Fast interpolation-based t-SNE for improved visualization of single-cell RNA-seq data

Simultaneous Protein Colorful Imaging via Raman Signal Classification

Protein imaging aids diagnosis and drug development by revealing protein-drug interactions or protein levels. However, the challenges of imaging multiple proteins, reduced sensitivity, and high reliance on specific protein properties such as Raman peaks or refractive index hinder the understanding. Here, we introduce multiprotein colorful imaging through Raman signal classification. Our method utilized machine learning-assisted classification of Raman signals, which are the distinctive features of label-free proteins. As a result, three types of proteins could be imaged simultaneously. In addition, we could quantify individual proteins from a mixture of multiple proteins over a wide detection range (10 fg/mL-1 μg/mL). These results showed a 1000-fold improvement in sensitivity and a 30-fold increase in the upper limit of detection compared to existing methods. These advances will enhance our understanding of biology and facilitate the development of disease diagnoses and treatments.

A comprehensive survey of dimensionality reduction and clustering methods for single-cell and spatial transcriptomics data

In recent years, the application of single-cell transcriptomics and spatial transcriptomics analysis techniques has become increasingly widespread. Whether dealing with single-cell transcriptomic or spatial transcriptomic data, dimensionality reduction and clustering are indispensable. Both single-cell and spatial transcriptomic data are often high-dimensional, making the analysis and visualization of such data challenging. Through dimensionality reduction, it becomes possible to visualize the data in a lower-dimensional space, allowing for the observation of relationships and differences between cell subpopulations. Clustering enables the grouping of similar cells into the same cluster, aiding in the identification of distinct cell subpopulations and revealing cellular diversity, providing guidance for downstream analyses. In this review, we systematically summarized the most widely recognized algorithms employed for the dimensionality reduction and clustering analysis of single-cell transcriptomic and spatial transcriptomic data. This endeavor provides valuable insights and ideas that can contribute to the development of novel tools in this rapidly evolving field.

Novel machine learning insights into the QM7b and QM9 quantum mechanics datasets

This paper (i) explores the internal structure of two quantum mechanics datasets (QM7b, QM9), composed of several thousands of organic molecules and described in terms of electronic properties, and (ii) further explores an inverse design approach to molecular design consisting of using machine learning methods to approximate the atomic composition of molecules, using QM9 data. Understanding the structure and characteristics of this kind of data is important when predicting the atomic composition from physical-chemical properties in inverse molecular designs. Intrinsic dimension analysis, clustering, and outlier detection methods were used in the study. They revealed that for both datasets the intrinsic dimensionality is several times smaller than the descriptive dimensions. The QM7b data is composed of well-defined clusters related to atomic composition. The QM9 data consists of an outer region predominantly composed of outliers, and an inner, core region that concentrates clustered inliner objects. A significant relationship exists between the number of atoms in the molecule and its outlier/inliner nature. The spatial structure exhibits a relationship with molecular weight. Despite the structural differences between the two datasets, the predictability of variables of interest for inverse molecular design is high. This is exemplified by models estimating the number of atoms of the molecule from both the original properties and from lower dimensional embedding spaces. In the generative approach the input is given by a set of desired properties of the molecule and the output is an approximation of the atomic composition in terms of its constituent chemical elements. This could serve as the starting region for further search in the huge space determined by the set of possible chemical compounds. The quantum mechanic's dataset QM9 is used in the study, composed of 133,885 small organic molecules and 19 electronic properties. Different multi-target regression approaches were considered for predicting the atomic composition from the properties, including feature engineering techniques in an auto-machine learning framework. High-quality models were found that predict the atomic composition of the molecules from their electronic properties, as well as from a subset of only 52.6% size. Feature selection worked better than feature generation. The results validate the generative approach to inverse molecular design.

Integration of metalloproteome and immunoproteome reveals a tight link of iron-related proteins with COVID-19 pathogenesis and immunity

Increasing clinical data show that the imbalance of host metallome is closely associated with different kinds of disease, however, the intrinsic mechanisms of action of metals in immunity and pathogenesis of disease remain largely undefined. There is lack of multiplexed profiling system to integrate the metalloproteome-immunoproteome information at systemic level for exploring the roles of metals in immunity and disease pathogenesis. In this study, we build up a metal-coding assisted multiplexed proteome assay platform for serum metalloproteomic and immunoproteomic profiling. By taking COVID-19 as a showcase, we unbiasedly uncovered the most evident modulation of iron-related proteins, i.e., Ft and Tf, in serum of severe COVID-19 patients, and the value of Ft/Tf could work as a robust biomarker for COVID-19 severity stratification, which overtakes the well-established clinical risk factors (cytokines). We further uncovered a tight association of transferrin with inflammation mediator IL-10 in COVID-19 patients, which was proved to be mainly governed by the monocyte/macrophage of liver, shedding light on new pathophysiological and immune regulatory mechanisms of COVID-19 disease. We finally validated the beneficial effects of iron chelators as anti-viral agents in SARS-CoV-2-infected K18-hACE2 mice through modulation of iron dyshomeostasis and alleviating inflammation response. Our findings highlight the critical role of liver-mediated iron dysregulation in COVID-19 disease severity, providing solid evidence on the involvement of iron-related proteins in COVID-19 pathophysiology and immunity.

Targeting pro-inflammatory T cells as a novel therapeutic approach to potentially resolve atherosclerosis in humans

Atherosclerosis (AS), a leading cause of cardio-cerebrovascular disease worldwide, is driven by the accumulation of lipid contents and chronic inflammation. Traditional strategies primarily focus on lipid reduction to control AS progression, leaving residual inflammatory risks for major adverse cardiovascular events (MACEs). While anti-inflammatory therapies targeting innate immunity have reduced MACEs, many patients continue to face significant risks. Another key component in AS progression is adaptive immunity, but its potential role in preventing AS remains unclear. To investigate this, we conducted a retrospective cohort study on tumor patients with AS plaques. We found that anti-programmed cell death protein 1 (PD-1) monoclonal antibody (mAb) significantly reduces AS plaque size. With multi-omics single-cell analyses, we comprehensively characterized AS plaque-specific PD-1+ T cells, which are activated and pro-inflammatory. We demonstrated that anti-PD-1 mAb, when captured by myeloid-expressed Fc gamma receptors (FcγRs), interacts with PD-1 expressed on T cells. This interaction turns the anti-PD-1 mAb into a substitute PD-1 ligand, suppressing T-cell functions in the PD-1 ligands-deficient context of AS plaques. Further, we conducted a prospective cohort study on tumor patients treated with anti-PD-1 mAb with or without Fc-binding capability. Our analysis shows that anti-PD-1 mAb with Fc-binding capability effectively reduces AS plaque size, while anti-PD-1 mAb without Fc-binding capability does not. Our work suggests that T cell-targeting immunotherapy can be an effective strategy to resolve AS in humans.

IL-27 promotes pathogenic T cells in a mouse model of Sjögren's disease

Sjögren's disease (SjD) is a chronic autoimmune disease characterized by focal lymphocytic inflammation in lacrimal and salivary glands. We recently identified IL-27 as a requisite signal for the spontaneous SjD-like manifestations in nonobese diabetic (NOD) mice. Here, we define T cell-intrinsic effects of IL-27 in lacrimal gland disease in NOD mice. IL-27 receptor was required by both CD4 T effector (Te) cells and CD8 T cells to mediate focal inflammation. Intrinsic IL-27 signaling was associated with PD-1 and ICOS expressing T follicular helper (Tfh)-like CD4 Te cells within lacrimal glands, including subsets defined by CD73 or CD39 expression. CD8 T cells capable of IL-27 signaling also expressed PD-1 with subsets expressing ICOS and CD73 demonstrating a T follicular cytotoxic (Tfc)-like cell phenotype and others expressing a CD39hi exhausted-like phenotype. These findings suggest IL-27 is a key early signal driving a follicular-type response in lacrimal gland inflammation in NOD mice.

GNNGL-PPI: multi-category prediction of protein-protein interactions using graph neural networks based on global graphs and local subgraphs

Most proteins exert their functions by interacting with other proteins, making the identification of protein-protein interactions (PPI) crucial for understanding biological activities, pathological mechanisms, and clinical therapies. Developing effective and reliable computational methods for predicting PPI can significantly reduce the time-consuming and labor-intensive associated traditional biological experiments. However, accurately identifying the specific categories of protein-protein interactions and improving the prediction accuracy of the computational methods remain dual challenges. To tackle these challenges, we proposed a novel graph neural network method called GNNGL-PPI for multi-category prediction of PPI based on global graphs and local subgraphs. GNNGL-PPI consisted of two main components: using Graph Isomorphism Network (GIN) to extract global graph features from PPI network graph, and employing GIN As Kernel (GIN-AK) to extract local subgraph features from the subgraphs of protein vertices. Additionally, considering the imbalanced distribution of samples in each category within the benchmark datasets, we introduced an Asymmetric Loss (ASL) function to further enhance the predictive performance of the method. Through evaluations on six benchmark test sets formed by three different dataset partitioning algorithms (Random, BFS, DFS), GNNGL-PPI outperformed the state-of-the-art multi-category prediction methods of PPI, as measured by the comprehensive performance evaluation metric F1-measure. Furthermore, interpretability analysis confirmed the effectiveness of GNNGL-PPI as a reliable multi-category prediction method for predicting protein-protein interactions.

A review of single-cell transcriptomics and epigenomics studies in maternal and child health

Single-cell sequencing technologies enhance our understanding of cellular dynamics throughout pregnancy. We outlined the workflow of single-cell sequencing techniques and reviewed single-cell studies in maternal and child health. We conducted a literature review of single cell studies on maternal and child health using PubMed. We summarized the findings from 16 single-cell atlases of the human and mammalian placenta across gestational stages and 31 single-cell studies on maternal exposures and complications including infection, obesity, diet, gestational diabetes, pre-eclampsia, environmental exposure and preterm birth. Single-cell studies provides insights on novel cell types in placenta and cell type-specific marks associated with maternal exposures and complications.

A novel approach for credit card fraud transaction detection using deep reinforcement learning scheme

Online transactions are still the backbone of the financial industry worldwide today. Millions of consumers use credit cards for their daily transactions, which has led to an exponential rise in credit card fraud. Over time, many variations and schemes of fraudulent transactions have been reported. Nevertheless, it remains a difficult task to detect credit card fraud in real-time. It can be assumed that each person has a unique transaction pattern that may change over time. The work in this article aims to (1) understand how deep reinforcement learning can play an important role in detecting credit card fraud with changing human patterns, and (2) develop a solution architecture for real-time fraud detection. Our proposed model utilizes the Deep Q network for real-time detection. The Kaggle dataset available online was used to train and test the model. As a result, a validation performance of 97.10% was achieved with the proposed deep learning component. In addition, the reinforcement learning component has a learning rate of 80%. The proposed model was able to learn patterns autonomously based on previous events. It adapts to the pattern changes over time and can take them into account without further manual training.

Immunophenotypic predictors of influenza vaccine immunogenicity in pediatric hematopoietic cell transplant recipients

ABSTRACT

Pediatric hematopoietic cell transplant (HCT) recipients exhibit poor serologic responses to influenza vaccination early after transplant. To facilitate the optimization of influenza vaccination timing, we sought to identify B- and T-cell subpopulations associated with influenza vaccine immunogenicity in this population. We used mass cytometry to phenotype peripheral blood mononuclear cells collected from pediatric HCT recipients enrolled in a multicenter influenza vaccine trial comparing high- and standard-dose formulations over 3 influenza seasons (2016-2019). We fit linear regression models to estimate relationships between immune cell subpopulation numbers before vaccination and prevaccination to postvaccination geometric mean fold rises in antigen-specific (A/H3N2, A/H1N1, and B/Victoria) serum hemagglutination inhibition antibody titers (28-42 days, and ∼6 months after 2 doses). For cell subpopulations identified as predictive of a response to all 3 antigens, we conducted a sensitivity analysis including time after transplant as an additional covariate. Among 156 HCT recipients, we identified 33 distinct immune cell subpopulations; 7 significantly predicted responses to all 3 antigens 28 to 42 days after a 2-dose vaccine series, irrespective of vaccine dose. We also found evidence that baseline absolute numbers of naïve B cells, naïve CD4+ T cells, and circulating T follicular helper cells predicted peak and sustained vaccine-induced titers irrespective of dose or timing of posttransplant vaccine administration. In conclusion, several B- and T-cell subpopulations predicted influenza vaccine immunogenicity in pediatric HCT recipients. This study provides insights into the immune determinants of vaccine responses and may help guide the development of tailored vaccination strategies for this vulnerable population.

Evaluating cell culture reliability in pediatric brain tumor primary cells through DNA methylation profiling

In vitro models of pediatric brain tumors (pBT) are instrumental for better understanding the mechanisms contributing to oncogenesis and testing new therapies; thus, ideally, they should recapitulate the original tumor. We applied DNA methylation (DNAm) and copy number variation (CNV) profiling to characterize 241 pBT samples, including 155 tumors and 86 pBT-derived cell cultures, considering serum vs serum-free conditions, late vs early passages, and dimensionality (2D vs 3D cultures). We performed a t-SNE classification and identified differentially methylated regions in tumors compared to cell models. Early cell cultures recapitulate the original tumor, but serum media and 2D culturing were demonstrated to significantly contribute to the divergence of DNAm profiles from the parental ones. All divergent cells clustered together acquiring a common deregulated epigenetic signature suggesting a shared selective pressure. We identified a set of hypomethylated genes shared among unfaithful cells converging on response to growth factors and migration pathways, such as signaling cascade activation, tissue organization, and cellular migration. In conclusion, DNAm and CNV are informative tools that should be used to assess the recapitulation of pBT-cells from parental tumors.

The implications of single-cell RNA-seq analysis in prostate cancer: unraveling tumor heterogeneity, therapeutic implications and pathways towards personalized therapy

In recent years, advancements in single-cell and spatial transcriptomics, which are highly regarded developments in the current era, particularly the emerging integration of single-cell and spatiotemporal transcriptomics, have enabled a detailed molecular comprehension of the complex regulation of cell fate. The insights obtained from these methodologies are anticipated to significantly contribute to the development of personalized medicine. Currently, single-cell technology is less frequently utilized for prostate cancer compared with other types of tumors. Starting from the perspective of RNA sequencing technology, this review outlined the significance of single-cell RNA sequencing (scRNA-seq) in prostate cancer research, encompassing preclinical medicine and clinical applications. We summarize the differences between mouse and human prostate cancer as revealed by scRNA-seq studies, as well as a combination of multi-omics methods involving scRNA-seq to highlight the key molecular targets for the diagnosis, treatment, and drug resistance characteristics of prostate cancer. These studies are expected to provide novel insights for the development of immunotherapy and other innovative treatment strategies for castration-resistant prostate cancer. Furthermore, we explore the potential clinical applications stemming from other single-cell technologies in this review, paving the way for future research in precision medicine.

Graph Convolutional Network-Enhanced Model for Screening Persistent, Mobile, and Toxic and Very Persistent and Very Mobile Substances

The global management for persistent, mobile, and toxic (PMT) and very persistent and very mobile (vPvM) substances has been further strengthened with the rapid increase of emerging contaminants. The development of a ready-to-use and publicly available tool for the high-throughput screening of PMT/vPvM substances is thus urgently needed. However, the current model building with the coupling of conventional algorithms, small-scale data set, and simplistic features hinders the development of a robust model for screening PMT/vPvM with wide application domains. Here, we construct a graph convolutional network (GCN)-enhanced model with feature fusion of a molecular graph and molecular descriptors to effectively utilize the significant correlation between critical descriptors and PMT/vPvM substances. The model is built with 213,084 substances following the latest PMT classification criteria. The application domains of the GCN-enhanced model assessed by kernel density estimation demonstrate the high suitability for high-throughput screening PMT/vPvM substances with both a high accuracy rate (86.6%) and a low false-negative rate (6.8%). An online server named PMT/vPvM profiler is further developed with a user-friendly web interface (http://www.pmt.zj.cn/). Our study facilitates a more efficient evaluation of PMT/vPvM substances with a globally accessible screening platform.

SuperCellCyto: enabling efficient analysis of large scale cytometry datasets

Advancements in cytometry technologies have enabled quantification of up to 50 proteins across millions of cells at single cell resolution. Analysis of cytometry data routinely involves tasks such as data integration, clustering, and dimensionality reduction. While numerous tools exist, many require extensive run times when processing large cytometry data containing millions of cells. Existing solutions, such as random subsampling, are inadequate as they risk excluding rare cell subsets. To address this, we propose SuperCellCyto, an R package that builds on the SuperCell tool which groups highly similar cells into supercells. SuperCellCyto is available on GitHub ( https://github.com/phipsonlab/SuperCellCyto ) and Zenodo ( https://doi.org/10.5281/zenodo.10521294 ).

Single-cell RNA-seq revealed heterogeneous responses and functional differentiation of hemocytes against white spot syndrome virus infection in Litopenaeus vannamei

Shrimp hemocytes are the vital immune cells participating in innate immune response to defend against viruses. However, the lack of specific molecular markers for shrimp hemocyte hindered the insightful understanding of their functional clusters and differential roles in combating microbial infections. In this study, we used single-cell RNA sequencing to map the transcriptomic landscape of hemocytes from the white spot syndrome virus (WSSV)-infected Litopenaeus vannamei and conjointly analyzed with our previous published single-cell RNA sequencing technology data from the healthy hemocytes. A total of 16 transcriptionally distinct cell clusters were identified, which occupied different proportions in healthy and WSSV-infected hemocytes and exerted differential roles in antiviral immune response. Following mapping of the sequencing data to the WSSV genome, we found that all types of hemocytes could be invaded by WSSV virions, especially the cluster 8, which showed the highest transcriptional levels of WSSV genes and exhibited a cell type-specific antiviral response to the viral infection. Further evaluation of the cell clusters revealed the delicate dynamic balance between hemocyte immune response and viral infestation. Unsupervised pseudo-time analysis of hemocytes showed that the hemocytes in immune-resting state could be significantly activated upon WSSV infection and then functionally differentiated to different hemocyte subsets. Collectively, our results revealed the differential responses of shrimp hemocytes and the process of immune-functional differentiation post-WSSV infection, providing essential resource for the systematic insight into the synergistic immune response mechanism against viral infection among hemocyte subtypes.

IMPORTANCE

Current knowledge of shrimp hemocyte classification mainly comes from morphology, which hinder in-depth characterization of cell lineage development, functional differentiation, and different immune response of hemocyte types during pathogenic infections. Here, single-cell RNA sequencing was used for mapping hemocytes during white spot syndrome virus (WSSV) infection in Litopenaeus vannamei, identifying 16 cell clusters and evaluating their potential antiviral functional characteristics. We have described the dynamic balance between viral infestation and hemocyte immunity. And the functional differentiation of hemocytes under WSSV stimulation was further characterized. Our results provided a comprehensive transcriptional landscape and revealed the heterogeneous immune response in shrimp hemocytes during WSSV infection.

Sirpα on tumor-associated myeloid cells restrains antitumor immunity in colorectal cancer independent of its interaction with CD47

Immunosuppressive myeloid cells hinder immunotherapeutic efficacy in tumors, but the precise mechanisms remain undefined. Here, by performing single-cell RNA sequencing in colorectal cancer tissues, we found tumor-associated macrophages and granulocytic myeloid-derived suppressor cells increased most compared to their counterparts in normal tissue and displayed the highest immune-inhibitory signatures among all immunocytes. These cells exhibited significantly increased expression of immunoreceptor tyrosine-based inhibitory motif-bearing receptors, including SIRPA. Notably, Sirpa-/- mice were more resistant to tumor progression than wild-type mice. Moreover, Sirpα deficiency reprogramed the tumor microenvironment through expansion of TAM_Ccl8hi and gMDSC_H2-Q10hi subsets showing strong antitumor activity. Sirpa-/- macrophages presented strong phagocytosis and antigen presentation to enhance T cell activation and proliferation. Furthermore, Sirpa-/- macrophages facilitated T cell recruitment via Syk/Btk-dependent Ccl8 secretion. Therefore, Sirpα deficiency enhances innate and adaptive immune activation independent of expression of CD47 and Sirpα blockade could be a promising strategy to improve cancer immunotherapy efficacy.

HM_ADET: a hybrid model for automatic detection of eyelid tumors based on photographic images

BACKGROUND

The accurate detection of eyelid tumors is essential for effective treatment, but it can be challenging due to small and unevenly distributed lesions surrounded by irrelevant noise. Moreover, early symptoms of eyelid tumors are atypical, and some categories of eyelid tumors exhibit similar color and texture features, making it difficult to distinguish between benign and malignant eyelid tumors, particularly for ophthalmologists with limited clinical experience.

METHODS

We propose a hybrid model, HM_ADET, for automatic detection of eyelid tumors, including YOLOv7_CNFG to locate eyelid tumors and vision transformer (ViT) to classify benign and malignant eyelid tumors. First, the ConvNeXt module with an inverted bottleneck layer in the backbone of YOLOv7_CNFG is employed to prevent information loss of small eyelid tumors. Then, the flexible rectified linear unit (FReLU) is applied to capture multi-scale features such as texture, edge, and shape, thereby improving the localization accuracy of eyelid tumors. In addition, considering the geometric center and area difference between the predicted box (PB) and the ground truth box (GT), the GIoU_loss was utilized to handle cases of eyelid tumors with varying shapes and irregular boundaries. Finally, the multi-head attention (MHA) module is applied in ViT to extract discriminative features of eyelid tumors for benign and malignant classification.

RESULTS

Experimental results demonstrate that the HM_ADET model achieves excellent performance in the detection of eyelid tumors. In specific, YOLOv7_CNFG outperforms YOLOv7, with AP increasing from 0.763 to 0.893 on the internal test set and from 0.647 to 0.765 on the external test set. ViT achieves AUCs of 0.945 (95% CI 0.894-0.981) and 0.915 (95% CI 0.860-0.955) for the classification of benign and malignant tumors on the internal and external test sets, respectively.

CONCLUSIONS

Our study provides a promising strategy for the automatic diagnosis of eyelid tumors, which could potentially improve patient outcomes and reduce healthcare costs.

Statistical method scDEED for detecting dubious 2D single-cell embeddings and optimizing t-SNE and UMAP hyperparameters

Two-dimensional (2D) embedding methods are crucial for single-cell data visualization. Popular methods such as t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP) are commonly used for visualizing cell clusters; however, it is well known that t-SNE and UMAP's 2D embeddings might not reliably inform the similarities among cell clusters. Motivated by this challenge, we present a statistical method, scDEED, for detecting dubious cell embeddings output by a 2D-embedding method. By calculating a reliability score for every cell embedding based on the similarity between the cell's 2D-embedding neighbors and pre-embedding neighbors, scDEED identifies the cell embeddings with low reliability scores as dubious and those with high reliability scores as trustworthy. Moreover, by minimizing the number of dubious cell embeddings, scDEED provides intuitive guidance for optimizing the hyperparameters of an embedding method. We show the effectiveness of scDEED on multiple datasets for detecting dubious cell embeddings and optimizing the hyperparameters of t-SNE and UMAP.

Comparative analysis of human gut- and blood-derived mononuclear cells: contrasts in function and phenotype

Introduction

Alterations in the gut immune system have been implicated in various diseases.The challenge of obtaining gut tissues from healthy individuals, commonly performed via surgical explants, has limited the number of studies describing the phenotype and function of gut-derived immune cells in health.

Methods

Here, by means of recto-sigmoid colon biopsies obtained during routine care (colon cancer screening in healthy adults), the phenotype and function of immune cells present in the gut were described and compared to those found in blood.

Results

The proportion of CD4+, CD8+, MAIT, γδ+ T, and NK cells phenotype, expression of integrins, and ability to produce cytokine in response to stimulation with PMA and ionomycin. T cells in the gut were found to predominantly have a memory phenotype as compared to T cells in blood where a naïve phenotype predominates. Recto-sigmoid mononuclear cells also had higher PD-1 and Ki67 expression. Furthermore, integrin expression and cytokine production varied by cell type and location in blood vs. gut.

Discussion

These findings demonstrate the differences in functionality of these cells when compared to their blood counterparts and validate previous studies on phenotype within gut-derived immune cells in humans (where cells have been obtained through surgical means). This study suggests that recto-sigmoid biopsies collected during colonoscopy can be a reliable yet more accessible sampling method for follow up of alterations of gut derived immune cells in clinical settings.

Tumor-targeted nanodrug FSGG/siGal-9 for transdermal photothermal immunotherapy of melanoma

Photothermal therapy (PTT) is a cancer-targeted treatment approach.The occurrence of tumors may be related to microbial infections (Viruses, bacteria, fungi, etc.), which probably provokes anti-tumor immunity. However, T cells in the context of cancer become exhausted and dysfunctional. Galectin-9 (Gal-9) is highly expressed in normal tissues and associates with body immune tolerance, and was firstly evidenced with much higher expression on the primary solid tumors than CD80/86 (B7) and CD274 (PD-L1) here, which suggests that Gal-9 may be a key factor in inhibiting the anti-tumor immunity, and its receptor T cell immunoglobulin and mucin domain 3 (TIM-3) was discovered on the cytotoxic T lymphocytes (CTL) with high expression as well based on the single cell analysis. The immune checkpoint communications showed that the Gal-9/TIM-3 axis played the most vital role on negatively regulating the anti-tumor immunity of CTL for melanoma. Then, we used a novel transdermal photothermal nanosensitizer (FSGG) loading Gal-9 siRNA (FSGG/siGal-9) for knocking the tumor cells down Gal-9 to block the Gal-9/TIM-3 axis and prohibit CTL exhaustion synergizing PTT against melanoma, which evidenced good effects on inhibiting tumor growth and enhancing anti-tumor immunity, named "photothermal immunotherapy". This paper provides a new perspective for tumor prevention and treatment.

Isoform-selective TGF-β3 inhibition for systemic sclerosis

BACKGROUND

Transforming growth factor β (TGF-β) is implicated as a key mediator of pathological fibrosis, but its pleiotropic activity in a range of homeostatic functions presents challenges to its safe and effective therapeutic targeting. There are three isoforms of TGF-β, TGF-β1, TGF-β2, and TGF-β3, which bind to a common receptor complex composed of TGF-βR1 and TGF-βR2 to induce similar intracellular signals in vitro. We have recently shown that the cellular expression patterns and activation thresholds of TGF-β2 and TGF-β3 are distinct from those of TGF-β1 and that selective short-term TGF-β2 and TGF-β3 inhibition can attenuate fibrosis in vivo without promoting excessive inflammation. Isoform-selective inhibition of TGF-β may therefore provide a therapeutic opportunity for patients with chronic fibrotic disorders.

METHODS

Transcriptomic profiling of skin biopsies from patients with systemic sclerosis (SSc) from multiple clinical trials was performed to evaluate the role of TGF-β3 in this disease. Antibody humanization, biochemical characterization, crystallization, and pre-clinical experiments were performed to further characterize an anti-TGF-β3 antibody.

FINDINGS

In the skin of patients with SSc, TGF-β3 expression is uniquely correlated with biomarkers of TGF-β signaling and disease severity. Crystallographic studies establish a structural basis for selective TGF-β3 inhibition with a potent and selective monoclonal antibody that attenuates fibrosis effectively in vivo at clinically translatable exposures. Toxicology studies suggest that, as opposed to pan-TGF-β inhibitors, this anti-TGF-β3 antibody has a favorable safety profile for chronic administration.

CONCLUSION

We establish a rationale for targeting TGF-β3 in SSc with a favorable therapeutic index.

FUNDING

This study was funded by Genentech, Inc.

Single-Nucleus Analysis Reveals Tumor Heterogeneity of Aldosterone-Producing Adenoma

BACKGROUND

Recent advances in omics techniques have allowed detailed genetic characterization of aldosterone-producing adenoma (APA). The pathogenesis of APA is characterized by tumorigenesis-associated aldosterone synthesis. The pathophysiological intricacies of APAs have not yet been elucidated at the level of individual cells. Therefore, a single-cell level analysis is speculated to be valuable in studying the differentiation process of APA.

METHODS

We conducted single-nucleus RNA sequencing of APAs with KCNJ5 mutation and nonfunctional adenomas obtained from 3 and 2 patients, respectively.

RESULTS

The single-nucleus RNA sequencing revealed the intratumoral heterogeneity of APA and identified cell populations consisting of a shared cluster of nonfunctional adenoma and APA. In addition, we extracted 2 cell fates in APA and obtained a cell population specialized in aldosterone synthesis. Genes related to ribosomes and neurodegenerative diseases were upregulated in 1 of these fates, whereas those related to the regulation of glycolysis were upregulated in the other fate. Furthermore, the total RNA reads in the nucleus were higher in hormonally activated clusters, indicating a marked activation of transcription per cell.

CONCLUSIONS

The single-nucleus RNA sequencing revealed intratumoral heterogeneity of APA with KCNJ5 mutation. The observation of 2 cell fates in KCNJ5-mutated APAs provides the postulation that a heterogeneous process of cellular differentiation was implicated in the pathophysiological mechanisms underlying APA tumors.

Single-Cell Transcriptome Profiling of Pancreatic Islets From Early Diabetic Mice Identifies Anxa10 for Ca2+ Allostasis Toward β-Cell Failure

Type 2 diabetes is a progressive disorder denoted by hyperglycemia and impaired insulin secretion. Although a decrease in β-cell function and mass is a well-known trigger for diabetes, the comprehensive mechanism is still unidentified. Here, we performed single-cell RNA sequencing of pancreatic islets from prediabetic and diabetic db/db mice, an animal model of type 2 diabetes. We discovered a diabetes-specific transcriptome landscape of endocrine and nonendocrine cell types with subpopulations of β- and α-cells. We recognized a new prediabetic gene, Anxa10, that was induced by and regulated Ca2+ influx from metabolic stresses. Anxa10-overexpressed β-cells displayed suppression of glucose-stimulated intracellular Ca2+ elevation and potassium-induced insulin secretion. Pseudotime analysis of β-cells predicted that this Ca2+-surge responder cluster would proceed to mitochondria dysfunction and endoplasmic reticulum stress. Other trajectories comprised dedifferentiation and transdifferentiation, emphasizing acinar-like cells in diabetic islets. Altogether, our data provide a new insight into Ca2+ allostasis and β-cell failure processes.

ARTICLE HIGHLIGHTS

The transcriptome of single-islet cells from healthy, prediabetic, and diabetic mice was studied. Distinct β-cell heterogeneity and islet cell-cell network in prediabetes and diabetes were found. A new prediabetic β-cell marker, Anxa10, regulates intracellular Ca2+ and insulin secretion. Diabetes triggers β-cell to acinar cell transdifferentiation.

GHGPR-PPIS: A graph convolutional network for identifying protein-protein interaction site using heat kernel with Generalized PageRank techniques and edge self-attention feature processing block

Accurately pinpointing protein-protein interaction site (PPIS) on the molecular level is of utmost significance for annotating protein function and comprehending the mechanisms underpinning various diseases. While numerous computational methods for predicting PPIS have emerged, they have indeed mitigated the labor and time constraints associated with traditional experimental methods. However, the predictive accuracy of these methods has yet to reach the desired threshold. In this context, we proposed a groundbreaking graph-based computational model called GHGPR-PPIS. This innovative model leveraged a graph convolutional network using heat kernel (GraphHeat) in conjunction with Generalized PageRank techniques (GHGPR) to predict PPIS. Additionally, building upon the GHGPR framework, we devised an edge self-attention feature processing block, further augmenting the performance of the model. Experimental findings conclusively demonstrated that GHGPR-PPIS surpassed all competing state-of-the-art models when evaluated on the benchmark test set. Impressively, on two distinct independent test sets and a specific protein chain, GHGPR-PPIS consistently demonstrated superior generalization performance and practical applicability compared to the comparative model, AGAT-PPIS. Lastly, leveraging the t-SNE dimensionality reduction algorithm and clustering visualization technique, we delved into an interpretability analysis of the effectiveness of GHGPR-PPIS by meticulously comparing the outputs from different stages of the model.

OPTIMAL: An OPTimized Imaging Mass cytometry AnaLysis framework for benchmarking segmentation and data exploration

Analysis of imaging mass cytometry (IMC) data and other low-resolution multiplexed tissue imaging technologies is often confounded by poor single-cell segmentation and suboptimal approaches for data visualization and exploration. This can lead to inaccurate identification of cell phenotypes, states, or spatial relationships compared to reference data from single-cell suspension technologies. To this end we have developed the "OPTimized Imaging Mass cytometry AnaLysis (OPTIMAL)" framework to benchmark any approaches for cell segmentation, parameter transformation, batch effect correction, data visualization/clustering, and spatial neighborhood analysis. Using a panel of 27 metal-tagged antibodies recognizing well-characterized phenotypic and functional markers to stain the same Formalin-Fixed Paraffin Embedded (FFPE) human tonsil sample tissue microarray over 12 temporally distinct batches we tested several cell segmentation models, a range of different arcsinh cofactor parameter transformation values, 5 different dimensionality reduction algorithms, and 2 clustering methods. Finally, we assessed the optimal approach for performing neighborhood analysis. We found that single-cell segmentation was improved by the use of an Ilastik-derived probability map but that issues with poor segmentation were only really evident after clustering and cell type/state identification and not always evident when using "classical" bivariate data display techniques. The optimal arcsinh cofactor for parameter transformation was 1 as it maximized the statistical separation between negative and positive signal distributions and a simple Z-score normalization step after arcsinh transformation eliminated batch effects. Of the five different dimensionality reduction approaches tested, PacMap gave the best data structure with FLOWSOM clustering out-performing phenograph in terms of cell type identification. We also found that neighborhood analysis was influenced by the method used for finding neighboring cells with a "disc" pixel expansion outperforming a "bounding box" approach combined with the need for filtering objects based on size and image-edge location. Importantly, OPTIMAL can be used to assess and integrate with any existing approach to IMC data analysis and, as it creates .FCS files from the segmentation output and allows for single-cell exploration to be conducted using a wide variety of accessible software and algorithms familiar to conventional flow cytometrists.

Single-cell characterization of human GBM reveals regional differences in tumor-infiltrating leukocyte activation

Glioblastoma (GBM) harbors a highly immunosuppressive tumor microenvironment (TME) which influences glioma growth. Major efforts have been undertaken to describe the TME on a single-cell level. However, human data on regional differences within the TME remain scarce. Here, we performed high-depth single-cell RNA sequencing (scRNAseq) on paired biopsies from the tumor center, peripheral infiltration zone and blood of five primary GBM patients. Through analysis of >45,000 cells, we revealed a regionally distinct transcription profile of microglia (MG) and monocyte-derived macrophages (MdMs) and an impaired activation signature in the tumor-peripheral cytotoxic-cell compartment. Comparing tumor-infiltrating CD8+ T cells with circulating cells identified CX3CR1high and CX3CR1int CD8+ T cells with effector and memory phenotype, respectively, enriched in blood but absent in the TME. Tumor CD8+ T cells displayed a tissue-resident memory phenotype with dysfunctional features. Our analysis provides a regionally resolved mapping of transcriptional states in GBM-associated leukocytes, serving as an additional asset in the effort towards novel therapeutic strategies to combat this fatal disease.

Nicotinamide mononucleotide impacts HIV-1 infection by modulating immune activation in T lymphocytes and humanized mice

BACKGROUND

HIV-1-associated immune activation drives CD4+ T cell depletion and the development of acquired immunodeficiency syndrome. We aimed to determine the role of nicotinamide mononucleotide (NMN), the direct precursor of nicotinamide adenine dinucleotide (NAD) co-enzyme, in CD4+ T cell modulation during HIV-1 infection.

METHODS

We examined HIV-1 integrated DNA or transcribed RNA, intracellular p24 protein, and T cell activation markers in CD4+ T cells including in vitro HIV-1-infected cells, reactivated patient-derived cells, and in HIV-1-infected humanized mice, under NMN treatment. RNA-seq and CyTOF analyses were used for investigating the effect of NMN on CD4+ T cells.

FINDINGS

We found that NMN increased the intracellular NAD amount, resulting in suppressed HIV-1 p24 production and proliferation in infected CD4+ T cells, especially in activated CD25+CD4+ T cells. NMN also inhibited CD25 expression on reactivated resting CD4+ T cells derived from cART-treated people living with HIV-1 (PLWH). In HIV-1-infected humanized mice, the frequency of CD4+ T cells was reconstituted significantly by combined cART and NMN treatment as compared with cART or NMN alone, which correlated with suppressed hyperactivation of CD4+ T cells.

INTERPRETATION

Our results highlight the suppressive role of NMN in CD4+ T cell activation during HIV-1 infection. It warrants future clinical investigation of NMN as a potential treatment in combination with cART in PLWH.

FUNDING

This work was supported by the Hong Kong Research Grants Council Theme-Based Research Scheme (T11-706/18-N), University Research Committee of The University of Hong Kong, the Collaborative Research with GeneHarbor (Hong Kong) Biotechnologies Limited and National Key R&D Program of China (Grant2021YFC2301900).

Vascular senescence and leak are features of the early breakdown of the blood-brain barrier in Alzheimer's disease models

Alzheimer's disease (AD) is an age-related disease, with loss of integrity of the blood-brain barrier (BBB) being an early feature. Cellular senescence is one of the reported nine hallmarks of aging. Here, we show for the first time the presence of senescent cells in the vasculature in AD patients and mouse models of AD. Senescent endothelial cells and pericytes are present in APP/PS1 transgenic mice but not in wild-type littermates at the time of amyloid deposition. In vitro, senescent endothelial cells display altered VE-cadherin expression and loss of cell junction formation and increased permeability. Consistent with this, senescent endothelial cells in APP/PS1 mice are present at areas of vascular leak that have decreased claudin-5 and VE-cadherin expression confirming BBB breakdown. Furthermore, single cell sequencing of endothelial cells from APP/PS1 transgenic mice confirms that adhesion molecule pathways are among the most highly altered pathways in these cells. At the pre-plaque stage, the vasculature shows significant signs of breakdown, with a general loss of VE-cadherin, leakage within the microcirculation, and obvious pericyte perturbation. Although senescent vascular cells were not directly observed at sites of vascular leak, senescent cells were close to the leak area. Thus, we would suggest in AD that there is a progressive induction of senescence in constituents of the neurovascular unit contributing to an increasing loss of vascular integrity. Targeting the vasculature early in AD, either with senolytics or with drugs that improve the integrity of the BBB may be valid therapeutic strategies.

Immune pathway activation in neurons triggers neural damage after stroke

Ischemic brain injury is a severe medical condition with high incidences in elderly people without effective treatment for the resulting neural damages. Using a unilateral mouse stroke model, we analyze single-cell transcriptomes of ipsilateral and contralateral cortical penumbra regions to objectively reveal molecular events with single-cell resolution at 4 h and 1, 3, and 7 days post-injury. Here, we report that neurons are among the first cells that sense the lack of blood supplies by elevated expression of CCAAT/enhancer-binding protein β (C/EBPβ). To our surprise, the canonical inflammatory cytokine gene targets for C/EBPβ, including interleukin-1β (IL-1β) and tumor necrosis factor α (TNF-α), are subsequently induced also in neuronal cells. Neuronal-specific silencing of C/EBPβ or IL-1β and TNF-α substantially alleviates downstream inflammatory injury responses and is profoundly neural protective. Taken together, our findings reveal a neuronal inflammatory mechanism underlying early pathological triggers of ischemic brain injury.

Model-based evaluation of spatiotemporal data reduction methods with unknown ground truth through optimal visualization and interpretability metrics

Optimizing and benchmarking data reduction methods for dynamic or spatial visualization and interpretation (DSVI) face challenges due to many factors, including data complexity, lack of ground truth, time-dependent metrics, dimensionality bias and different visual mappings of the same data. Current studies often focus on independent static visualization or interpretability metrics that require ground truth. To overcome this limitation, we propose the MIBCOVIS framework, a comprehensive and interpretable benchmarking and computational approach. MIBCOVIS enhances the visualization and interpretability of high-dimensional data without relying on ground truth by integrating five robust metrics, including a novel time-ordered Markov-based structural metric, into a semi-supervised hierarchical Bayesian model. The framework assesses method accuracy and considers interaction effects among metric features. We apply MIBCOVIS using linear and nonlinear dimensionality reduction methods to evaluate optimal DSVI for four distinct dynamic and spatial biological processes captured by three single-cell data modalities: CyTOF, scRNA-seq and CODEX. These data vary in complexity based on feature dimensionality, unknown cell types and dynamic or spatial differences. Unlike traditional single-summary score approaches, MIBCOVIS compares accuracy distributions across methods. Our findings underscore the joint evaluation of visualization and interpretability, rather than relying on separate metrics. We reveal that prioritizing average performance can obscure method feature performance. Additionally, we explore the impact of data complexity on visualization and interpretability. Specifically, we provide optimal parameters and features and recommend methods, like the optimized variational contractive autoencoder, for targeted DSVI for various data complexities. MIBCOVIS shows promise for evaluating dynamic single-cell atlases and spatiotemporal data reduction models.

Development of predictive models for lymphedema by using blood tests and therapy data

Lymphedema is a disease that refers to tissue swelling caused by an accumulation of protein-rich fluid that is usually drained through the lymphatic system. Detection of lymphedema is often based on expensive diagnoses such as bioimpedance spectroscopy, shear wave elastography, computed tomography, etc. In current machine learning models for lymphedema prediction, reliance on observable symptoms reported by patients introduces the possibility of errors in patient-input data. Moreover, these symptoms are often absent during the initial stages of lymphedema, creating challenges in its early detection. Identifying lymphedema before these observable symptoms manifest would greatly benefit patients by potentially minimizing the discomfort caused by these symptoms. In this study, we propose to use new data, such as complete blood count, serum, and therapy data, to develop predictive models for lymphedema. This approach aims to compensate for the limitations of using only observable symptoms data. We collected data from 2137 patients, including 356 patients with lymphedema and 1781 patients without lymphedema, with the lymphedema status of each patient confirmed by clinicians. The data for each patient included: (1) a complete blood count (CBC) test, (2) a serum test, and (3) therapy information. We used various machine learning algorithms (i.e. random forest, gradient boosting, decision tree, logistic regression, and artificial neural network) to develop predictive models on the training dataset (i.e. 80% of the data) and evaluated the models on the external validation dataset (i.e. 20% of the data). After selecting the best predictive models, we created a web application to aid medical doctors and clinicians in the rapid screening of lymphedema patients. A dataset of 2137 patients was assembled from Seoul National University Bundang Hospital. Predictive models based on the random forest algorithm exhibited satisfactory performance (balanced accuracy = 87.0 ± 0.7%, sensitivity = 84.3 ± 0.6%, specificity = 89.1 ± 1.5%, precision = 97.4 ± 0.7%, F1 score = 90.4 ± 0.4%, and AUC = 0.931 ± 0.007). We developed a web application to facilitate the swift screening of lymphedema among medical practitioners: https://snubhtxt.shinyapps.io/SNUBH_Lymphedema . Our study introduces a novel tool for the early detection of lymphedema and establishes the foundation for future investigations into predicting different stages of the condition.

Nanomedicine-based co-delivery of a calcium channel inhibitor and a small molecule targeting CD47 for lung cancer immunotherapy

Pro-tumoral macrophages in lung tumors present a significant challenge in immunotherapy. Here, we introduce a pH-responsive nanomedicine approach for activating anti-tumoral macrophages and dendritic cells. Using a layered double hydroxide nanosheet carrier, we co-deliver a T-type calcium channel inhibitor (TTA-Q6) and a CD47 inhibitor (RRX-001) into lung tumors. In the tumor acidic environment, TTA-Q6 is released, disrupting cancer cell calcium uptake, causing endoplasmic reticulum stress and inducing calreticulin transfer to the cell surface. Surface calreticulin activates macrophages and triggers dendritic cell maturation, promoting effective antigen presentation and therefore activating antitumor T cells. Simultaneously, RRX-001 reduces CD47 protein levels, aiding in preventing immune escape by calreticulin-rich cancer cells. In lung tumor models in male mice, this combined approach shows anti-tumor effects and immunity against tumor re-exposure, highlighting its potential for lung cancer immunotherapy.

Th1/interferon-γ bias in 22q11.2 deletion syndrome is driven by memory T cells and exacerbated by IL-7

The aim of this study was to investigate the impact of thymic dysplasia on the phenotypic and functional characteristics of T cells in patients with 22q11.2 deletion syndrome, including T-cell phenotype, transcriptional profile, cytokine production, as well as the possibility of utilizing IL-7 to recover their numbers and function. We found a strong bias towards Th1 response in pediatric and young adult 22q11.2DS patients, expansion of CXCR5+ follicular helper cells and CXCR3+CCR6- Th1 cells, increased production of cytokines IFN-γ, IL-10, IL-2, IL-21 and TNF-α. This Th1 skew was primarily driven by expanded terminally differentiated T cells. IL-7 further reduced naive T cells, increased cytokine production and caused an upregulation of exhaustion markers. Thus, Th1 bias in T cell populations persists from infancy into adolescence and is accompanied by accelerated maturation of T cells into memory stages. This phenotype is exacerbated by IL-7 which causes further decrease in naïve T cells in vitro.

GoM DE: interpreting structure in sequence count data with differential expression analysis allowing for grades of membership

Parts-based representations, such as non-negative matrix factorization and topic modeling, have been used to identify structure from single-cell sequencing data sets, in particular structure that is not as well captured by clustering or other dimensionality reduction methods. However, interpreting the individual parts remains a challenge. To address this challenge, we extend methods for differential expression analysis by allowing cells to have partial membership to multiple groups. We call this grade of membership differential expression (GoM DE). We illustrate the benefits of GoM DE for annotating topics identified in several single-cell RNA-seq and ATAC-seq data sets.

Preanalytical conditions for multiparameter platelet flow cytometry

Background

Flow cytometry is an important technique for understanding multiple aspects of blood platelet biology. Despite the widespread use of the platform for assessing platelet function, the optimization and careful consideration of preanalytical conditions, sample processing techniques, and data analysis strategies should be regularly assessed. When set up and designed with optimal conditions, it can ensure the acquisition of robust and reproducible flow cytometry data. However, these parameters are rarely described despite their importance.

Objectives

We aimed to characterize the effects of several preanalytical variables on the analysis of blood platelets by multiparameter fluorescent flow cytometry.

Methods

We assessed anticoagulant choice, sample material, sample processing, and storage times on 4 distinct and commonly used markers of platelet activation, including fibrinogen binding, expression of CD62P and CD42b, and phosphatidylserine exposure.

Results

The use of suboptimal conditions led to increases in basal platelet activity and reduced sensitivities to stimulation; however, the use of optimal conditions protected the platelets from artifactual stimulation and preserved basal activity and sensitivity to activation.

Conclusion

The optimal preanalytical conditions identified here for the measurement of platelet phenotype by flow cytometry suggest a framework for future development of multiparameter platelet assays for high-quality data sets and advanced analysis.

Dissection of key factors correlating with H5N1 avian influenza virus driven inflammatory lung injury of chicken identified by single-cell analysis

Chicken lung is an important target organ of avian influenza virus (AIV) infection, and different pathogenic virus strains lead to opposite prognosis. Using a single-cell RNA sequencing (scRNA-seq) assay, we systematically and sequentially analyzed the transcriptome of 16 cell types (19 clusters) in the lung tissue of chickens infected with H5N1 highly pathogenic avian influenza virus (HPAIV) and H9N2 low pathogenic avian influenza virus (LPAIV), respectively. Notably, we developed a valuable catalog of marker genes for these cell types. Compared to H9N2 AIV infection, H5N1 AIV infection induced extensive virus replication and the immune reaction across most cell types simultaneously. More importantly, we propose that infiltrating inflammatory macrophages (clusters 0, 1, and 14) with massive viral replication, pro-inflammatory cytokines (IFN-β, IL1β, IL6 and IL8), and emerging interaction of various cell populations through CCL4, CCL19 and CXCL13, potentially contributed to the H5N1 AIV driven inflammatory lung injury. Our data revealed complex but distinct immune response landscapes in the lung tissue of chickens after H5N1 and H9N2 AIV infection, and deciphered the potential mechanisms underlying AIV-driven inflammatory reactions in chicken. Furthermore, this article provides a rich database for the molecular basis of different cell-type responses to AIV infection.

Development of a 1:1-binding biparatopic anti-TNFR2 antagonist by reducing signaling activity through epitope selection

Conventional bivalent antibodies against cell surface receptors often initiate unwanted signal transduction by crosslinking two antigen molecules. Biparatopic antibodies (BpAbs) bind to two different epitopes on the same antigen, thus altering crosslinking ability. In this study, we develop BpAbs against tumor necrosis factor receptor 2 (TNFR2), which is an attractive immune checkpoint target. Using different pairs of antibody variable regions specific to topographically distinct TNFR2 epitopes, we successfully regulate the size of BpAb-TNFR2 immunocomplexes to result in controlled agonistic activities. Our series of results indicate that the relative positions of the two epitopes recognized by the BpAb are critical for controlling its signaling activity. One particular antagonist, Bp109-92, binds TNFR2 in a 1:1 manner without unwanted signal transduction, and its structural basis is determined using cryo-electron microscopy. This antagonist suppresses the proliferation of regulatory T cells expressing TNFR2. Therefore, the BpAb format would be useful in designing specific and distinct antibody functions.

DISCERN: deep single-cell expression reconstruction for improved cell clustering and cell subtype and state detection

BACKGROUND

Single-cell sequencing provides detailed insights into biological processes including cell differentiation and identity. While providing deep cell-specific information, the method suffers from technical constraints, most notably a limited number of expressed genes per cell, which leads to suboptimal clustering and cell type identification.

RESULTS

Here, we present DISCERN, a novel deep generative network that precisely reconstructs missing single-cell gene expression using a reference dataset. DISCERN outperforms competing algorithms in expression inference resulting in greatly improved cell clustering, cell type and activity detection, and insights into the cellular regulation of disease. We show that DISCERN is robust against differences between batches and is able to keep biological differences between batches, which is a common problem for imputation and batch correction algorithms. We use DISCERN to detect two unseen COVID-19-associated T cell types, cytotoxic CD4+ and CD8+ Tc2 T helper cells, with a potential role in adverse disease outcome. We utilize T cell fraction information of patient blood to classify mild or severe COVID-19 with an AUROC of 80% that can serve as a biomarker of disease stage. DISCERN can be easily integrated into existing single-cell sequencing workflow.

CONCLUSIONS

Thus, DISCERN is a flexible tool for reconstructing missing single-cell gene expression using a reference dataset and can easily be applied to a variety of data sets yielding novel insights, e.g., into disease mechanisms.

Temporal and spatial dynamics of immune cells in spontaneous liver transplant tolerance

The liver has long been deemed a tolerogenic organ. We employed high-dimensional mass cytometry and immunohistochemistry to depict the temporal and spatial dynamics of immune cells in the spleen and liver in a murine model of spontaneous liver allograft acceptance. We depicted the immune landscape of spontaneous liver tolerance throughout the rejection and acceptance stages after liver transplantation and highlighted several points of importance. Of note, the CD4+/CD8+ T cell ratio remained low, even in the tolerance phase. Furthermore, a PhenoGraph clustering analysis revealed that exhausted CD8+ T cells were the most dominant metacluster in graft-infiltrating lymphocytes (GILs), which highly expressed the costimulatory molecule CD86. The temporal and spatial dynamics of immune cells revealed by high-dimensional analyses enable a fine-grained analysis of GIL subsets, contribute to new insights for the discovery of immunological mechanisms of liver tolerance, and provide potential ways to achieve clinical operational tolerance after liver transplantation.

scDEED: a statistical method for detecting dubious 2D single-cell embeddings and optimizing t-SNE and UMAP hyperparameters

Two-dimensional (2D) embedding methods are crucial for single-cell data visualization. Popular methods such as t-SNE and UMAP are commonly used for visualizing cell clusters; however, it is well known that t-SNE and UMAP's 2D embedding might not reliably inform the similarities among cell clusters. Motivated by this challenge, we developed a statistical method, scDEED, for detecting dubious cell embeddings output by any 2D-embedding method. By calculating a reliability score for every cell embedding, scDEED identifies the cell embeddings with low reliability scores as dubious and those with high reliability scores as trustworthy. Moreover, by minimizing the number of dubious cell embeddings, scDEED provides intuitive guidance for optimizing the hyperparameters of an embedding method. Applied to multiple scRNA-seq datasets, scDEED demonstrates its effectiveness for detecting dubious cell embeddings and optimizing the hyperparameters of t-SNE and UMAP.

Immune checkpoint blockade therapy mitigates systemic inflammation and affects cellular FLIP-expressing monocytic myeloid-derived suppressor cells in non-progressor non-small cell lung cancer patients

Cancer cells favor the generation of myeloid cells with immunosuppressive and inflammatory features, including myeloid-derived suppressor cells (MDSCs), which support tumor progression. The anti-apoptotic molecule, cellular FLICE (FADD-like interleukin-1β-converting enzyme)-inhibitory protein (c-FLIP), which acts as an important modulator of caspase-8, is required for the development and function of monocytic (M)-MDSCs. Here, we assessed the effect of immune checkpoint inhibitor (ICI) therapy on systemic immunological landscape, including FLIP-expressing MDSCs, in non-small cell lung cancer (NSCLC) patients. Longitudinal changes in peripheral immunological parameters were correlated with patients' outcome. In detail, 34 NSCLC patients were enrolled and classified as progressors (P) or non-progressors (NP), according to the RECIST evaluation. We demonstrated a reduction in pro-inflammatory cytokines such as IL-8, IL-6, and IL-1β in only NP patients after ICI treatment. Moreover, using t-distributed stochastic neighbor embedding (t-SNE) and cluster analysis, we characterized in NP patients a significant increase in the amount of lymphocytes and a slight contraction of myeloid cells such as neutrophils and monocytes. Despite this moderate ICI-associated alteration in myeloid cells, we identified a distinctive reduction of c-FLIP expression in M-MDSCs from NP patients concurrently with the first clinical evaluation (T1), even though NP and P patients showed the same level of expression at baseline (T0). In agreement with the c-FLIP expression, monocytes isolated from both P and NP patients displayed similar immunosuppressive functions at T0; however, this pro-tumor activity was negatively influenced at T1 in the NP patient cohort exclusively. Hence, ICI therapy can mitigate systemic inflammation and impair MDSC-dependent immunosuppression.

GoM DE: interpreting structure in sequence count data with differential expression analysis allowing for grades of membership

Parts-based representations, such as non-negative matrix factorization and topic modeling, have been used to identify structure from single-cell sequencing data sets, in particular structure that is not as well captured by clustering or other dimensionality reduction methods. However, interpreting the individual parts remains a challenge. To address this challenge, we extend methods for differential expression analysis by allowing cells to have partial membership to multiple groups. We call this grade of membership differential expression (GoM DE). We illustrate the benefits of GoM DE for annotating topics identified in several single-cell RNA-seq and ATAC-seq data sets.

CRUSTY: a versatile web platform for the rapid analysis and visualization of high-dimensional flow cytometry data

Flow cytometry (FCM) can investigate dozens of parameters from millions of cells and hundreds of specimens in a short time and at a reasonable cost, but the amount of data that is generated is considerable. Computational approaches are useful to identify novel subpopulations and molecular biomarkers, but generally require deep expertize in bioinformatics and the use of different platforms. To overcome these limitations, we introduce CRUSTY, an interactive, user-friendly webtool incorporating the most popular algorithms for FCM data analysis, and capable of visualizing graphical and tabular results and automatically generating publication-quality figures within minutes. CRUSTY also hosts an interactive interface for the exploration of results in real time. Thus, CRUSTY enables a large number of users to mine complex datasets and reduce the time required for data exploration and interpretation. CRUSTY is accessible at https://crusty.humanitas.it/ .

GCSTI: A Single-Cell Pseudotemporal Trajectory Inference Method Based on Graph Compression

The single-cell pseudotemporal trajectory inference is an important way to explore the process of developmental changes within a cell. Due to the uneven rate of cell growth, changes in gene expression depend less on the time of data collection and more on a cell's "internal clock". To overcome the challenges of gene analysis, and replicate biological developmental processes, several strategies have been put forth. However, due to the size of single-cell datasets, locating relevant signposts usually necessitate clustering analysis or a sizable amount of priori information. To this end, we propose a novel single-cell pseudotemporal trajectory inference technique: GCSTI method, which is based on graph compression and doesn't rely on a priori knowledge or clustering procedures, can handle the trajectory inference problem for a large network in a stable and efficient manner. Additionally, we simultaneously improve the pseudotime defining method currently employed in this study in order to obtain more trustworthy and beneficial outcomes for trajectory inference. Finally, we validate the efficacy and stability of the GCSTI method using datasets from human skeletal muscle myogenic cells and four simulated datasets.

Mge-cluster: a reference-free approach for typing bacterial plasmids

Extrachromosomal elements of bacterial cells such as plasmids are notorious for their importance in evolution and adaptation to changing ecology. However, high-resolution population-wide analysis of plasmids has only become accessible recently with the advent of scalable long-read sequencing technology. Current typing methods for the classification of plasmids remain limited in their scope which motivated us to develop a computationally efficient approach to simultaneously recognize novel types and classify plasmids into previously identified groups. Here, we introduce mge-cluster that can easily handle thousands of input sequences which are compressed using a unitig representation in a de Bruijn graph. Our approach offers a faster runtime than existing algorithms, with moderate memory usage, and enables an intuitive visualization, classification and clustering scheme that users can explore interactively within a single framework. Mge-cluster platform for plasmid analysis can be easily distributed and replicated, enabling a consistent labelling of plasmids across past, present, and future sequence collections. We underscore the advantages of our approach by analysing a population-wide plasmid data set obtained from the opportunistic pathogen Escherichia coli, studying the prevalence of the colistin resistance gene mcr-1.1 within the plasmid population, and describing an instance of resistance plasmid transmission within a hospital environment.

Monocyte pathology in human tuberculosis is due to plasma milieu changes and aberrant STAT signalling

Monocyte-derived macrophages contribute centrally to immune protection in Mycobacterium tuberculosis infection and changes in monocyte phenotype characterize immunopathology in tuberculosis patients. Recent studies highlighted an important role of the plasma milieu in tuberculosis immunopathology. Here, we investigated monocyte pathology in patients with acute tuberculosis and determined tuberculosis plasma milieu effects on phenotype as well as cytokine signalling of reference monocytes. Patients with tuberculosis (n = 37) and asymptomatic contacts (controls n = 35) were recruited as part of a hospital-based study in the Ashanti region of Ghana. Multiplex flow cytometry phenotyping of monocyte immunopathology was performed and effects of individual blood plasma samples on reference monocytes prior to and during treatment were characterized. Concomitantly, cell signalling pathways were analysed to elucidate underlying mechanisms of plasma effects on monocytes. Multiplex flow cytometry visualization characterized changes in monocyte subpopulations and detected higher expression of CD40, CD64 and PD-L1 in monocytes from tuberculosis patients as compared to controls. Aberrant expression normalized during anti-mycobacterial treatment and also CD33 expression decreased markedly. Notably, higher CD33, CD40 and CD64 expression was induced in reference monocytes when cultured in the presence of plasma samples from tuberculosis patients as compared to controls. STAT signalling pathways were affected by the aberrant plasma milieu and higher levels of STAT3 and STAT5 phosphorylation was found in tuberculosis plasma-treated reference monocytes. Importantly, high pSTAT3 levels were associated with high CD33 expression and pSTAT5 correlated with CD40 as well as CD64 expression. These results suggested plasma milieu effects with potential implications on monocyte phenotype and function in acute tuberculosis.

A unique cytotoxic CD4+ T cell-signature defines critical COVID-19

Objectives

SARS-CoV-2 infection causes a spectrum of clinical disease presentation, ranging from asymptomatic to fatal. While neutralising antibody (NAb) responses correlate with protection against symptomatic and severe infection, the contribution of the T-cell response to disease resolution or progression is still unclear. As newly emerging variants of concern have the capacity to partially escape NAb responses, defining the contribution of individual T-cell subsets to disease outcome is imperative to inform the development of next-generation COVID-19 vaccines.

Methods

Immunophenotyping of T-cell responses in unvaccinated individuals was performed, representing the full spectrum of COVID-19 clinical presentation. Computational and manual analyses were used to identify T-cell populations associated with distinct disease states.

Results

Critical SARS-CoV-2 infection was characterised by an increase in activated and cytotoxic CD4+ lymphocytes (CTL). These CD4+ CTLs were largely absent in asymptomatic to severe disease states. In contrast, non-critical COVID-19 was associated with high frequencies of naïve T cells and lack of activation marker expression.

Conclusion

Highly activated and cytotoxic CD4+ T-cell responses may contribute to cell-mediated host tissue damage and progression of COVID-19. Induction of these potentially detrimental T-cell responses should be considered when developing and implementing effective COVID-19 control strategies.

Targeted ablation of signal transducer and activator of transduction 1 alleviates inflammation by microglia/macrophages and promotes long-term recovery after ischemic stroke

BACKGROUND

Brain microglia and macrophages (Mi/MΦ) can shift to a harmful or advantageous phenotype following an ischemic stroke. Identification of key molecules that regulate the transformation of resting Mi/MΦ could aid in the development of innovative therapies for ischemic stroke. The transcription factor signal transducer and activator of transduction 1 (STAT1) has been found to contribute to acute neuronal death (in the first 24 h) following ischemic stroke, but its effects on Mi/MΦ and influence on long-term stroke outcomes have yet to be determined.

METHODS

We generated mice with tamoxifen-induced, Mi/MΦ-specific knockout (mKO) of STAT1 driven by Cx3cr1^CreER. Expression of STAT1 was examined in the brain by flow cytometry and RNA sequencing after ischemic stroke induced by transient middle cerebral artery occlusion (MCAO). The impact of STAT1 mKO on neuronal cell death, Mi/MΦ phenotype, and brain inflammation profiles were examined 3-5 days after MCAO. Neurological deficits and the integrity of gray and white matter were assessed for 5 weeks after MCAO by various neurobehavioral tests and immunohistochemistry.

RESULTS

STAT1 was activated in Mi/MΦ at the subacute stage (3 days) after MCAO. Selective deletion of STAT1 in Mi/MΦ did not alter neuronal cell death or infarct size at 24 h after MCAO, but attenuated Mi/MΦ release of high mobility group box 1 and increased arginase 1-producing Mi/MΦ 3d after MCAO, suggesting boosted inflammation-resolving responses of Mi/MΦ. As a result, STAT1 mKO mice had mitigated brain inflammation at the subacute stage after MCAO and less white matter injury in the long term. Importantly, STAT1 mKO was sufficient to improve functional recovery for at least 5 weeks after MCAO in both male and female mice.

CONCLUSIONS

Mi/MΦ-targeted STAT1 KO does not provide immediate neuroprotection but augments inflammation-resolving actions of Mi/MΦ, thereby facilitating long-term functional recovery after stroke. STAT1 is, therefore, a promising therapeutic target to harness beneficial Mi/MΦ responses and improve long-term outcomes after ischemic stroke.

IndepthPathway: an integrated tool for in-depth pathway enrichment analysis based on single-cell sequencing data

MOTIVATION

Single-cell sequencing enables exploring the pathways and processes of cells, and cell populations. However, there is a paucity of pathway enrichment methods designed to tolerate the high noise and low gene coverage of this technology. When gene expression data are noisy and signals are sparse, testing pathway enrichment based on the genes expression may not yield statistically significant results, which is particularly problematic when detecting the pathways enriched in less abundant cells that are vulnerable to disturbances.

RESULTS

In this project, we developed a Weighted Concept Signature Enrichment Analysis specialized for pathway enrichment analysis from single-cell transcriptomics (scRNA-seq). Weighted Concept Signature Enrichment Analysis took a broader approach for assessing the functional relations of pathway gene sets to differentially expressed genes, and leverage the cumulative signature of molecular concepts characteristic of the highly differentially expressed genes, which we termed as the universal concept signature, to tolerate the high noise and low coverage of this technology. We then incorporated Weighted Concept Signature Enrichment Analysis into an R package called "IndepthPathway" for biologists to broadly leverage this method for pathway analysis based on bulk and single-cell sequencing data. Through simulating technical variability and dropouts in gene expression characteristic of scRNA-seq as well as benchmarking on a real dataset of matched single-cell and bulk RNAseq data, we demonstrate that IndepthPathway presents outstanding stability and depth in pathway enrichment results under stochasticity of the data, thus will substantially improve the scientific rigor of the pathway analysis for single-cell sequencing data.

AVAILABILITY AND IMPLEMENTATION

The IndepthPathway R package is available through: https://github.com/wangxlab/IndepthPathway.

CCL19+ dendritic cells potentiate clinical benefit of anti-PD-(L)1 immunotherapy in triple-negative breast cancer

BACKGROUND

The extensive involvement of dendritic cells (DCs) in immune contexture indicates their potent value in cancer immunotherapy. Understanding DC diversity in patient cohorts may strengthen the clinical benefit of immune checkpoint inhibitors (ICIs).

METHODS

Single-cell profiling of breast tumors from two clinical trials was performed to investigate DC heterogeneity. Multiomics, tissue characterization, and pre-clinical experiments were used to evaluate the role of the identified DCs in the tumor microenvironment. Four independent clinical trials were leveraged to explore biomarkers to predict ICI and chemotherapy outcomes.

FINDINGS

We identified a distinct CCL19-expressing functional state of DCs associated with favorable responses to anti-programmed death (ligand)-1 (PD-(L)1), which displayed migratory and immunomodulatory phenotypes. These cells were correlated with antitumor T cell immunity and the presence of tertiary lymphoid structures and lymphoid aggregates, defining immunogenic microenvironments in triple-negative breast cancer. In vivo, CCL19⁺ DC deletion by Ccl19 gene ablation dampened CCR7⁺CD8⁺ T cells and tumor elimination in response to anti-PD-1. Notably, high circulating and intratumoral CCL19 levels were associated with superior response and survival in patients receiving anti-PD-1 but not chemotherapy.

CONCLUSIONS

We uncovered a critical role of DC subsets in immunotherapy, which has implications for designing novel therapies and patient stratification strategies.

FUNDING

This study was funded by the National Key Research and Development Project of China, the National Natural Science Foundation of China, the Program of Shanghai Academic/Technology Research Leader, the Natural Science Foundation of Shanghai, the Shanghai Key Laboratory of Breast Cancer, the Shanghai Hospital Development Center (SHDC), and the Shanghai Health Commission.