Biology subjects

Liao, C.

Publications and source records attributed to Liao, C..

At least 55 records · Page 3Linked to original sources

bioRxiv2024

Transcriptomic and epigenomic consequences of heterozygous loss of function mutations in AKAP11, the first large-effect shared risk gene for bipolar disorder and schizophrenia

The gene A-kinase anchoring protein 11 (AKAP11) recently emerged as a shared risk factor between bipolar disorder and schizophrenia, driven by large-effect loss-of-function (LoF) variants. Recent research has uncovered the neurophysiological characteristics and synapse proteomics profile of Akap11-mutant mouse models. Considering the role of AKAP11 in binding cAMP-dependent protein kinase A (PKA) and mediating phosphorylation of numerous substrates, such as transcription factors and epigenetic regulators, and given that chromatin alterations have been implicated in the brains of patients with bipolar disorder and schizophrenia, it is crucial to uncover the transcriptomic and chromatin dysregulations following the heterozygous knockout of AKAP11, particularly in human neurons. In this study, we use genome-wide approaches to investigate such aberrations in human induced pluripotent stem cell (iPSC)-derived neurons. We show the impact of heterozygous AKAP11 LoF mutations on the gene expression landscape and profile the methylomic and acetylomic modifications. Altogether we highlight the involvement of aberrant activity of intergenic and intronic enhancers, which are enriched in PBX homeobox 2 (PBX2) and Nuclear Factor-1 (NF1) known binding motifs, respectively, in transcription dysregulations of genes functioning as DNA-binding transcription factors, actin and cytoskeleton regulators, and cytokine receptors, as well as genes involved in G-protein-coupled receptors (GPCRs) binding and signaling. A better understanding of the dysregulations resulting from haploinsufficiency in AKAP11 improves our knowledge of the biological roots and pathophysiology of BD and SCZ, paving the way for better therapeutic approaches.

genomics↗

bioRxiv2024

Spherical Echo-Planar Time-resolved Imaging (sEPTI) for rapid 3D quantitative T2* and Susceptibility imaging

PurposeTo develop a 3D spherical EPTI (sEPTI) acquisition and a comprehensive reconstruction pipeline for rapid high-quality whole-brain submillimeter T2* and QSM quantification. MethodsFor the sEPTI acquisition, spherical k-space coverage is utilized with variable echo-spacing and maximum kx ramp-sampling to improve efficiency and incoherency when compared to existing EPTI approaches. For reconstruction, an iterative rank-shrinking B0 estimation and odd-even high-order phase correction algorithms were incorporated into the reconstruction to better mitigate artifacts from field imperfections. A physics-informed unrolled network was utilized to boost the SNR, where 1-mm and 0.75-mm isotropic whole-brain imaging were performed in 45 and 90 seconds, respectively. These protocols were validated through simulations, phantom, and in vivo experiments. Ten healthy subjects were recruited to provide sufficient data for the unrolled network. The entire pipeline was validated on additional 5 healthy subjects where different EPTI sampling approaches were compared. Two additional pediatric patients with epilepsy were recruited to demonstrate the generalizability of the unrolled reconstruction. ResultssEPTI achieved 1.4 x faster imaging with improved image quality and quantitative map precision compared to existing EPTI approaches. The B0 update and the phase correction provide improved reconstruction performance with lower artifacts. The unrolled network boosted the SNR, achieving high-quality T2* and QSM quantification with single average data. High-quality reconstruction was also obtained in the pediatric patient using this network. ConclusionsEPTI achieved whole-brain distortion-free multi-echo imaging and T2* and QSM quantification at 0.75 mm in 90 seconds which has the potential to be useful for wide clinical applications.

bioengineering↗

bioRxiv2024

Retrotransposon-mediated variation of a chitin synthase gene confers insect resistance to Bacillus thuringiensis Vip3Aa toxin

Bacillus thuringiensis (Bt) crops expressing Vip3Aa are highly efficacious in controlling major lepidopteran pests and delaying evolution of pest resistance. Although practical resistance to Vip3Aa in the field has not been reported, to proactively manage the pest resistance, there is an urgent need to better understand the genetic basis of resistance to Vip3Aa. This is particularly important for the fall armyworm (Spodoptera frugiperda), one of the most destructive pests around the world, which has evolved practical resistance to Bt crystal (Cry) toxins. Here, a highly Vip3Aa-resistant (resistance ratio: 5,562-fold) strain of S. frugiperda was selected in the laboratory. Results from bulked segregant analysis, fine-scale mapping, and genetic linkage analysis indicate that a mutation in the midgut-specific chitin synthase gene, SfCHS2, is strongly associated with high-level resistance to Vip3Aa. The resistance is ascribed to the transcriptional variation caused by retrotransposon insertion. The same variation of SfCHS2 was also detected in a field population. Importantly, knockout of SfCHS2 via CRISPR/Cas9 in susceptible S. frugiperda confers its complete resistance (>10,000-fold) to Vip3Aa. Also, we demonstrate that knockout of CHS2 can result in complete resistance to Vip3Aa in additional lepidopteran species, suggesting a general role of this gene in Vip3Aa resistance among lepidopteran pests. These results reported here would contribute to monitor and management of pest resistance to Vip3Aa.

genetics↗

bioRxiv2024

Probing the Link Between Vision and Language in Material Perception

We can visually discriminate and recognize a wide range of materials. Meanwhile, we use language to express our subjective understanding of visual input and communicate relevant information about the materials. Here, we investigate the relationship between visual judgment and language expression in material perception to understand how visual features relate to semantic representations. We use deep generative networks to construct an expandable image space to systematically create materials of well-defined and ambiguous categories. From such a space, we sampled diverse stimuli and compared the representations of materials from two behavioral tasks: visual material similarity judgments and free-form verbal descriptions. Our findings reveal a moderate but significant correlation between vision and language on a categorical level. However, analyzing the representations with an unsupervised alignment method, we discover structural differences that arise at the image-to-image level, especially among materials morphed between known categories. Moreover, visual judgments exhibit more individual differences compared to verbal descriptions. Our results show that while verbal descriptions capture material qualities on the coarse level, they may not fully convey the visual features that characterize the materials optical properties. Analyzing the image representation of materials obtained from various pre-trained data-rich deep neural networks, we find that human visual judgments similarity structures align more closely with those of the text-guided visual-semantic model than purely vision-based models. Our findings suggest that while semantic representations facilitate material categorization, non-semantic visual features also play a significant role in discriminating materials at a finer level. This work illustrates the need to consider the vision-language relationship in building a comprehensive model for material perception. Moreover, we propose a novel framework for quantitatively evaluating the alignment and misalignment between representations from different modalities, leveraging information from human behaviors and computational models.

neuroscience↗

bioRxiv2023

Reply to: Contribution of carbon inputs to soil carbon accumulation cannot be neglected

In the accompanying Comment1, He et al. argue that the determinant role of microbial carbon use efficiency in global soil organic carbon (SOC) storage shown in Tao et al. (2023)2 was overestimated because carbon inputs were neglected in our data analysis while they suggest that our model-based analysis could be biased and model-dependent. Their argument is based on a different choice of independent variables in the data analysis and a sensitivity analysis of two process-based models other than that used in our study. We agree that both carbon inputs and outputs (as mediated by microbial processes) matter when predicting SOC storage - the question is their relative contributions. While we encourage further studies to examine how the evaluation of the relative importance of CUE to global SOC storage may vary with different model structures, He et al.s claims about Tao et al. (2023) need to be taken as an alternative, unproven hypothesis until empirical data support their specific parameterization. Here we show that an additional literature assessment of global data does not support He et al.s argument, in contrast to our study, and that further study on this topic is essential.

ecology↗

bioRxiv2023

In situ structural insights into the excitation contraction coupling mechanism of skeletal muscle

Excitation-contraction coupling (ECC) is a fundamental mechanism in control of skeletal muscle contraction and occurs at triad junctions, where dihydropyridine receptors (DHPRs) on transverse tubules sense excitation signals and then cause calcium release from the sarcoplasmic reticulum via coupling to type 1 ryanodine receptors (RyR1s), inducing the subsequent contraction of muscle filaments. However, the molecular mechanism remains unclear due to the lack of structural details. Here, we explored the nanometre-resolution architecture of triad junction by cryo-electron tomography, solved the in situ structure of RyR1 in complex with FKBP12 and calmodulin, and discovered the intact RyR1-DHPR supercomplex. RyR1s arrange into two rows on the terminal cisternae membrane by forming right-hand corner-to-corner contacts, and tetrads of DHPRs bind to RyR1s in an alternating manner, forming another two rows on the transverse tubule membrane. Such unique arrangement is important for synergistic calcium release and provides direct evidence of physical coupling in ECC.

biophysics↗

bioRxiv2023

A CRISPR-based genetic screen in Bacteriodes thetaiotamicron reveals a small RNA modulator of bile susceptibility

Microbiota-centric interventions are limited by our incomplete understanding of the gene functions of many of its constituent species. This applies in particular to small RNAs (sRNAs), which are emerging as important regulators in microbiota species, yet tend to be missed by traditional functional genomics approaches. Here, we establish CRISPR interference (CRISPRi) in the abundant microbiota member Bacteroides thetaiotaomicron for genome-wide sRNA screens. By assessing the abundance of different protospacer-adjacent motifs, we identify the Prevotella bryantii B14 Cas12a as a suitable nuclease for CRISPR screens in these bacteria and generate an inducible Cas12a expression system. Using a luciferase reporter strain, we infer guide design rules and use this knowledge to assemble a computational pipeline for automated gRNA design. By subjecting the resulting guide library to a phenotypic screen, we uncover the previously uncharacterized sRNA BatR to increase susceptibility to bile salts, likely through the regulation of genes involved in Bacteroides cell surface structure. Our study lays the groundwork for unlocking the genetic potential of these major human gut mutualists and, more generally, for discovering hidden functions of bacterial sRNAs.

microbiology↗

bioRxiv2023

Development of Synthetic Modulator Enabling Long-Term Propagation and Neurogenesis of Human-Derived Neural Progenitor Cells

Neural progenitor cells (NPCs) are important cells for in vitro drug screening and the cell-based therapy for brain-related disorders, which requires well-defined and reproducible culture systems. Current strategy the use of protein growth factors presents challenges in terms of reproducibility and cost. In this study, we have developed a novel DNA-based modulator to regulate FGFR signaling of NPCs, enabling maintenance of the stemness over 50 passages and neurogenesis towards neurons. The DNA-based FGFR-agonist effectively promoted FGFR1 phosphorylation and activated the downstream ERK signaling pathway in FGFR1-positive cells. Using human embryonic stem cell lines, we differentiated them into NPCs and replaced basic fibroblast growth factor (bFGF) in the regulator culture medium with DNA-based FGFR-agonist for artificially elicited FGFR signaling. The results demonstrated that the FGFR-agonist could promote NPCs proliferation and neurosphere formation, recapitulating the function of bFGF. Notably, transcriptomic analysis revealed that FGFR-agonist could customize the stemness-associated transcription program, while decouples the neuronal differentiation program, highly resembling that the native ligand, bFGF. Moreover, our culture condition facilitated the successful propagation of NPCs for over 50 passages, while retaining their ability to efficiently differentiate into neurons. Overall, our approach provides a highly effective method for expanding NPCs, offering new opportunities for disease-in-dish research and drug screening for neural degeneration.

neuroscience↗

bioRxiv2023

Fitness and transcriptomic analysis of pathogenic Vibrio parahaemolyticus in seawater at different shellfish harvesting temperatures

To better characterize the population dynamics of Vibrio parahaemolyticus (Vp) containing different virulence genes, two Vp strains were inoculated into seawater separately and incubated at temperatures (30 and 10 {degrees}C) mimicking summer and winter pre-harvest shellfish rearing seasons. The cellular responses of these two strains, one containing the thermostable direct hemolysin (tdh+) gene and the other one containing tdh-related hemolysin (trh+) gene, were studied at the transcriptomic level. Results showed that, at 30 {degrees}C, tdh+ and trh+ strains reached 6.77 {+/-} 0.20 and 6.14 {+/-} 0.07 Log CFU/ml respectively after 5 days. During this time, higher growth rate was observed in the tdh+ strain than the trh+ strain. When being kept at 10 {degrees}C, both Vp strains persisted at ca. 3.0 Log CFU/ml in seawater with no difference observed between them. Growth and persistence predictive models were then established based on the Baranyi equation. The goodness of fit scores ranged from 0.674 to 0.950. RNA sequencing results showed that downregulated central energy metabolism and weakened degradation of branched chain amino acid were observed only in trh+ strain not in tdh+ strain at 30 {degrees}C. This might be one reason for the lower growth rates of the trh+ strain at 30 {degrees}C. Histidine metabolism and biofilm formation pathways were significantly downregulated in both strains at 10 {degrees}C. No significant difference was observed for virulence-associated gene expression between 10 and 30 {degrees}C, regardless of the strains. SIGINIFICANCEGiven the involvement of Vp in a wide range of seafood outbreaks, a systematical characterization of Vp fitness and transcriptomic changes at temperatures of critical importance for seafood production and storage is needed. In this study, predictive models describing the behavior of Vp strains containing different virulence factors are established. While no difference was observed at the lower temperature (10 C), tdh+ strain had faster growth rate than the trh+ strain. Transcriptomic analysis showed that significantly higher number of genes were upregulated at 30 {degrees}C than 10 {degrees}C. Majority of differentially expressed genes of Vp at 30 {degrees}C were annotated to functional categories supporting cellular growth. At the lower temperature, the down regulation of the biofilm formation pathway and histidine metabolism indicates that the current practice of storing seafood at lower temperatures not only protect the seafood quality but also ensure the seafood safety.

microbiology↗

bioRxiv2023

Deep Learning Initialized Compressed Sensing (Deli-CS) in Volumetric Spatio-Temporal Subspace Reconstruction

IntroductionSpatio-temporal MRI methods enable whole-brain multi-parametric mapping at ultra-fast acquisition times through efficient k-space encoding, but can have very long reconstruction times, which limit their integration into clinical practice. Deep learning (DL) is a promising approach to accelerate reconstruction, but can be computationally intensive to train and deploy due to the large dimensionality of spatio-temporal MRI. DL methods also need large training data sets and can produce results that dont match the acquired data if data consistency is not enforced. The aim of this project is to reduce reconstruction time using DL whilst simultaneously limiting the risk of deep learning induced hallucinations, all with modest hardware requirements. MethodsDeep Learning Initialized Compressed Sensing (Deli-CS) is proposed to reduce the reconstruction time of iterative reconstructions by "kick-starting" the iterative reconstruction with a DL generated starting point. The proposed framework is applied to volumetric multi-axis spiral projection MRF that achieves whole-brain T1 and T2 mapping at 1-mm isotropic resolution for a 2-minute acquisition. First, the traditional reconstruction is optimized from over two hours to less than 40 minutes while using more than 90% less RAM and only 4.7 GB GPU memory, by using a memory-efficient GPU implementation. The Deli-CS framework is then implemented and evaluated against the above reconstruction. ResultsDeli-CS achieves comparable reconstruction quality with 50% fewer iterations bringing the full reconstruction time to 20 minutes. ConclusionDeli-CS reduces the reconstruction time of subspace reconstruction of volumetric spatio-temporal acquisitions by providing a warm start to the iterative reconstruction algorithm.

neuroscience↗

bioRxiv2023

A novel lethal cuticular structural protein, AaCPR100A and its upstream interaction protein, G12-like, function in cuticle and egg shell formation in the yellow fever mosquito, Aedes aegypti

AaCPR100A is a structural protein, found in the soft cuticle of Ae. Aegypti. RNAi of AaCPR100A resulted in high mortality in Ae. Aegypti and abnormal egg development in the surviving mosquitoes. Over thirty proteins that could interact with AaCPR100A were screened out by yeast two-hybrid assay, and subsequently, further verification by hybrid and GST pull-down assays identified that G12-like had the strongest interaction with AaCPR100A. RNAi of G12-like suggested it may be related to larva development. Interestingly, the adults in which the G12-like gene was knocked down were sensitive to low temperature, and their egg shell formation, production, and hatching were affected. G12-like has the opposite effect in the upstream expression of AaCPR100A, promoting AaCPR100A function in the larval stage and inhibiting AaCPR100A in the adult stage. In all, functional studies of AaCPR100A and its interaction protein G12-like provide insight into its involvement in cuticle development and formation and egg shell formation.

molecular biology↗

bioRxiv2022

5-MeO-DMT modifies innate behaviors and promotes structural neural plasticity in mice

Serotonergic psychedelics are gaining increasing interest as potential therapeutics for a range of mental illnesses. Compounds with short-lived subjective effects may be clinically useful because dosing time would be reduced, which may improve patient access. One short-acting psychedelic is 5-MeO-DMT, which has been associated with improvement in depression and anxiety symptoms in early clinical studies. However relatively little is known about the behavioral effects and neural mechanisms of 5-MeO-DMT in animal models. Here we characterized the effects of 5-MeO-DMT on innate behaviors and dendritic architecture in mice. We showed that 5-MeO-DMT induces a dose-dependent increase in head-twitch response that is shorter in duration than that induced by psilocybin at all doses tested. 5-MeO-DMT also substantially suppresses social ultrasonic vocalizations produced during mating behavior. 5-MeO-DMT produces long-lasting increases in dendritic spine density in the mouse medial frontal cortex that are driven by an elevated rate of spine formation. However, unlike psilocybin, 5-MeO-DMT did not affect the size of dendritic spines. These data provide insights into the behavioral and neural consequences underlying the action of 5-MeO-DMT and highlight similarities and differences with those of psilocybin.

neuroscience↗

bioRxiv2022

Relative enrichment of oral bacteria in feces indicates loss of intestinal commensals with implications for host health

The increased relative abundance of oral bacteria detected in fecal samples has been associated with intestinal diseases and digestive disorders. This observation raises two competing hypotheses: either oral bacteria invade the gut bacterial population and expand in the intestine (the Expansion hypothesis), or oral bacteria transit through and their relative increase in feces marks a depletion of the gut bacterial population (the Marker hypothesis). To address this, we conducted a comprehensive analysis of quantitative microbiome data from mouse experiments and diverse patient cohorts. Our findings consistently support the Marker hypothesis as the primary explanation. We further establish a robust inverse correlation between the total fraction of oral bacteria and decreased total bacterial abundance in feces. This correlation underlies the associations between the oral bacterial fraction and multiple patient outcomes consistent with a depleted gut microbiota. By distinguishing between the two hypotheses, our study guides the interpretation of microbiome compositional data and their links with human health.

microbiology↗

bioRxiv2022

The TaxUMAP atlas: efficient display of large clinical microbiome data reveals ecological competition involved in protection against bacteremia

The microbiome is associated with health and disease, but causal effects are hard to quantify-- especially in humans where controlled experiments are nearly impossible. Akin to natural experiments, closely monitored patients offer an alternative to characterize microbiome effects. We present TaxUMAP, a taxonomically-informed visualization method to effectively display diverse microbiome states. TaxUMAP charts a microbiome atlas from 1,870 cancer patients as they progress through therapy-induced perturbations, and quantifies the microbiome contribution to patients risk for life-threatening bacteremia. We find that the lowest diversity states (gut dominations) that follow antibiotic treatments are stable, and that diverse communities harbor more diverse antimicrobial resistance genes than dominations. We reveal that certain Klebsiella species are associated with reduced risk for bacteremia, an effect driven by bacterial competition that we validate experimentally in vitro and in vivo. TaxUMAP effectively maps longitudinal microbiome data that can facilitate research into causal microbiome effects on human health. HIGHLIGHTSO_LITaxUMAP charts an atlas of patients microbiome states and their clinical context to reveal new causal effects. C_LIO_LIAntibiotics deplete the biodiversity and reduce the number of different antimicrobial resistance genes in the gut microbiome. C_LIO_LICertain Klebsiella species are associated with lower risk of bacteremia by other gut-borne pathogens. C_LIO_LIThese Klebsiella outcompete other gram-negative pathogens in vivo. C_LI

microbiology↗

bioRxiv2022

Alignment-free microbiome-based classification of fresh produce safety and quality

Small samples sizes and loss of up to 50-70% of sequencing data during the data denoising step of preprocessing can limit the statistical power of fresh produce microbiome analyses and prevent detection of important bacterial species associated with produce contamination or quality reduction. Here, we explored an alignment-free analysis strategy using k-mer hashes to identify DNA signatures predictive of produce safety and produce quality, and compared it against the amplicon sequence variant (ASV) strategy that uses a typical denoising step. Random forests (RF)-based classifiers for fresh produce safety and quality using 7-mer hash datasets had significantly higher classification accuracy than those using the ASV datasets. We also demonstrated that the proposed combination of integrating multiple datasets and leveraging an alignment-free 7-mer hash strategy leads to better classification performance for fresh produce safety and quality. Results generated from this study lay the foundation for future studies that wish and need to incorporate and/or compare different microbiome sequencing datasets for the application of machine learning in the area of microbial safety and quality of food.

microbiology↗

bioRxiv2022

Translucency perception emerges in deep generativerepresentations for natural image synthesis

Humans constantly assess the appearance of materials to plan actions, such as stepping on icy roads without slipping. Visual inference of materials is challenging because a given material can appear dramatically different in various scenes. This problem especially stands out for translucent materials (e.g., skin, soap), whose appearance strongly depends on lighting, geometry, and viewpoint. Despite this, humans can still distinguish between different materials, and it remains unsolved how to systematically discover visual features pertinent to material inference. Here, we developed an unsupervised style-based image generation model to identify perceptually relevant dimensions for material appearances from natural photographs. We find the model can synthesize images of convincing material appearances. Importantly, without supervision, human-understandable scene attributes, including objects shape, material, and body color, spontaneously emerge in the models layer-wise latent space in a scale-specific manner. Crucially, the middle-layers of the latent space selectively encode translucency features correlated with perception, suggesting that translucent impressions are established in mid-to-low spatial scale features. Our findings indicate that humans may use the scale-specific statistical structure of natural images to efficiently represent material properties across contexts, and our approach is widely applicable in discovering perceptually relevant features from complex stimuli for many visual inference tasks.

neuroscience↗

bioRxiv2022

Teasing out Missing Reactions in Genome-scale Metabolic Networks through Deep Learning

GEnome-scale Metabolic models (GEMs) are powerful tools to predict cellular metabolism and physiological states in living organisms. However, due to our imperfect knowledge of metabolic processes, even highly curated GEMs have knowledge gaps (e.g., missing reactions). Existing gap-filling methods typically require phenotypic data as input to tease out missing reactions. We still lack a computational method for rapid and accurate gap-filling of metabolic networks before experimental data is available. Here we present a deep learning-based method -- CHEbyshev Spectral HyperlInk pREdictor (CHESHIRE) -- to predict missing reactions in GEMs purely from metabolic network topology. We demonstrate that CHESHIRE outperforms other topology-based methods in predicting artificially removed reactions over 926 high- and intermediate-quality GEMs. Furthermore, CHESHIRE is able to improve the phenotypic predictions of 49 draft GEMs for fermentation products and amino acids secretions. Both types of validation suggest that CHESHIRE is a powerful tool for GEM curation to reveal unknown links between reactions and observed metabolic phenotypes.

bioinformatics↗

bioRxiv2022

Distinct Single-cell Immune Ecosystems Distinguish True and De Novo HBV-related Hepatocellular Carcinoma Recurrences

Revealing differential tumor immune microenvironment (TIME) characteristics between true versus de novo hepatocellular carcinoma (HCC) recurrence could help optimal development and use of immunotherapies. Here, we studied the TIME of recurrent HBV-related HCCs by 5and VDJ single-cell and bulk RNA-sequencing, flow cytometry, and multiplexed immunofluorescence. Analyses of mutational profiles, evolutionary trajectories, and clonal architecture using whole-exome sequencing identified de novo versus true recurrences, some of which occurred before clinical diagnosis. The TIME of truly recurrent HCCs was characterized by an increased abundance in KLRB1+CD8+ T cells with memory phenotype and low cytotoxicity. In contrast, we found an enrichment in cytotoxic and exhausted CD8+ T cells in the TIME of de novo recurrent HCCs. Transcriptomic and interaction analyses showed an upregulated GDF15 expression level on HCC cells in proximity to dendritic cells, which may have dampened antigen presentation and inhibited anti-tumor immunity in the TIME of truly recurrent lesions. In contrast, we found that myeloid cells crosstalk with T cells mediated T cell exhaustion and immunosuppression in the TIME of de novo recurrent HCC. In conclusion, our results support genomic diagnosis and immune profiling for guiding immunotherapy implementation based on the type of HCC recurrence and TIME. HighlightsO_LITruly recurrent lesions are seeded before primary tumor diagnosis, and that de novo cancer can occur earlier than the clinically used 2-year limit. C_LIO_LIScRNA-seq unravels distinct immune ecosystems in true versus de novo HCC recurrences, highlighting the need for different immunotherapy strategies for two types of HCC recurrence. C_LIO_LICD8+ T cells in de novo recurrence displayed cytotoxic and exhausted phenotypes while those in truly recurrent lesions showed a memory phenotype with weak cytotoxicity. C_LIO_LIHCC cells expressing the inhibitory molecule GDF15 were in the proximity of DCs only in truly recurrent lesions. C_LIO_LIHigh GDF15 expression level was associated with truly recurrent HCC and worse prognosis. C_LI Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=186 SRC="FIGDIR/small/494526v1_ufig1.gif" ALT="Figure 1"> View larger version (71K): org.highwire.dtl.DTLVardef@1b76bf0org.highwire.dtl.DTLVardef@10f889aorg.highwire.dtl.DTLVardef@d16b52org.highwire.dtl.DTLVardef@dceb7a_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗