RNA Velocity Research Articles

Human umbilical cord mesenchymal stem cell-derived exosomes promote murine skin wound healing by neutrophil and macrophage modulations revealed by single-cell RNA sequencing.

Full-thickness skin wound healing remains a serious undertaking for patients. While stem cell-derived exosomes have been proposed as a potential therapeutic approach, the underlying mechanism of action has yet to be fully elucidated. The current study aimed to investigate the impact of exosomes derived from human umbilical cord mesenchymal stem cells (hucMSC-Exosomes) on the single-cell transcriptome of neutrophils and macrophages in the context of wound healing. Utilizing single-cell RNA sequencing, the transcriptomic diversity of neutrophils and macrophages was analyzed in order to predict the cellular fate of these immune cells under the influence of hucMSC-Exosomes and to identify alterations of ligand-receptor interactions that may influence the wound microenvironment. The validity of the findings obtained from this analysis was subsequently corroborated by immunofluorescence, ELISA, and qRT-PCR. Neutrophil origins were characterized based on RNA velocity profiles. The expression of RETNLG and SLC2A3 was associated with migrating neutrophils, while BCL2A1B was linked to proliferating neutrophils. The hucMSC-Exosomes group exhibited significantly higher levels of M1 macrophages (215 vs 76, p < 0.00001), M2 macrophages (1231 vs 670, p < 0.00001), and neutrophils (930 vs 157, p < 0.00001) when compared to control group. Additionally, it was observed that hucMSC-Exosomes elicit alterations in the differentiation trajectories of macrophages towards more anti-inflammatory phenotypes, concomitant with changes in ligand-receptor interactions, thereby facilitating healing. This study has revealed the transcriptomic heterogeneity of neutrophils and macrophages in the context of skin wound repair following hucMSC-Exosomes interventions, providing a deeper understanding of cellular responses to hucMSC-Exosomes, a rising target of wound healing intervention.

Identification of phenotypically, functionally, and anatomically distinct stromal niche populations in human bone marrow based on single-cell RNA sequencing.

Hematopoiesis is regulated by the bone marrow (BM) stroma. However, cellular identities and functions of the different BM stromal elements in humans remain poorly defined. Based on single-cell RNA sequencing (scRNAseq), we systematically characterized the human non-hematopoietic BM stromal compartment and we investigated stromal cell regulation principles based on the RNA velocity analysis using scVelo and studied the interactions between the human BM stromal cells and hematopoietic cells based on ligand-receptor (LR) expression using CellPhoneDB. scRNAseq led to the identification of six transcriptionally and functionally distinct stromal cell populations. Stromal cell differentiation hierarchy was recapitulated based on RNA velocity analysis and in vitro proliferation capacities and differentiation potentials. Potential key factors that might govern the transition from stem and progenitor cells to fate-committed cells were identified. In situ localization analysis demonstrated that different stromal cells were localized in different niches in the bone marrow. In silico cell-cell communication analysis further predicted that different stromal cell types might regulate hematopoiesis through distinct mechanisms. These findings provide the basis for a comprehensive understanding of the cellular complexity of the human BM microenvironment and the intricate stroma-hematopoiesis crosstalk mechanisms, thus refining our current view on human hematopoietic niche organization.

Increased retinoic acid signaling decreases lung metastasis in salivary adenoid cystic carcinoma by inhibiting the noncanonical Notch1 pathway

MYB-NFIB fusion and NOTCH1 mutation are common hallmark genetic events in salivary gland adenoid cystic carcinoma (SACC). However, abnormal expression of MYB and NOTCH1 is also observed in patients without MYB-NFIB fusion and NOTCH1 mutation. Here, we explore in-depth the molecular mechanisms of lung metastasis through single-cell RNA sequencing (scRNA-seq) and exome target capture sequencing in two SACC patients without MYB-NFIB fusion and NOTCH1 mutation. Twenty-five types of cells in primary and metastatic tissues were identified via Seurat clustering and categorized into four main stages ranging from near-normal to cancer-based on the abundance of each cell cluster in normal tissue. In this context, we identified the Notch signaling pathway enrichment in almost all cancer cells; RNA velocity, trajectory, and sub-clustering analyses were performed to deeply investigate cancer progenitor-like cell clusters in primary tumor-associated lung metastases, and signature genes of progenitor-like cells were enriched in the “MYC_TARGETS_V2” gene set. In vitro, we detected the NICD1-MYB-MYC complex by co-immunoprecipitation (Co-IP) and incidentally identified retinoic acid (RA) as an endogenous antagonist of genes in the “MYC_TARGETS_V2” gene set. Following this, we confirmed that all-trans retinoic acid (ATRA) suppresses the lung metastasis of SACC by correcting erroneous cell differentiation mainly caused by aberrant NOTCH1 or MYB expression. Bioinformatic, RNA-seq, and immunohistochemical (IHC) analyses of primary tissues and metastatic lung tissues from patients with SACC suggested that RA system insufficiency partially promotes lung metastasis. These findings imply the value of the RA system in diagnosis and treatment.

Transcriptional and open chromatin analysis of bovine skeletal muscle development by single-cell sequencing.

Skeletal muscle is a complex heterogeneous tissue and characterizing its cellular heterogeneity and transcriptional and epigenetic signatures are important for understanding the details of its ontogeny. In our study, we applied scRNA-seq and scATAC-seq to investigate the cell types, molecular features, transcriptional and epigenetic regulation, and patterns of developing bovine skeletal muscle from gestational, lactational and adult stages. Detailed molecular analyses were used to dissect cellular heterogeneity, and we deduced the differentiation trajectory of myogenic cells and uncovered their dynamic gene expression profiles. SCENIC analysis was performed to demonstrate key regulons during cell fate decisions. We explored the future expression states of these heterogeneous cells by RNA velocity analysis and found extensive networks of intercellular communication using the toolkit CellChat. Moreover, the transcriptomic and chromatin accessibility modalities were confirmed to be highly concordant, and integrative analysis of chromatin accessibility and gene expression revealed key transcriptional regulators acting during myogenesis. In bovine skeletal muscle, by scRNA-seq and scATAC-seq analysis, different cell types such as adipocytes, endothelial cells, fibroblasts, lymphocytes, monocytes, pericyte cells and eight skeletal myogenic subpopulations were identified at the three developmental stages. The pseudotime trajectory exhibited a distinct sequential ordering for these myogenic subpopulations and eight distinct gene clusters were observed according to their expression pattern. Moreover, specifically expressed TFs (such as MSC, MYF5, MYOD1, FOXP3, ESRRA, BACH1, SIX2 and ATF4) associated with muscle development were predicted, and likely future transcriptional states of individual cells and the developmental dynamics of differentiation among neighbouring cells were predicted. CellChat analysis on the scRNA-seq data set then classified many ligand-receptor pairs among these cell clusters, which were further categorized into significant signalling pathways, including BMP, IGF, WNT, MSTN, ANGPTL, TGFB, TNF, VEGF and FGF. Finally, scRNA-seq and scATAC-seq results were successfully integrated to reveal a series of specifically expressed TFs that are likely to be candidates for the promotion of cell fate transition during bovine skeletal muscle development. Overall, our results outline a single-cell dynamic chromatin/transcriptional landscape for normal bovine skeletal muscle development; these provide an important resource for understanding the structure and function of mammalian skeletal muscle, which will promote research into its biology.

Single cell sequencing of the mouse anterior palate reveals mesenchymal heterogeneity.

Cleft palate is one of the most prevalent birth defects. Mice are useful for studying palate development because of their morphological and genetic similarities to humans. In mice, palate development occurs between embryonic days (E)11.5 to 15.5. Single cell transcriptional profiles of palate cell populations have been a valuable resource for the craniofacial research community, but we lack a single cell transcriptional profile for anterior palate at E13.5, at the transition from proliferation to shelf elevation. A detailed single cell RNA sequencing analysis reveals heterogeneity in expression profiles of the cell populations of the E13.5 anterior palate. Hybridization chain reaction RNA fluorescent in situ hybridization (HCR RNA FISH) reveals epithelial populations segregate into layers. Mesenchymal populations spatially segregate into four domains. One of these mesenchymal populations expresses ligands and receptors distinct from the rest of the mesenchyme, suggesting that these cells have a unique function. RNA velocity analysis shows two terminal cell states that contribute to either the proximal or distal palatal regions emerge from a single progenitor pool. This single cell resolution expression data and detailed analysis from E13.5 anterior palate provides a powerful resource for mechanistic insight into secondary palate morphogenesis for the craniofacial research community.

Using single cell atlas data to reconstruct regulatory networks.

Inference of global gene regulatory networks from omics data is a long-term goal of systems biology. Most methods developed for inferring transcription factor (TF)-gene interactions either relied on a small dataset or used snapshot data which is not suitable for inferring a process that is inherently temporal. Here, we developed a new computational method that combines neural networks and multi-task learning to predict RNA velocity rather than gene expression values. This allows our method to overcome many of the problems faced by prior methods leading to more accurate and more comprehensive set of identified regulatory interactions. Application of our method to atlas scale single cell data from 6 HuBMAP tissues led to several validated and novel predictions and greatly improved on prior methods proposed for this task.

DOP65 T-cell derived TGFβ1 remodels the rectum mesenchyme in Crohn’s Disease patients with perianal fistulizing disease

Abstract Background Perianal fistulizing disease (PFD) is a complication occurring in about 20% of Crohn’s Disease (CD) patients. It consists of the formation of new epithelised tracks connecting the intestine with the outer perianal region, which dramatically worsens a patient’s quality of life. The mechanisms driving fistula formation have not been elucidated and therefore, there are currently no specific treatments for PFD. Methods We collected rectum biopsies of CD patients with and without PFD in the absence of rectal inflammation (CD and CD+PFD), and with active rectal inflammation (CDinf and CD+PFDinf) (Fig.1). Rectal tissue samples were processed for RT-qPCR, or histologic analysis, and single-cell RNA sequencing (scRNAseq) for each patient. Results Differential expression gene (DEG) analysis of CD4+ and CD8+ T cells comparing CD vs CD+PFD and CDinf vs CD+PFDinf patients, respectively, revealed increased expression of TGFB1 in patients with fistulae. To investigate the main cellular subsets TGFB1 was interacting with, we performed ligand-receptor interaction analysis using CellphoneDB, which indicated that T-cell derived TGFB1 was primary interacting with the stromal compartment (Fig.2). 365 DEG were found in lamina propria (LP) fibroblasts of CD+PFD compared to CD, including TGFβ1-inducible genes such as TFPI2, SERPINE2 and MMPs 1 and 3. These changes were validated by bulk RT-qPCR, underscoring their potential use as PFD biomarkers (Fig.3). Moreover, we also identified a signature of elastin fibre assembly genes dysregulated in CD+PFD fibroblasts, suggesting molecular changes in the formation of elastin fibres in these patients. To determine whether the differentiation of fibroblasts was altered in the context of PFD, we conducted trajectory analysis using RNA velocity. Whilst fibroblasts in CD patients preferentially differentiate towards myofibroblasts, patients with PFD are biased towards LP fibroblasts, indicating an overall structural rearrangement of the mesenchyme in the rectum of patients with PFD, regardless of their inflammatory status (Fig.4). Importantly, analysis of an independent scRNAseq dataset of CD and ulcerative colitis (UC) patients showed higher expression of TGFB1 in T cells of active CD compared to active UC patients, suggesting that these alterations are already present in CD patients before the onset of a fistulizing phenotype and might potentially contribute to explain the absence of PFD in UC patients. Conclusion Based on these results, we suggest that the rectum of patients suffering from PFD present cellular and transcriptomic differences compared to CD patients without PFD. These changes are inflammation independent and point towards the TGFβ pathway as the main driver of this complication.

Comparing the transcriptome of developing native and iPSC-derived mouse retinae by single cell RNA sequencing

We report the generation and analysis of single-cell RNA-Seq data (> 38,000 cells) from mouse native retinae and induced pluripotent stem cell (iPSC)-derived retinal organoids at four matched stages of development spanning the emergence of the major retinal cell types. We combine information from temporal sampling, visualization of 3D UMAP manifolds, pseudo-time and RNA velocity analyses, to show that iPSC-derived 3D retinal organoids broadly recapitulate the native developmental trajectories. However, we observe relaxation of spatial and temporal transcriptome control, premature emergence and dominance of photoreceptor precursor cells, and susceptibility of dynamically regulated pathways and transcription factors to culture conditions in retinal organoids. We demonstrate that genes causing human retinopathies are enriched in cell-type specifying genes and identify a subset of disease-causing genes with expression profiles that are highly conserved between human retinae and murine retinal organoids. This study provides a resource to the community that will be useful to assess and further improve protocols for ex vivo recapitulation and study of retinal development.

Single-cell RNA sequencing reveals resident progenitor and vascularization-associated cell subpopulations in rat annulus fibrosus.

One of the main causes of low back pain is intervertebral disc degeneration (IDD). Annulus fibrosus (AF) is important for the integrity and functions of the intervertebral disc (IVD). However, the resident functional cell components such as progenitors and vascularization-associated cells in AF are yet to be fully identified. Identification of functional AF cell subpopulations including resident progenitors and vascularization-associated cells. In this study, the single-cell RNA sequencing data of rat IVDs from a public database were analyzed using Seurat for cell clustering, gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) for functional analysis, StemID for stem cell identification, Monocle and RNA velocity for pseudotime differentiation trajectory validation, single-cell regulatory network inference and clustering (SCENIC) for gene regulatory network (GRN) analysis, and CellChat for cell-cell interaction analysis. Immunostaining on normal and degenerated rat IVDs, as well as human AF, was used for validations. From the data analysis, seven AF cell clusters were identified, including two newly discovered functional clusters, the Grem1 + subpopulation and the Lum +​subpopulation. The Grem1 + subpopulation had progenitor characteristics, while the Lum +​subpopulation was associated with vascularization during IDD. The GRN analysis showed that Sox9 and Id1 were among the key regulators in the Grem1 + subpopulation, and Nr2f2 and Creb5 could be responsible for the vascularization function in the Lum +​subpopulation. Cell-cell interaction analysis revealed highly regulated cellular communications between these cells, and multiple signaling networks including PDGF and MIF signaling pathways were involved in the interactions. Our results revealed two new functional AF cell subpopulations, with stemness and vascularization induction potential, respectively. These findings complement our knowledge about IVDs, especially the AF, and in return provide potential cell source and regulation targets for IDD treatment and tissue repair. The existence of the cell subpopulations was also validated in human AF, which strengthen the clinical relevance of the findings.

Transcriptional advantage influence odorant receptor gene choice.

Odorant receptors (ORs) obey mutual exclusivity and monoallelic mode of expression. Efforts are ongoing to decipher the molecular mechanism that drives the 'one-neuron-one-receptor' rule of olfaction. Recently, single-cell profiling of olfactory sensory neurons (OSNs) revealed the expression of multiple ORs in the immature neurons, suggesting that the OR gene choice mechanism is much more complex than previously described by the silence-all-and-activate-one model. These results also led to the genesis of two possible mechanistic models i.e. winner-takes-all and stochastic selection. We developed Reverse Cell Tracking (RCT), a novel computational framework that facilitates OR-guided cellular backtracking by leveraging Uniform Manifold Approximation and Projection embeddings from RNA Velocity Workflow. RCT-based trajectory backtracking, coupled with statistical analysis, revealed the OR gene choice bias for the transcriptionally advanced (highest expressed) OR during neuronal differentiation. Interestingly, the observed selection bias was uniform for all ORs across different spatial zones or their relative expression within the olfactory organ. We validated these findings on independent datasets and further confirmed that the OR gene selection may be regulated by Upf3b. Lastly, our RNA dynamics-based tracking of the differentiation cascade revealed a transition cell state that harbors mixed molecular identities of immature and mature OSNs, and their relative abundance is regulated by Upf3b.

An atlas of late prenatal human neurodevelopment resolved by single-nucleus transcriptomics

Late prenatal development of the human neocortex encompasses a critical period of gliogenesis and cortical expansion. However, systematic single-cell analyses to resolve cellular diversity and gliogenic lineages of the third trimester are lacking. Here, we present a comprehensive single-nucleus RNA sequencing atlas of over 200,000 nuclei derived from the proliferative germinal matrix and laminating cortical plate of 15 prenatal, non-pathological postmortem samples from 17 to 41 gestational weeks, and 3 adult controls. This dataset captures prenatal gliogenesis with high temporal resolution and is provided as a resource for further interrogation. Our computational analysis resolves greater complexity of glial progenitors, including transient glial intermediate progenitor cell (gIPC) and nascent astrocyte populations in the third trimester of human gestation. We use lineage trajectory and RNA velocity inference to further characterize specific gIPC subpopulations preceding both oligodendrocyte (gIPC-O) and astrocyte (gIPC-A) lineage differentiation. We infer unique transcriptional drivers and biological pathways associated with each developmental state, validate gIPC-A and gIPC-O presence within the human germinal matrix and cortical plate in situ, and demonstrate gIPC states being recapitulated across adult and pediatric glioblastoma tumors.

A Guide to Trajectory Inference and RNA Velocity.

Technological developments have led to an explosion of high-throughput single-cell data, which are revealing unprecedented perspectives on cell identity. Recently, significant attention has focused on investigating, from single-cell RNA-sequencing (scRNA-seq) data, cellular dynamic processes, such as cell differentiation, cell cycle and cell (de)activation. In particular, trajectory inference methods, by ordering cells along a trajectory, allow estimating a differentiation tree of cells. While trajectory inference tools typically work with gene expression levels, common scRNA-seq protocols allow the identification and quantification of unspliced pre-mRNAs and mature spliced mRNAs for each gene. By exploiting the abundance of unspliced and spliced mRNA, one can infer the RNA velocity of individual cells, i.e., the time derivative of the gene expression state of cells. Whereas traditional trajectory inference methods reconstruct cellular dynamics given a population of cells of varying maturity, RNA velocity relies on a dynamical model describing splicing dynamics. Here, we initially discuss conceptual and theoretical aspects of both approaches, then illustrate how they can be combined together, and finally present an example use case on real data.

DeepVelo: Single-cell transcriptomic deep velocity field learning with neural ordinary differential equations.

Recent advances in single-cell sequencing technologies have provided unprecedented opportunities to measure the gene expression profile and RNA velocity of individual cells. However, modeling transcriptional dynamics is computationally challenging because of the high-dimensional, sparse nature of the single-cell gene expression measurements and the nonlinear regulatory relationships. Here, we present DeepVelo, a neural network-based ordinary differential equation that can model complex transcriptome dynamics by describing continuous-time gene expression changes within individual cells. We apply DeepVelo to public datasets from different sequencing platforms to (i) formulate transcriptome dynamics on different time scales, (ii) measure the instability of cell states, and (iii) identify developmental driver genes via perturbation analysis. Benchmarking against the state-of-the-art methods shows that DeepVelo can learn a more accurate representation of the velocity field. Furthermore, our perturbation studies reveal that single-cell dynamical systems could exhibit chaotic properties. In summary, DeepVelo allows data-driven discoveries of differential equations that delineate single-cell transcriptome dynamics.

Uncovering single‐nucleus RNA velocity variations in neuropathologic Alzheimer’s disease

AbstractBackgroundDifferential single‐nucleus (snRNA‐seq) gene expression analyses in Alzheimer’s disease (AD) provide fixed snapshots of cellular alterations, failing to detect the temporal dynamics of genes at individual cells. To overcome this limitation, here we analyze single‐nucleus RNA velocities (RNA‐vel) from the prefrontal cortex to characterize dynamic genetic and cellular differences in neuropathological AD progression. RNA‐vel is the rate of change of gene expression obtained by comparing intronic and exonic sequence counts. Comparison of AD pathology associated gene expression with parallel RNA‐vel differences reveals sets of genes and molecular pathways that underlie the static and putative dynamic regimes of cell type‐specific dysregulations underlying the disease.MethodWe used snRNA‐seq data from the prefrontal cortex of 48 subjects1 with varied levels of AD pathology from ROSMAP2. We then calculated both the expression and RNA‐vel differences between low AD‐pathology and mild‐to‐severe AD‐pathology. These differences were cell‐specific across six major cell types: excitatory neurons, inhibitory neurons, astrocytes, microglia, oligodendrocytes, and oligodendrocyte progenitor cells.1. We characterized RNA‐vel differences associated with four AD neuropathology traits: neuritic plaque, neuronal neurofibrillary tangle, overall β‐amyloid load, and PHF tau tangle density.2. We demonstrated the reproducibility of the RNA‐vel differences in two independent datasets: dorsolateral prefrontal cortex (N = 24) from ROSMAP3 and superior frontal gyrus (N = 6) from GEO4.ResultOnly 10 of 843 (1.2%) genes overlap between the differential RNA expression and velocity analyses, suggesting substantial AD‐pathology related differences between these two RNA descriptors. The genes with only velocity differences relate to cell developmental and synaptic processes. Conversely, the genes with only differential expression are majorly associated with mitochondrial activity and ribosomal processes. Many genes were reproduced in the independent datasets; these overlapping genes are again associated with neural development and synaptic activities across many cell types, as well as vascular‐ and immune‐related processes in astrocytes and microglia, respectivelyConclusionWe use RNA velocity to characterize, for the first time to our knowledge, the dynamical multicellular processes underlying neuropathological AD progression. The results support the validity of the novel RNA‐vel concept for achieving a complementary molecular characterization of AD, which may be obscured by typical analysis of RNA abundance alone.

Simulation-based inference of differentiation trajectories from RNA velocity fields

We report Cytopath, a method for trajectory inference that takes advantage of transcriptional activity information from the RNA velocity of single cells to perform trajectory inference. Cytopath performs this task by defining a Markov chain model, simulating an ensemble of possible differentiation trajectories, and constructing a consensus trajectory. We show that Cytopath can recapitulate the topological and molecular characteristics of the differentiation process under study. In our analysis, we include differentiation trajectories with varying bifurcated, circular, convergent, and mixed topologies studied in single-snapshot as well as time-series single-cell RNA sequencing experiments. We demonstrate the capability to reconstruct differentiation trajectories, assess the association of RNA velocity-based pseudotime with actually elapsed process time, and identify drawbacks in current state-of-the art trajectory inference approaches.

Cancer immunotherapies transition endothelial cells into HEVs that generate TCF1+ T lymphocyte niches through a feed-forward loop.

The lack of Tcell infiltrates is a major obstacle to effective immunotherapy in cancer. Conversely, the formation of tumor-associated tertiary-lymphoid-like structures (TA-TLLSs), which are the local site of humoral and cellular immune responses against cancers, is associated with good prognosis, and they have recently been detected in immune checkpoint blockade (ICB)-responding patients. However, how these lymphoid aggregates develop remains poorly understood. By employing single-cell transcriptomics, endothelial fate mapping, and functional multiplex immune profiling, we demonstrate that antiangiogenic immune-modulating therapies evoke transdifferentiation of postcapillary venules into inflamed high-endothelial venules (HEVs) via lymphotoxin/lymphotoxin beta receptor (LT/LTβR) signaling. In turn, tumor HEVs boost intratumoral lymphocyte influx and foster permissive lymphocyte niches for PD1- and PD1+TCF1+ CD8 Tcell progenitors that differentiate into GrzB+PD1+ CD8T effector cells. Tumor-HEVs require continuous CD8 and NK cell-derived signals revealing that tumor HEV maintenance is actively sculpted by the adaptive immune system through a feed-forward loop.

Cascade Forest-Based Model for Prediction of RNA Velocity.

In recent years, single-cell RNA sequencing technology (scRNA-seq) has developed rapidly and has been widely used in biological and medical research, such as in expression heterogeneity and transcriptome dynamics of single cells. The investigation of RNA velocity is a new topic in the study of cellular dynamics using single-cell RNA sequencing data. It can recover directional dynamic information from single-cell transcriptomics by linking measurements to the underlying dynamics of gene expression. Predicting the RNA velocity vector of each cell based on its gene expression data and formulating RNA velocity prediction as a classification problem is a new research direction. In this paper, we develop a cascade forest model to predict RNA velocity. Compared with other popular ensemble classifiers, such as XGBoost, RandomForest, LightGBM, NGBoost, and TabNet, it performs better in predicting RNA velocity. This paper provides guidance for researchers in selecting and applying appropriate classification tools in their analytical work and suggests some possible directions for future improvement of classification tools.

Trib1 Inhibition Promotes the Klr Effector State and Augments Checkpoint Blockade to Enhance T Cell Function during T Cell Exhaustion

Background: T cell exhaustion is a homeostatic state that limits the immune system's ability to clear persistent antigen such as cancer or chronic infection. During the initial stages of antigen exposure, T cells have a robust response. Then proliferation and function dampen, and they settle into a state of exhaustion (TEX) marked by increased expression of inhibitory receptors. However, some cells in the exhaustion pathway retain effector function and gene signatures, a feature that is associated with expression of KLRs (killer lectin-like receptors) and effector genes, such as granzymes, with decreased expression of inhibitory receptors. Growing evidence in humans and mice suggests these cells are important cytotoxic effectors. Despite their potential benefit, the signaling molecules regulating the differentiation of this population are poorly understood. Our lab recently identified the pseudokinase Tribbles 1 (Trib1) as a novel regulator of T cell effector function during chronic infection (Rome et al, JEM 2020). In the current study, we use scRNA-seq, RNA velocity, flow cytometry, and TCR clonal analysis to demonstrate that deleting Trib1 shifts TEX towards KLR+CD8+ T cells (TKLR), away from a terminal TEX fate, and that loss of Trib1 augments anti-PDL1 blockade to improve viral control during chronic infection. Methods and Results: We previously performed scRNA-seq on CD44+TCRβ+ T cells from WT (CD4-Cre+Trib1+/+) and Trib1 cKO (CD4-Cre+Trib1f/f) mice that were infected with LCMV clone 13 for 15 days. When CD4+ and CD8+ T cells were analyzed, we noted that a CD8+KLRG1+ cluster was uniquely enriched in Trib1 cKO cells compared to WT cells, and that this subset had a transcriptional profile strikingly similar to effector T cells from acute infection (Rome et al, JEM 2020). To better understand how Trib1 regulates CD8+ T cell biology during chronic infection, we specifically analyzed only CD8+ T cells from this data set and found that CD8+ Trib1 cKO cells are enriched for multiple KLR genes such as Klrk1, Klrg1, Klrc1, Klrd1, Klrb1c, Klre1, and KLR-associated genes such as Hcst (encoding Dap10), as well as Gzma and Cx3cr1 compared to WT cells, which express higher levels of exhaustion-associated genes such as Pdcd1, Tox, and Eomes. Total CD8+ cells were clustered using TooManyCells (Schwartz et al, Nat. Methods 2020), and based on gene expression profiles the tree was annotated to include progenitors (TPROG), transitional cells (TTRANS), terminally exhausted TTERM, and TKLR (Fig. 1). RNA velocity was used to predict the directional flow of cell differentiation states within the CD8+ cells in each cluster, and both WT and Trib1 cKO CD8+ cells progress from TPROG to TTRANS states, which then bifurcate into either TTERM or TKLR subsets. However, in Trib1 cKO mice, the propensity towards TKLR differentiation was stronger than in WT mice. scRNA-seq and flow cytometry confirmed that Trib1 cKO mice are enriched for the TKLR subset, while WT mice are enriched for TTERM cells, confirming this shift in cell fate decision (Fig. 1). Single cell TCR sequencing further revealed that Trib1 cKO CD8+ cells preferentially clonally expand at the TTRANS and TKLR stage, reinforcing that Trib1 deletion during T cell exhaustion redirects the fate of CD8+ cells towards an effector-like TKLR state. Finally, Trib1 deletion from T cells of mice treated with immune checkpoint inhibition using an anti-PDL1 antibody during chronic viral infection improves viral load (2.2 x 104 PFU/ml in Trib1 WT vs 9.1 x 103 PFU/ml in Trib1 cKO mice; p = 0.024) and this is associated with expanded TKLR subsets. Conclusions: Trib1 deletion from mouse T cells during chronic infection intrinsically promotes a TKLR fate by increasing the expression of KLR-associated genes at all stages of exhaustion and by driving the clonal expansion and differentiation of TTRANS towards an effector-like TKLR identity. We demonstrate for the first time that deletion of a single signaling molecule can promote a TKLR fate while diverting the pathway of exhaustion towards a more effector-like state, and that this is accompanied by improved viral control. Additionally, Trib1 deletion improved anti-PDL1 treatment in chronically infected mice by enhancing the expansion of TKLR. Thus, manipulating Trib1, or it's signaling partners, could provide a novel means to improve T cell function during chronic antigen exposure, such as infection or cancer therapy. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Monocyte/Macrophage Heterogeneity during Skin Wound Healing in Mice.

Monocytes (Mos)/macrophages (Mϕs) orchestrate biological processes critical for efficient skin wound healing. However, current understanding of skin wound Mo/Mϕ heterogeneity is limited by traditional experimental approaches such as flow cytometry and immunohistochemistry. Therefore, we sought to more fully explore Mo/Mϕ heterogeneity and associated state transitions during the course of excisional skin wound healing in mice using single-cell RNA sequencing. The live CD45+CD11b+Ly6G- cells were isolated from skin wounds of C57BL/6 mice on days 3, 6, and 10 postinjury and captured using the 10x Genomics Chromium platform. A total of 2813 high-quality cells were embedded into a uniform manifold approximation and projection space, and eight clusters of distinctive cell populations were identified. Cluster dissimilarity and differentially expressed gene analysis categorized those clusters into three groups: early-stage/proinflammatory, late-stage/prohealing, and Ag-presenting phenotypes. Signature gene and Gene Ontology analysis of each cluster provided clues about the different functions of the Mo/Mϕ subsets, including inflammation, chemotaxis, biosynthesis, angiogenesis, proliferation, and cell death. Quantitative PCR assays validated characteristics of early- versus late-stage Mos/Mϕs inferred from our single-cell RNA sequencing dataset. Additionally, cell trajectory analysis by pseudotime and RNA velocity and adoptive transfer experiments indicated state transitions between early- and late-state Mos/Mϕs as healing progressed. Finally, we show that the chemokine Ccl7, which was a signature gene for early-stage Mos/Mϕs, preferentially induced the accumulation of proinflammatory Ly6C+F4/80lo/- Mos/Mϕs in mouse skin wounds. In summary, our data demonstrate the complexity of Mo/Mϕ phenotypes, their dynamic behavior, and diverse functions during normal skin wound healing.

Combined Inhibition of CDK2 and BCL2 Overrides Resistance to Targeting BTK and CDK4/6 in Mantle Cell Lymphoma Therapy

Mantle cell lymphoma (MCL) remains incurable due to the development of drug resistance, despite the plethora of therapies available. Each successive treatment failure is associated with a more rapidly proliferating disease and fewer treatment options. Understanding the drug-resistance mechanism and developing effective, well-tolerated, new therapies for MCL are urgently needed. Unrestrained proliferation of MCL cells is mainly driven by aberrant cyclin D1 expression and dysregulated CDK4 activity that promotes cell cycle progression from G1 to S. In our phase 1 clinical trial inhibiting CDK4/6 with palbociclib and BTK with ibrutinib (PALIBR) in recurrent MCL, the complete response rate (CR) was 42% compared to 21% in response to ibrutinib alone, despite a comparable 67% overall response rate. Moreover, 5 patients remain on therapy (2 CR and 3 PR) for ~ 8 years, suggesting that CDK4/6 inhibition deepens and prolongs the clinical response to BTK inhibition in MCL. However, resistance emerges in some patients. To elucidate the underpinnings, we undertook integrative longitudinal genomic analysis of sequential tissue and blood specimens (n=53) of 27 MCL patients before, during therapy and on progression, and 4 treatment-naïve MCL patients and 4 normal subjects as controls. Whole exome and whole transcriptome sequencing of freshly isolated MCL cells and B cells revealed that resistance to PALIBR is rooted in homozygous deletion of RB1 (direct target of CDK4/6) or compound copy number variation (cCNV) - concurrent hemizygous loss of RB1 and CDKN2A (p16, CDK4/6 inhibitor) and gain of CDK4. In other resistant patients, the RB protein is silenced by translational termination mutation. Single-cell RNA-sequencing (scRNA-seq, 210,000 cells) of corresponding PBMCs using a unique in-house MCL RNA reference library reveals that MCL cells comprise 4 transcriptomically distinct clusters. Cluster 1 (C1) is similar to quiescent normal B cells; C2 resembles hyper-activated B cells enriched for signatures of BCR and cytokine signaling; C3 represents non-proliferating, long-lived MCL cells; and C4 is highly proliferative, expanding with disease progression. RNA velocity analysis further uncovered that C2 fuels the transition to C4 with disease progression, whereas C3 is a terminal node (Fig 1). In C2 and C3 MCL cells, CDK2 is highly elevated, indicating evasion of CDK4/6 inhibition, and PIK3IP1 (negative PI3K regulator) is silenced, suggesting that compensatory PI3K activation in BTKi resistance (Chiron et al, Cancer Discovery 2014) is secondary to loss of cell cycle control. BCL2 is markedly elevated in C2, greater in C3 MCL cells along with increases in anti-apoptotic BCL2L1(BCL-XL) and BCL-2A1 and decreases in pro-apoptotic PMAIP1 (NOXA) expression. This suggests combined CDK2 and BCL2 inhibition as a rational approach to override CDK4/6i and BTKi resistance. Indeed, combined inhibition of CDK2 with PF-07104091 and BCL2 with venetoclax (PF-VEN) cooperatively impaired growth and killed ibrutinib-resistant MAVER-1 cells, more effectively in the isogenic derivative MAVER-1R cells which we generated to recapitulate cCNV in PALIBR resistance - depletion of RB protein through translational termination mutation (W99, AAF 75%) in RB1, loss of CDKN2A and gain of CDK4. Mechanistically, MAVER-1R cells are accelerated in G1 to S progression, based on elevation of CDK2 and TK1 expression and cell cycle profiling. They are intrinsically apoptotic and vulnerable to BCL2 inhibition due to a marked reduction in cells expressing anti-apoptotic BCL2L1 and IRF4, and universal expression of pro-apoptotic PMAIP1 (NOXA). Importantly, combined CDK2 and BCL2 inhibition synergistically kills MCL cells of a patient resistant to PALIBR ex vivo. In summary, by longitudinal genomic analysis, we have provided the first evidence that 1) resistance to CDK4/6i and BTKi stems from MCL intrinsic cCNV; 2) MCL cells comprise 4 major transcriptomically distinct clusters; cluster 2 expressing copious CDK2 and BCL2 is pivotal in fueling proliferation of MCL cells in resistance; 3) cluster 3 cells also accumulate with resistance due to longevity; 4) combined inhibition of CDK2 and BCL2 overrides resistance to CDK4i and BTKi, in MCL cell lines and primary MCL cells from PALIBR resistant patient. Combined inhibition of CDK2 and BCL2, therefore, represents a mechanism-based strategy to overcome CDK4/6i and BTKi resistance in MCL therapy. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal