April Top 10 Publications - Advances in Single Cell Sequencing
Our monthly single cell Literature Blitz column is back again. April featured a plethora of exciting new studies that advanced our understanding of single cell methods and disease treatments. Here is our selection of top publications featuring new advances and applications of single cell sequencing in immune profiling of leukemia, cerebellum landscape via machine learning, fallopian tube, glioblastoma and bladder cancer cellular landscapes, worm whole brain topography, retinal organoids, autoimmune prostate disease and worm infestation microenvironment.
What are your April top choices?
TCR Profiling in Leukemia [↑]
T-cell large granular lymphocyte leukemia (T-GLL) is a lymphoproliferative disease and bone marrow failure syndrome which responds to immunosuppressive therapies. It typically is characterised by clonal expansion of terminally differentiated effector-memory cytotoxic T cells.
Gao et al. applied single-cell TCR sequencing coupled with scRNA-seq to CD3+ T-cells obtained from a relatively large cohort of T-GLL patients and sought to characterise TCR repertoire to define pathophysiologic mechanisms at the single-cell level. They also wanted to determine mechanisms of action of an immune therapy in this disease and found that T-GLL patients demonstrate an expansion of CD8+ effector T-cells and loss of TCR clonality, which can be modulated by immunosuppressive treatment.
Figure 1: Loss of TCR repertoire diversity in T-LGLL patients. (Left) Percentages of effector memory T cell and CD8+/CD4+ T cell ratio comparisons between patients. (Right) Dot plot showing frequency of a sum of top 3 clones in individual patients and healthy donors. (Down). Circos plots: individual cells yielding rearranged TCR sequences. Black lines connected clones sharing identical CDR3 sequences among individuals.
Cerebellum and Machine Learning [↑]
The cerebellum acts as a master regulator in the brain as well as controlling fine motor skills. Disruption of cerebellar function is linked to schizophrenia, autism, and depression. The cerebellum is made up of 4 main cell types and it is different between species. Gaining insights into those differences can shine a light on evolutionary processes in the brain.
Li et al. analyzed hundreds of thousands of cells from single-cell sequencing datasets and applied machine learning (ML) algorithms to classify cells within a species and differentiate between cerebellar cells between humans and mice. They showed that 648 features can be used to distinguish mouse and human cerebellar cells by combining several ML algorithms. This study defined gene combinations that can distinguish mouse and human cerebellar cells.
Figure 2: Box plots of MCC values yielded by classifiers with randomly selected gene features on datasets of four cerebellum cell types. (A, B) Box plots on datasets for Granule cells, (C). Box plots on datasets for Interneuron cells, (D) Box plots on datasets for Unipolar brush cells. The figure also shows tables containing cell type annotation markers and statistics.
Fallopian Tube Cellular Landscapes. [↑]
Fallopian tube (FT) homeostasis requires dynamic regulation of heterogenous cell populations and is disrupted in fertility and ovarian cancer. To understand how FT cell numbers change in a disease state, the Ulrich et al. analysed samples from hydrosalpinx (blocked FT) and observed shifts in epithelial and stromal populations, and cell-type-specific changes in extracellular matrix and TGF-β signalling.
The authors identified six stromal subtypes in human fallopian tube. The finding was also confirmed in vivo with immunofluorescence or immunohistochemistry with representative markers. Multiple progenitor populations that contribute to epithelial and stromal cell homeostasis were identified using RNA velocity analysis, as well as roles of immune cells and cytokines in FT disease progression. The study provides a refined cell atlas of the pre-menopausal fallopian tube.
Figure 3: Major cell types and markers identified from single-cell RNA-seq analysis of healthy human fallopian tube. Identification of 12 major cell types from global clustering of cells from healthy subjects, visualized in UMAP space.
Glioblastoma Cellular Landscapes [↑]
Glioblastoma (GBM) is the most common and aggressive subtype of aggressive brain tumors, characterized by extensive inter- and intratumor heterogeneity. Patient-derived models, such as organoids and explants, have recently emerged as useful models to study such heterogeneity, although the extent to which they can recapitulate GBM genomic features remains unclear.
LeBlanc et al. used single cell sequencing to investigate cellular and molecular heterogeneity in GBM model systems. They determined the extent to which single-cell genomic analysis could detect phenotypic heterogeneity in addition to genetic heterogeneity in PDEs and revealed that PDEs but not GS lines retain some non-malignant cells present in their parent tumors.
Figure 4: UMAP visualization of all tissue (A), PDE (B), and GS (D) cells, colored by patient (key shown in C).
Topography of the entire C. Elegans nervous system [↑]
The complete anatomy and wiring diagram of the C. elegans nervous system were defined previously by serial section electron microscopy. A comprehensive investigation of the relationship between gene expression and neuroanatomy in C. elegans has not been provided until now.
Taylor et al. generated gene expression profiles for all 302 neurons comprising the entire nervous system of C. elegans using single cell RNA sequencing to characterize the transcriptomic profiles of FACS-isolated neuronal cells labelled by the transgenic enhancer trap lines. Gene expression profiles for all 118 neuron classes in C. elegans were obtained and each neuron type was shown to expresses a distinct code of neuropeptide genes and receptors. Also, they showed that cell adhesion molecules correlate with neuron-specific connectivity.
Figure 5: All known neuron types in the C. elegans nervous system are identified as individual clusters of scRNA-seq profiles. (A) All neuron types in the mature C. elegans hermaphrodite. (B) UMAP projection of all neuron types and subtypes of ten anatomically defined classes. (C) Graphical representation of neurons targeted in individual experiments. (D) the LUA cluster exclusively expressed C39H7.2.
Bladder Cancer Ecosystem [↑]
Cystitis glandularis (CG) usually occurs in chronic inflammation leading to epithelial hyperplasia and metaplasia, which although otherwise benign, could lead to bladder cancer (BLCA) in patients with intestinal metaplasia and urinary tract accumulation.
Luo et al. aimed to provide new biological knowledge about the cell composition of CG and the heterogeneity of different grades of bladder cancer. Authors performed single cell RNA sequencing and comprehensively described the expression characteristics of malignant epitheliums and immune cells, as well as the dynamic changes of cell percentages, and the heterogeneity of cell subtypes. They found that a macrophage cluster that highly expressed proinflammatory cytokine TNF, while T cells did not. Another macrophage cluster was shown to highly express immunosuppressive and angiogenic phenotype-related markers that promote tumor progression associated with CD163.
Figure 6: Changes of cell composition in microenvironment of different bladder samples. (Left) UMAP plot of immune cells and stromal cells, showing different cell types. (Right) T-cell pseudotime development trajectory.
Retinal Organoids [↑]
 Dorgau et al. Human Retinal Organoids Provide a Suitable Tool for Toxicological Investigations: A Comprehensive Validation Using Drugs and Compounds Affecting the Retina. Stem Cells Translational Medicine, 2022. [PubMed]
Retinal drug toxicity screening is essential for the development of safe treatment strategies for many diseases. To this end, retinal organoids derived from human pluripotent stem cells (hPSCs) provide a suitable screening platform due to their similarity to the human retina and the ease of generation in large-scale formats
Dorgau et al. aimed to investigate the effectiveness of hPSC-derived organoids on drug testing and validate differentiation by single cell RNA sequencing. They showed that hiPSC cells differentiated to retinal organoids in 150-200 days and tested 6 drugs for toxicity. Drug effects on retinal organoids were comparable to in vivo responses.
Figure 7: UMAP maps showed 21 transcriptionally distinct cell clusters, comprising all the main retinal cell types together with non-retinal cell clusters, including a proliferating cell and a Fibroblast cell cluster.
Non-invasive Breast Cancer Treatments [↑]
Minimally invasive therapies like microwave ablation (MWA), have been attempted in the treatment of early-stage breast cancer. In situ tumor ablation can create an antigen source and these tumor specific antigens are presented to lymphocytes by dendritic cells and macrophages.
Zhou et al. aimed to find key cellular subsets and underlying mechanism of MWA-induced immune response in the treatment of early-stage breast cancer and performed single cell RNA sequencing of PBMC immune profiles from breast cancer patients before and after MWA. They showed that B cells were important antigen-presenting cells (APCs) that initiate CD4+T cells in MWA-induced immune response. The procedure activated CD8+ T cells and increased NK cell cytotoxicity. Also, immune checkpoint inhibitors synergistically activated peripheral T cells after MWA in vitro.
Figure 8: UMAP analysis of peripheral NK and T cells showing 8 clusters before and after MWA.
Autoimmune Prostate Hyperplasia [↑]
Autoimmune (AI) diseases can affect many organs; however, the prostate has not been considered to be a primary target of these systemic inflammatory processes. Here, Vickman et al. utilized medical record data, patient samples, and in vivo models to evaluate the impact of inflammation, as seen in AI diseases, on prostate tissue. Human and mouse tissues were used to examine whether systemic targeting of inflammation limits prostatic inflammation in benign prostatic hyperplasia (BPH). They implicated BPH T cells and macrophages as sources of TNF and showed that TNF-antagonists (ETANECEPT) reduce prostate hyperplasia and inflammation in NOD mice.
Figure 9: Workflow scheme and UMAP plot of 69,850 individual cells from 14 patient samples, demonstrating dominant T cell and macrophage populations.
Mechanisms of Helminth Infestation [↑]
Human whipworms (Trichuris trichiura) infect hundreds of millions of people and cause trichuriasis, a major neglected tropical disease with high chronic morbidity and dire socio-economic consequences in affected countries. The molecular mechanisms allowing for the penetration of whipworm larvae to intestinal epithelia are unknown.
Duque-Correa et al. used single cell RNA sequencing to characterize the transcriptomic profiles of whipworm infected mouse caecum and established Murine caecaloids, the first in-vitro system for whipworm infection and organoid model for live helminths. They revealed that progression of infection results in cell damage and an expansion of enterocytes expressing of Isg15, potentially instigating the host immune response to the whipworm and tissue repair.
Figure 10: Host IECs responses to early infection with whipworms are dominated by a type-I IFN signature. a Bulk RNA-seq data was analyzed for cell signature genes in the IFN alpha pathway.
 Dorgauet al. Human Retinal Organoids Provide a Suitable Tool for Toxicological Investigations: A Comprehensive Validation Using Drugs and Compounds Affecting the Retina. Stem Cells Translational Medicine, 2022.[PubMed]
- THE END -