Wei Lin

Cancer-associated fibroblasts (CAFs) represent a major structural component of solid tumors and play crucial roles in cancer progression and drug resistance. However, their developmental origin, differentiation trajectory, and therapeutic potential remain poorly defined. Using advanced approaches—including inducible genetic lineage tracing, single-cell RNA sequencing, and spatial transcriptomic profiling—we identified a population of Cd34 ⁺ Pi16 ⁺ fibroblast progenitors (Cd34 ⁺ CAFs) in both melanoma and gastric cancer. We delineated their differentiation trajectory toward Acta2 ⁺ CAFs, driven by the upregulation of the transcription factor Foxs1. This work establishes the developmental origin of Acta2 ⁺ CAFs and experimentally validates the Cd34 ⁺ to Acta2 ⁺ transition. By reverse-matching the transcriptional signatures of Acta2 ⁺ CAF differentiation with the CMap/LINCS L1000 drug perturbation database, we identified four small-molecule candidates predicted to inhibit tumor-induced Foxs1 upregulation. These compounds effectively suppressed Cd34 ⁺ CAF differentiation, maintaining the progenitor-like Cd34 ⁺ state. Collectively, this study proposes a novel antitumor strategy that targets CAF lineage development to restrain tumor progression.

SUMMARY: Alzheimer’s disease (AD) is the most common form of dementia worldwide. Despite extensive progress, the cellular and molecular mechanisms of AD remain incompletely understood, partially due to inadequate disease models. To illuminate the earliest changes in hereditary (familial) Alzheimer’s disease, we developed an isogenic AD cerebrocortical organoid (CO) model. Our refined methodology produces COs containing excitatory and inhibitory neurons alongside glial cells, utilizing established isogenic wild-type and diseased human induced pluripotent stem cells (hiPSCs) carrying heterozygous familial AD mutations, namely PSEN1 ^ΔE9/WT , PSEN1 ^M146V/WT , or APP ^swe/WT . Our CO model reveals time-progressive accumulation of amyloid beta (Aβ) species, loss of monomeric Tau, and accumulation of aggregated high-molecular-weight (HMW) phospho(p)-Tau species. This is accompanied by neuronal hyperexcitability, as observed in early human AD cases on electroencephalography (EEG), and synapse loss. Single-cell RNA-sequencing analyses reveal significant differences in molecular abnormalities in excitatory vs. inhibitory neurons, helping explain AD clinical phenotypes. Finally, we show that chronic dosing with autophagy activators, including a novel CNS-penetrant mTOR inhibitor-independent drug candidate, normalizes pathologic accumulation of Aβ and HMW p-Tau, normalizes hyperexcitability, and rescues synaptic loss in COs. Collectively, our results demonstrate these COs are a useful human AD model suitable for assessing early features of familial AD etiology and for testing drug candidates that ameliorate or prevent molecular AD phenotypes.

In this study, we have devised a computational framework called Supervised Feature Learning and Scoring (SuperFeat) that allows for the training of a machine learning model and evaluates the canonical cellular status/features in pathological tissues that underlie the progression of disease. This framework also enables the identification of potential drugs that target the presumed detrimental cellular features. This framework was constructed on the basis of an artificial neural network with the gene expression profiles serving as input nodes. The training data comprised single-cell RNA sequencing datasets that encompassed the specific cell lineage during the developmental progression of cell features. A few models of the canonical cancer-involved cellular status/features were tested by such framework. Finally, we have illustrated the drug repurposing pipeline, utilizing the training parameters derived from the adverse cellular status/features, which has yielded successful validation results both in vitro and in vivo. SuperFeat is accessible at https://github.com/weilin-genomics/rSuperFeat.

In this study, we have devised a computational framework SuperFeat that allows for the training of a machine learning model and evaluate the canonical cellular states/features in pathological tissues that underlie the progression of disease. This framework also enables the identification of potential drugs that target the presumed detrimental cellular features. This framework was constructed on the basis of an artificial neural network with the gene expression profiles serving as input nodes. The training data comprised single-cell RNA-seq datasets that encompassed the specific cell lineage during the developmental progression of cell features. A few models of the canonical cancer-involved cellular states/features were tested by such framework. Finally, we have illustrated the drug repurposing pipeline, utilizing the training parameters derived from the adverse cellular states/features, which has yielded successful validation results both in vitro and in vivo . SuperFeat is accessible at https://github.com/weilin-genomics/rSuperFeat .

SUMMARY: Systematic understanding of immune dynamics across the entire human lifespan at single-cell resolution is currently lacking. Here, we performed single-cell RNA sequencing (scRNA-seq) and single-cell T cell receptor (TCR)/B cell receptor (BCR) sequencing (scTCR/BCR-seq) on over 380,000 peripheral blood mononuclear cells collected from 45 healthy participants aged 0 to over 90 years. We revealed that the functions of T cell subsets were most susceptible to senescence among all PBMCs, featured by increased NF-κB signaling and IFN-γ responses pathways, while reduced telomere maintenance and energy metabolism. We subsequently explored the rewiring of cell-cell interactions among different immune cells across the lifespan and revealed the major alteration of immune checkpoints in T cells within the cellular interaction networks. By combined analysis of scRNA-seq and scTCR-seq, we revealed that 1) GNLY ⁺ CD8 Effector T cells exhibited a high clonal expansion with distinct functional signatures in children and the elderly; 2) Naïve CD4 ⁺ T and naïve CD8 ⁺ T cells displayed different aging patterns in both transcriptomes and immune repertoires; and 3) CD8 ⁺ MAIT cells showed a higher cell abundance and clonal diversity in adolescents. Furthermore, we identified a unique cytotoxic B cell subset enriched in children by scRNA-seq and scBCR-seq analysis and experimental validations. In summary, our work provided valuable insights and rich resources for understanding the development and aging of the human immune system across the lifespan.

Eicosanoids, oxidized fatty acids that serve as cell-signaling molecules, have been broadly implicated in tumorigenesis. To identify eicosanoids relevant to pancreatic tumorigenesis, we profiled normal pancreas and pancreatic ductal adenocarcinoma (PDAC) in mouse models and patient samples using mass spectrometry. We interrogated RNA sequencing datasets for eicosanoid synthase or receptor expression. Findings were confirmed by immunostaining. In murine models, we identified elevated levels of PGD ₂ , prostacyclin, and thromboxanes in neoplasia while PGE ₂ , 12-HHTre, HETEs, and HDoHEs are elevated specifically in tumors. Analysis of scRNA-seq datasets suggests that PGE ₂ and prostacyclins are derived from fibroblasts, PGD ₂ and thromboxanes from myeloid cells, and PGD ₂ and 5-HETE from tuft cells. In patient samples, we identified a transition from PGD ₂ to PGE ₂ -producing enzymes in the epithelium during the transition to PDAC, fibroblast/tumor expression of PTGIS, and myeloid/tumor cell expression of TBXAS1. Altogether, our analyses identify key changes in eicosanoid species during pancreatic tumorigenesis and the cell types responsible for their synthesis.

We systematically investigated the transcriptomes of the peripheral immune cells from 6 inactivated vaccine, BBIBP-CorV recipients at 4 pivotal time points using single-cell RNA-seq technique. First, the significant variation of the canonical immune-responsive signals of both humoral and cellular immunity, as well as other possible symptom-driver signals were evaluated in the specific cell types. Second, we described and compared the common and distinct variation trends across COVID-19 vaccination, disease progression, and flu vaccination to achieve in-depth understandings of the manifestation of immune response in peripheral blood under different stimuli. Third, the expanded T cell and B cell clones were correlated to the specific phenotypes which allowed us to characterize the antigen-specific ones much easier in the future. At last, other than the coagulopathy, the immunogenicity of megakaryocytes in vaccination were highlighted in this study. In brief, our study provided a rich data resource and the related methodology to explore the details of the classical immunity scenarios.

ScRNA-seq has the ability to reveal accurate and precise cell types and states. Existing scRNA-seq platforms utilize bead-based technologies uniquely barcoding individual cells, facing practical challenges for precious samples with limited cell number. Here, we present a scRNA-seq platform, named Paired-seq, with high cells/beads utilization efficiency, cell-free RNAs removal capability, high gene detection ability and low cost. We utilize the differential flow resistance principle to achieve single cell/barcoded bead pairing with high cell utilization efficiency (95%). The integration of valves and pumps enables the complete removal of cell-free RNAs, efficient cell lysis and mRNA capture, achieving highest mRNA detection accuracy (R = 0.955) and comparable sensitivity. Lower reaction volume and higher mRNA capture and barcoding efficiency significantly reduce the cost of reagents and sequencing. The single-cell expression profile of mES and drug treated cells reveal cell heterogeneity, demonstrating the enormous potential of Paired-seq for cell biology, developmental biology and precision medicine.

Characterizing the cell identity in a heterogeneous tissue is essential to the in-depth understanding of this sample. Existing single-cell techniques (e.g., flow cytometry or in situ cell florescent imaging) allow us to do so using the high/low signal of a combination of multiple signature molecules or even of a single marker. Recent advance of single-cell RNA-seq technology profiles the entire transcriptome of individual cells. Using a few marker genes to characterize cell type in this new technique is less reliable due to the high noise level and the dynamic transcription behavior. Nonetheless, the “noisy” but high-throughput transcriptome profiles provide adequate information to reveal the cellular identity and to understand the detail of the molecular characteristics. In this chapter, we will demonstrate a new method that is based on the supervised learning of the single-cell transcriptome profiles of many different known cell types. We will demonstrate how this technique solves the cellular identity problem.

Granulosa cells (GCs) play a critical role in driving the formation of ovarian follicles and building the cumulus-oocyte complex surrounding the ovum. We are particularly interested in assessing oocyte quality by examining the detailed gene expression profiles of human cumulus single cells. Using single-cell RNAseq techniques, we extensively investigated the single-cell transcriptomes of the cumulus GC populations from two women with normal ovarian function. This allowed us to elucidate the endogenous heterogeneity of GCs by uncovering the hidden GC subpopulation. The subsequent validation results suggest that CD24(+) GCs are essential for triggering ovulation. Treatment with human chorionic gonadotropin (hCG) significantly increases the expression of CD24 in GCs. CD24 in cultured human GCs is associated with hCG-induced upregulation of prostaglandin synthase (ARK1C1, PTGS2, PTGES, and PLA2G4A) and prostaglandin transporter (SLCO2A1 and ABCC4) expression, through supporting the EGFR-ERK1/2 pathway. In addition, it was observed that the fraction of CD24(+) cumulus GCs decreases in PCOS patients compared to that of controls. Altogether, the results support the finding that CD24 is an important mediator of ovulation and that it may also be used for therapeutic target of ovulatory disorders.