HGPS mice by ~21%, implying that targeting cardiac atrophy may be an approach to HGPS treatment.

Jmjd4 Facilitates Pkm2 Degradation in Cardiomyocytes and Is Protective Against Dilated Cardiomyopathy.

BACKGROUND: A large portion of idiopathic and familial dilated cardiomyopathy (DCM) cases have no obvious causal genetic variant. Although altered response to metabolic stress has been implicated, the molecular mechanisms underlying the pathogenesis of DCM remain elusive. The JMJD family proteins, initially identified as histone deacetylases, have been shown to be involved in many cardiovascular diseases. Despite their increasingly diverse functions, whether JMJD family members play a role in DCM remains unclear. METHODS: We examined Jmjd4 expression in patients with DCM, and conditionally deleted and overexpressed Jmjd4 in cardiomyocytes in vivo to investigate its role in DCM. RNA sequencing, metabolites profiling, and mass spectrometry were used to dissect the molecular mechanism of Jmjd4-regulating cardiac metabolism and hypertrophy. RESULTS: We found that expression of Jmjd4 is significantly decreased in hearts of patients with DCM. Induced cardiomyocyte-specific deletion of Jmjd4 led to spontaneous DCM with severely impaired mitochondrial respiration. Pkm2, the less active pyruvate kinase compared with Pkm1, which is normally absent in healthy adult cardiomyocytes but elevated in cardiomyopathy, was found to be drastically accumulated in hearts with Jmjd4 deleted. Jmjd4 was found mechanistically to interact with Hsp70 to mediate degradation of Pkm2 through chaperone-mediated autophagy, which is dependent on hydroxylation of K66 of Pkm2 by Jmjd4. By enhancing the enzymatic activity of the abundant but less active Pkm2, TEPP-46, a Pkm2 agonist, showed a significant therapeutic effect on DCM induced by Jmjd4 deficiency, and heart failure induced by pressure overload, as well. CONCLUSIONS: Our results identified a novel role of Jmjd4 in maintaining metabolic homeostasis in adult cardiomyocytes by degrading Pkm2 and suggest that Jmjd4 and Pkm2 may be therapeutically targeted to treat DCM, and other cardiac diseases with metabolic dysfunction, as well.

Inhibition of Fap Promotes Cardiac Repair by Stabilizing BNP.

BACKGROUND: Myocardial infarction (MI) elicits cardiac fibroblast activation and extracellular matrix (ECM) deposition to maintain the structural integrity of the heart. Recent studies demonstrate that Fap (fibroblast activation protein)-a prolyl-specific serine protease-is an important marker of activated cardiac fibroblasts after MI. METHODS: Left ventricle and plasma samples from patients and healthy donors were used to analyze the expression level of FAP and its prognostic value. Echocardiography and histological analysis of heart sections were used to analyze cardiac functions, scar formation, ECM deposition and angiogenesis after MI. RNA-Sequencing, biochemical analysis, cardiac fibroblasts (CFs) and endothelial cells co-culture were used to reveal the molecular and cellular mechanisms by which Fap regulates angiogenesis. RESULTS: We found that Fap is upregulated in patient cardiac fibroblasts after cardiac injuries, while plasma Fap is downregulated and functions as a prognostic marker for cardiac repair. Genetic or pharmacological inhibition of Fap in mice significantly improved cardiac function after MI. Histological and transcriptomic analyses showed that Fap inhibition leads to increased angiogenesis in the peri-infarct zone, which promotes ECM deposition and alignment by cardiac fibroblasts and prevents their overactivation, thereby limiting scar expansion. Mechanistically, we found that BNP (brain natriuretic peptide) is a novel substrate of Fap that mediates postischemic angiogenesis. Fap degrades BNP to inhibit vascular endothelial cell migration and tube formation. Pharmacological inhibition of Fap in Nppb (encoding pre-proBNP) or Npr1 (encoding the BNP receptor)-deficient mice showed no cardioprotective effects, suggesting that BNP is a physiological substrate of Fap. CONCLUSIONS: This study identifies Fap as a negative regulator of cardiac repair and a potential drug target to treat MI. Inhibition of Fap stabilizes BNP to promote angiogenesis and cardiac repair.

Gamma-glutamyl transferase to high-density lipoprotein cholesterol ratio: A valuable predictor of type 2 diabetes mellitus incidence.

Background: Gamma-glutamyl transferase (GGT) and high-density lipoprotein cholesterol (HDL-C) have been proven to be valuable predictors of type 2 diabetes mellitus (T2DM). The aim of this study was to investigate the association between GGT/HDL-C ratio and incident T2DM. Methods: The study retrospectively analyzed 15453 participants from 2004 to 2015. Cox proportional hazards regression models and Kaplan-Meier curves were used to elucidate the effect of GGT/HDL-C ratio on T2DM. Restricted cubic spline (RCS) analysis was performed to explore any non-linear correlation between GGT/HDL-C ratio and the risk of T2DM. The predictive performance of GGT, HDL-C and GGT/HDL-C ratio for T2DM was evaluated utilizing receiver-operating-characteristic (ROC) curves. Results: During a median follow-up of 5.39 years, 373 cases of incident T2DM were observed. Kaplan-Meier curves showed that the cumulative probabilities of T2DM increased in the participants with higher GGT/HDL-C ratio significantly (P < 0.001). Cox models further clarified that high GGT/HDL-C ratio was an independent risk factor for T2DM (HR = 1.01, 95% CI = 1.00-1.01, P = 0.011). Linear positive correlation between GGT/HDL-C ratio and the risk of T2DM was demonstrated through RCS analysis. In the ROC analysis, GGT/HDL-C ratio (AUC = 0.75, 95% CI = 0.73-0.77) showed competitive role in the prediction of T2DM compared with single GGT and HDL-C. Conclusions: The GGT/HDL-C ratio could serve as a valuable predictor of T2DM, and the risk of T2DM increases in the condition of higher GGT/HDL-C ratio.

VideoABC: A Real-World Video Dataset for Abductive Visual Reasoning.

In this paper, we investigate the problem of abductive visual reasoning (AVR), which requires vision systems to infer the most plausible explanation for visual observations. Unlike previous work which performs visual reasoning on static images or synthesized scenes, we exploit long-term reasoning from instructional videos that contain a wealth of detailed information about the physical world. We conceptualize two tasks for this emerging and challenging topic. The primary task is AVR, which is based on the initial configuration and desired goal from an instructional video, and the model is expected to figure out what is the most plausible sequence of steps to achieve the goal. In order to avoid trivial solutions based on appearance information rather than reasoning, the second task called AVR++ is constructed, which requires the model to answer why the unselected options are less plausible. We introduce a new dataset called VideoABC, which consists of 46,354 unique steps derived from 11,827 instructional videos, formulated as 13,526 abductive reasoning questions with an average reasoning duration of 51 seconds. Through an adversarial hard hypothesis mining algorithm, non-trivial and high-quality problems are generated efficiently and effectively. To achieve human-level reasoning, we propose a Hierarchical Dual Reasoning Network (HDRNet) to capture the long-term dependencies among steps and observations. We establish a benchmark for abductive visual reasoning, and our method set state-of-the-arts on AVR (  âˆ¼ 74 %) and AVR++ (  âˆ¼ 45 %), and humans can easily achieve over 90% accuracy on these two tasks. The large performance gap reveals the limitation of current video understanding models on temporal reasoning and leaves substantial room for future research on this challenging problem. Our dataset and code are available at https://github.com/wl-zhao/VideoABC.

PLK inhibitors identified by high content phenotypic screening promote maturation of human PSC-derived cardiomyocytes.

Human pluripotent stem cells-derived cardiomyocytes (hPSC-CMs) provide an unlimited source of human cardiomyocytes for disease modeling, cell therapies, and other biomedical applications. However, hPSC-CMs remain developmentally immature which limits their suitability in translational applications. High Content Screening (HCS) is a powerful tool for identifying novel molecules and pathways regulating complex biological processes, but no HCS assay for hPSC-CM maturation has yet been reported. PCM1, a centriole satellite protein, is specifically restricted on nuclear envelope in mature cardiomyocytes. We developed a High Content Screen (HCS) based on PCM1 subcellular localization in hPSC-CMs to identify novel molecules promoting cardiomyocyte maturation, which identified 93 from 1693 compounds that enhance maturation of hPSC-CMs, including multiple PLK inhibitors. Volasertib and Centrinone, two PLK inhibitors, can enhance binucleation, and promote metabolic and electrophysiological maturation in hPSC-CMs. Furthermore, PI3K-AKT signaling pathway was found to be suppressed by PLK inhibitors, and VO-Ohpic, a PTEN inhibitor that activates AKT pathway, blunted the effect of PLK inhibitors on hPSC-CM maturation. In summary, our HCS assay found that PLK inhibitors can promote maturation of hPSC-CMs through suppressing AKT signaling pathway.

Circulating miR-19b-3p as a Novel Prognostic Biomarker for Acute Heart Failure.

Background Circulating microRNAs are emerging biomarkers for heart failure (HF). Our study aimed to assess the prognostic value of microRNA signature that is differentially expressed in patients with acute HF. Methods and Results Our study comprised a screening cohort of 15 patients with AHF and 5 controls, a PCR-discovery cohort of 50 patients with AHF and 26 controls and a validation cohort of 564 patients with AHF from registered study DRAGON-HF (Diagnostic, Risk Stratification and Prognostic Value of Novel Biomarkers in Patients With Heart Failure). Through screening by RNA-sequencing and verification by reverse-transcription quantitative polymerase chain reaction, 9 differentially expressed microRNAs were verified (miR-939-5p, miR-1908-5p, miR-7706, miR-101-3p, miR-144-3p, miR-4732-3p, miR-3615, miR-484 and miR-19b-3p). Among them, miR-19b-3p was identified as the microRNA signature with the highest fold-change of 8.4 and the strongest prognostic potential (area under curve with 95% CI, 0.791, 0.654-0.927). To further validate its prognostic value, in the validation cohort, the baseline level of miR-19b-3p was measured. During a follow-up period of 19.1 (17.7, 20.7) months, primary end point comprising of all-cause mortality or readmission due to HF occurred in 48.9% patients, while patients in the highest quartile of miR-19b-3p level presented the worst survival (Log-rank P<0.001). Multivariate Cox model showed that the level of miR-19b-3p could independently predict the occurrence of primary end point (adjusted hazard ratio,1.39; 95% CI, 1.18-1.64). In addition, miR-19b-3p positively correlated with soluble suppression of tumorigenicity 2 and echocardiographic indexes of left ventricular hypertrophy. Conclusions Circulating miR-19b-3p could be a valuable prognostic biomarker for AHF. In addition, a high level of circulating miR-19b-3p might indicate ventricular hypertrophy in AHF subjects. Registration URL: https://www.clinicaltrials.gov. Unique Identifier: NCT03727828.

Comprehensive Instructional Video Analysis: The COIN Dataset and Performance Evaluation.

Thanks to the substantial and explosively increased instructional videos on the Internet, novices are able to acquire knowledge for completing various tasks. Over the past decade, growing efforts have been devoted to investigating the problem on instructional video analysis. However, most existing instructional video datasets have limitations in diversity and scale, which makes them far from many real-world applications where more diverse activities occur. To address this, in this article, we propose a large-scale dataset called "COIN" for COmprehensive INstructional video analysis. Organized with a hierarchical structure, the COIN dataset contains 11,827 videos of 180 tasks in 12 domains (e.g., vehicles, gadgets, etc.) related to our daily life. With a new developed toolbox, all the videos are annotated efficiently with a series of step labels and the corresponding temporal boundaries. In order to provide a benchmark for instructional video analysis, we evaluate plenty of approaches on our COIN dataset under five different settings. Furthermore, we exploit two important characteristics (i.e., task-consistency and ordering-dependency) for localizing important steps in instructional videos. Accordingly, we propose two simple yet effective methods, which can be easily plugged into conventional proposal-based action detection models. We believe the introduction of the COIN dataset will promote the future in-depth research on instructional video analysis for the community. Our dataset, annotation toolbox and source codes are available at http://coin-dataset.github.io.

Learning Semantics-Preserving Attention and Contextual Interaction for Group Activity Recognition.

In this paper, we investigate the problem of group activity recognition by learning semantics-perserving attention and contextual interaction among different people. Conventional methods usually aggregate the features extracted from individual persons by pooling operations, which lack physical meaning and cannot fully explore the contextual information for group activity recognition. To address this, we develop a Semantics-Preserving Teacher-Student (SPTS) networks architecture. Our SPTS networks first learn a Teacher Network in the semantic domain that classifies the word of group activity based on the words of individual actions. Then we design a Student Network in the appearance domain that recognizes the group activity according to the input video. We enforce the Student Network to mimic the Teacher Network in the learning procedure. In this way, we allocate semantics-preserving attention to different people, which is more effective to seek the key people and discard the misleading people, while no extra labelled data are required. Moreover, a group of people inherently lie in a graphbased structure, where the people and their relationship can be regarded as the nodes and edges of a graph respectively. Based on this, we build two graph convolutional modules on both the Teacher Network and the Student Network to reason the dependency among different people. Furthermore, we extend our approach on action segmentation task based on its intermediate features. Experimental results on four datasets for group activity analysis clearly show the superior performance of our method in comparisons with the state-of-the-arts.