Spatial Transcriptomics Data Research Articles

Benchmarking mapping algorithms for cell-type annotating in mouse brain by integrating single-nucleus RNA-seq and Stereo-seq data.

Limited gene capture efficiency and spot size of spatial transcriptome (ST) data pose significant challenges in cell-type characterization. The heterogeneity and complexity of cell composition in the mammalian brain make it more challenging to accurately annotate ST data from brain. Many algorithms attempt to characterize subtypes of neuron by integrating ST data with single-nucleus RNA sequencing (snRNA-seq) or single-cell RNA sequencing. However, assessing the accuracy of these algorithms on Stereo-seq ST data remains unresolved. Here, we benchmarked 9 mapping algorithms using 10 ST datasets from four mouse brain regions in two different resolutions and 24 pseudo-ST datasets from snRNA-seq. Both actual ST data and pseudo-ST data were mapped using snRNA-seq datasets from the corresponding brain regions as reference data. After comparing the performance across different areas and resolutions of the mouse brain, we have reached the conclusion that both robust cell-type decomposition and SpatialDWLS demonstrated superior robustness and accuracy in cell-type annotation. Testing with publicly available snRNA-seq data from another sequencing platform in the cortex region further validated our conclusions. Altogether, we developed a workflow for assessing suitability of mapping algorithm that fits for ST datasets, which can improve the efficiency and accuracy of spatial data annotation.

#1766 Mincle receptor in macrophage and neutrophil contributes to the unresolved inflammation during the transition from AKI to CKD

Abstract Background and Aims Recent studies have demonstrated a strong association between acute kidney injury (AKI) and chronic kidney disease (CKD). Unresolved inflammation may contribute to the AKI-to-CKD transition. As a transmembrane pattern recognition receptor, Mincle (macrophage-inducible C-type lectin, Clec4e) was previously reported to be critical in the early immune response after AKI. However, the impact of Mincle on the chronic transition from AKI remains poorly understood. Method We performed single-cell RNA sequencing (scRNA-seq) with the clinically relevant unilateral ischemia-reperfusion (UIR) murine model of AKI at days 1, 3, 14 and 28 after injury. Potential effects and mechanism of Mincle on renal inflammation and fibrosis were further validated in vivo utilizing Mincle knockout mice. Results Single-cell transcriptome analysis showed predominant Mincle expression in macrophages and neutrophils during the AKI-to-CKD transition. For both cell types, Mincle was significantly upregulated on day 1 after AKI which decreased gradually in later stages, and exhibited a secondary peak at day 14. Flow cytometry confirmed the increased infiltration of Mincle-positive macrophages and neutrophils post-UIR. Next, we identified distinct subclusters of Minclehigh neutrophils and Minclehigh macrophages, both displaying time-dependent influx with dual peaks on day 1 and day 14 and significant pro-inflammatory and pro-fibrotic phenotypes as indicated by functional gene scoring. Spatial transcriptomics data showed a pronounced expression of Mincle at the outer medulla in the kidney similar to myeloid cells, possibly correlating with fibrotic areas represented by α-SMA patterns. To further elucidate the mechanisms underlying AKI-to-CKD transition mediated by Mincle-expressing macrophages and neutrophils, we identified 54 common genes from the up-regulated genes of these two cell subsets, which were involved in immune-inflammatory responses and extracellular matrix synthesis. Importantly, TNF-α was revealed as a pivotal hub gene and strongly associated with Mincle in these two Mincle-expressing subclusters. Correspondingly, a prominent increase in TNF-α expression primarily in macrophages and neutrophils on days 1 and 14 was confirmed, displaying consistent patterns with Mincle expression. Meanwhile, significant TNF-α reduction specifically in macrophages and neutrophils on day 14 was observed in Mincle knockout mice. Remarkably, Mincle-deficient mice showed improved renal pathology and reduced extracellular matrix deposition, accompanied by notable mRNA down-regulation of pro-inflammatory factors (TNF-α, IL-1β, Ccl2) and fibrosis associated factors (TGF-β, α-SMA, COL1A1) as well as decreased immune cells infiltration. Conclusion Together, the present findings have unveiled combined persistence of Minclehigh neutrophils and macrophages during AKI-to-CKD transition, contributing to unresolved inflammation followed by fibrosis via TNF-α as a central pro-inflammatory cytokine. Targeting Mincle may offer a novel therapeutic strategy for preventing the transition from AKI to CKD.

Accurately deciphering spatial domains for spatially resolved transcriptomics with stCluster.

Spatial transcriptomics provides valuable insights into gene expression within the native tissue context, effectively merging molecular data with spatial information to uncover intricate cellular relationships and tissue organizations. In this context, deciphering cellular spatial domains becomes essential for revealing complex cellular dynamics and tissue structures. However, current methods encounter challenges in seamlessly integrating gene expression data with spatial information, resulting in less informative representations of spots and suboptimal accuracy in spatial domain identification. We introduce stCluster, a novel method that integrates graph contrastive learning with multi-task learning to refine informative representations for spatial transcriptomic data, consequently improving spatial domain identification. stCluster first leverages graph contrastive learning technology to obtain discriminative representations capable of recognizing spatially coherent patterns. Through jointly optimizing multiple tasks, stCluster further fine-tunes the representations to be able to capture complex relationships between gene expression and spatial organization. Benchmarked against six state-of-the-art methods, the experimental results reveal its proficiency in accurately identifying complex spatial domains across various datasets and platforms, spanning tissue, organ, and embryo levels. Moreover, stCluster can effectively denoise the spatial gene expression patterns and enhance the spatial trajectory inference. The source code of stCluster is freely available at https://github.com/hannshu/stCluster.

Targeting PRKDC activates the efficacy of antitumor immunity while sensitizing to chemotherapy and targeted therapy in liver hepatocellular carcinoma.

Liver hepatocellular carcinoma (LIHC) ranks among the malignancies with the highest mortality rates, primarily due to chemoresistance culminating in treatment failure. Despite its impact, predictive models addressing disease progression, tumor microenvironment, and drug sensitivity remain elusive for LIHC patients. Recognizing the significant influence of various programmed cell death (PCD) modes on tumor evolution, this study investigates PCD genes to elucidate their implications on the prognosis and immune landscape of LIHC. To develop the classification and model, we employed a total of 17 genes associated with PCD patterns. To collect data, we acquired gene expression profiles, somatic mutation information, copy number variation data, and corresponding clinical data from the TCGA database, specifically from LIHC patients. Moreover, we obtained spatial transcriptome data and additional bulk datasets from the Gene Expression Omnibus (GEO) database to conduct further analysis. Various experiments were conducted to validate the role of the PCD gene PRKDC in proliferation, migration, invasion, EMT, cell cycle, therapeutic sensitivity, and antitumor immunity. A novel LIHC classification based on these genes divided patients into two clusters, C1 and C2. The C2 cluster exhibited characteristics indicative of poor prognosis and an immune-activated microenvironment. This group showed greater response potential to immune checkpoint inhibitors, displaying higher levels of certain immune signatures and receptors. A programmed cell death index (PCDI) was constructed using 17 selected PCD genes. This index could effectively predict patient prognosis, with higher PCDI indicating poorer survival rates and a more pro-tumor microenvironment. Immune landscapes revealed varying interactions with PCDI, suggesting therapeutic targets and insights into treatment resistance. Moreover, experiments results suggested that PRKDC can augment the invasive nature and growth of malignant cells and it can serve as a potential target for therapy, offering hope for ameliorating the prognosis of LIHC patients. The study uncovers the insights of programmed cell death in the prognosis and potential therapeutic interventions. And we found that PRKDC can serve as a target for enhancing the efficacy of antitumor immunity while sensitizing chemotherapy and targeted therapy in liver hepatocellular carcinoma.

CoCo-ST: Comparing and Contrasting Spatial Transcriptomics data sets using graph contrastive learning.

Traditional feature dimension reduction methods have been widely used to uncover biological patterns or structures within individual spatial transcriptomics data. However, these methods are designed to yield feature representations that emphasize patterns or structures with dominant high variance, such as the normal tissue spatial pattern in a precancer setting. Consequently, they may inadvertently overlook patterns of interest that are potentially masked by these high-variance structures. Herein we present our graph contrastive feature representation method called CoCo-ST (Comparing and Contrasting Spatial Transcriptomics) to overcome this limitation. By incorporating a background data set representing normal tissue, this approach enhances the identification of interesting patterns in a target data set representing precancerous tissue. Simultaneously, it mitigates the influence of dominant common patterns shared by the background and target data sets. This enables discerning biologically relevant features crucial for capturing tissue-specific patterns, a capability we showcased through the analysis of serial mouse precancerous lung tissue samples.

O13 Benchmarking a human pluripotent stem cell-derived hair-bearing skin organoid model to prenatal skin

Abstract Introduction and aims Molecular regulation of mammalian skin development has been largely derived from murine studies. The lack of human studies stems primarily from the difficulties in prenatal human skin access. There is an important need to study human prenatal skin owing to the notable differences between human and mouse skin and the paucity of mechanistic understanding of congenital skin disorders. Methods We assembled the first comprehensive multiomic reference atlas of prenatal human skin (21 samples from 7 to 16 postconception weeks), combining single-cell (235 201 cells) and spatial transcriptomic data. We followed 84 skin organoids (SkOs) using single-cell analysis over a 5-month culture period and characterized the spatial organization of the skin in mature SkOs. Prenatal skin, SkO and any relevant pre-existing single-cell RNA-Seq datasets were integrated into the analysis. Cell–cell interactions and cell-lineage determination were explored using CellPhoneDB and trajectory inferences. We evaluated the differentiation trajectory alignment between prenatal skin and SkO using the Genes2Genes analysis. Results Our findings reveal predicted factors regulating the epithelial–mesenchymal crosstalk during human hair follicle formation; progressive expression of scar-promoting genes in later gestation, potentially contributing to scarless healing properties of human prenatal skin observed during the early stages of gestation and expression of genes causing genetic skin disorders in the same cell types in prenatal human skin and SkO. We also found that SkOs morphologically recapitulate human skin development from early (periderm formation and shedding) to late (cornification and hair follicle formation) stages. We also demonstrated that SkO and prenatal skin mesenchyme differentiation into hair follicle dermal condensate and dermal papilla is highly conserved at the cellular and molecular level. Conclusions SkOs faithfully recapitulate human prenatal skin development and are a valuable model for mechanistic studies on skin development including hair follicle biology and congenital skin disorders.

Computational modeling for deciphering tissue microenvironment heterogeneity from spatially resolved transcriptomics

Spatial transcriptomics techniques, while measuring gene expression, retain spatial location information, aiding in situ studies of organismal tissue architecture and the progression of pathological processes. These techniques generate vast amounts of omics data, necessitating the development of computational methods to reveal the underlying tissue microenvironment heterogeneity. The main directions in spatial transcriptomics data analysis are spatial domain detection and spatial deconvolution, which can identify spatial functional regions and parse the distribution of cell types in spatial transcriptomics data by integrating single-cell transcriptomics data. In these two research directions, many computational methods have been successively proposed. This article will categorize them into three types: machine learning-based methods, probabilistic models-based methods, and deep learning-based methods. It will list and discuss the representative algorithms of each type along with their advantages and disadvantages and describe the datasets and evaluation metrics used to assess these computational methods, facilitating researchers in selecting suitable computational methods according to their research needs. Finally, combining the latest technological developments and the advantages and disadvantages of current algorithms, this article will look forward to the future directions of computational method development.

PVRIG is Expressed on Stem-Like T Cells in Dendritic Cell-Rich Niches in Tumors and Its Blockade May Induce Immune Infiltration in Non-Inflamed Tumors.

Cancers that are poorly immune infiltrated pose a substantial challenge, with current immunotherapies yielding limited clinical success. Stem-like memory T cells (TSCM) have been identified as a subgroup of T cells that possess strong proliferative capacity and that can expand and differentiate following interactions with dendritic cells (DCs). In this study, we explored the pattern of expression of a recently discovered inhibitory receptor poliovirus receptor-related immunoglobulin domain protein (PVRIG) and its ligand, poliovirus receptor-related ligand 2 (PVRL2), in the human tumor microenvironment. Using spatial and single-cell RNA transcriptomics data across diverse cancer indications, we found that among the T-cell checkpoints, PVRIG is uniquely expressed on TSCM and PVRL2 is expressed on DCs in immune aggregate niches in tumors. PVRIG blockade could therefore enhance TSCM-DC interactions and efficiently drive T-cell infiltration to tumors. Consistent with these data, following PVRIG blockade in patients with poorly infiltrated tumors, we observed immune modulation including increased tumor T-cell infiltration, T-cell receptor (TCR) clonality, and intratumoral T-cell expansion, all of which were associated with clinical benefit. These data suggest PVRIG blockade as a promising strategy to induce potent antitumor T-cell responses, providing a novel approach to overcome resistance to immunotherapy in immune-excluded tumors.

GAGE-seq concurrently profiles multiscale 3D genome organization and gene expression in single cells.

The organization of mammalian genomes features a complex, multiscale three-dimensional (3D) architecture, whose functional significance remains elusive because of limited single-cell technologies that can concurrently profile genome organization and transcriptional activities. Here, we introduce genome architecture and gene expression by sequencing (GAGE-seq), a scalable, robust single-cell co-assay measuring 3D genome structure and transcriptome simultaneously within the same cell. Applied to mouse brain cortex and human bone marrow CD34+ cells, GAGE-seq characterized the intricate relationships between 3D genome and gene expression, showing that multiscale 3D genome features inform cell-type-specific gene expression and link regulatory elements to target genes. Integration with spatial transcriptomic data revealed in situ 3D genome variations in mouse cortex. Observations in human hematopoiesis unveiled discordant changes between 3D genome organization and gene expression, underscoring a complex, temporal interplay at the single-cell level. GAGE-seq provides a powerful, cost-effective approach for exploring genome structure and gene expression relationships at the single-cell level across diverse biological contexts.

Methods to Enable Spatial Transcriptomics of Bone Tissues.

Understanding the relationship between the cells and their location within each tissue is critical to uncover the biological processes associated with normal development and disease pathology. Spatial transcriptomics is a powerful method that enables the analysis of the whole transcriptome within tissue samples, thus providing information about the cellular gene expression and the histological context in which the cells reside. While this method has been extensively utilized for many soft tissues, its application for the analyses of hard tissues such as bone has been challenging. The major challenge resides in the inability to preserve good quality RNA and tissue morphology while processing the hard tissue samples for sectioning. Therefore, a method is described here to process freshly obtained bone tissue samples to effectively generate spatial transcriptomics data. The method allows for the decalcification of the samples, granting successful tissue sections with preserved morphological details while avoiding RNA degradation. In addition, detailed guidelines are provided for samples that were previously paraffin-embedded, without demineralization, such assamples collected from tissue banks. Using these guidelines, high-quality spatial transcriptomics data generated from tissue bank samples of primary tumor and lung metastasis of bone osteosarcoma are shown.

Abstract PO3-25-07: Characterization of tertiary lymphoid structures and its association with response to immune checkpoint inhibitors in triple negative breast cancer using spatial transcriptomics

Abstract Background Tertiary lymphoid structures (TLS) are ectopic lymphoid organs playing a role in adaptive antitumor immune response. They were shown to be associated with favorable outcome and appears to be a promising biomarker for response to immune checkpoint inhibitors (ICI). Yet, there is no consensus for their detection and quantification. Here we aimed to derive a specific TLS signature using high resolution spatial transcriptomics data from a large series of triple negative breast cancer (TNBC) samples and to assess its clinical relevance including response to ICI. Methods Spatial transcriptomics (ST) was performed on a series of 94 early-stage TNBC surgical samples with detailed clinicopathological and outcome data. TLS and tumor infiltrating lymphocytes (TILs) were scored using hematoxylin and eosin and double immunohistochemistry (CD3/CD20) stained slides by a dedicated breast pathologist. Regression based on morphological annotations and deconvolution methods estimated the fraction of gene expression related to TLS in each ST spot allowing to derive specific TLS signature. Its prognostic and predictive value of response to ICI was assessed using publicly available breast cancer (METABRIC, SCAN-B, and I-SPY2) and other tumor type datasets treated with ICI (Bareche et al. Ann Oncol, 2022) from bulk gene expression data. Results Deconvolution of the TLS compartments using the ST gene expression data showed that TLS are enriched in all B cell subsets, CD4+ (central) memory and naïve T cells, as well as mast cells. A functional analysis showed an enrichment of ‘vascular endothelial cell proliferation’, ‘mitotic G2/M transition checkpoint’, ‘V/D/J recombination’ and ‘regulation of cell chemotaxis to fibroblast growth factor’ suggesting that TLS aggregates are associated with angiogenesis, immune cell proliferation, adaptative humoral response and tissue remodeling. A comparison of gene expression data between TLS and non-aggregated TILs led to the development of a specific TLS gene signature comprising B and T cell-specific genes, immunoglobulin genes as well as genes associated with TLS initiation. Its projection on tumor slides overlapped with the regions annotated as TLS by the pathologist, demonstrating its specificity for TLS detection. As expected, the highest levels of the TLS signature were observed in the immunomodulatory TNBC molecular subtype and fully inflamed tumors, whereas the lowest levels were observed in the mesenchymal molecular subtype and immune desert tumors (p< 0.05). Of note, high levels of the TLS signature were associated with better prognosis in TNBC patients from METABRIC (p=6.5 10−5) and SCAN-B datasets (p=1.6 10−3) as well as with higher pathological complete response rates in all breast cancer patients treated with neoadjuvant pembrolizumab in the I-SPY2 study (p=0.026). Similar results were found in other tumor types treated with immunotherapy in the metastatic setting including metastatic melanoma, pancreatic and bladder cancers (PFS, p=6 10−6). Of interest, the TLS signature outperformed the predictive value of other immune-related signatures including TILs suggesting a key role of TLS in obtaining a sustainable, adaptive antitumor immune response in the vicinity of the tumor area. Conclusion By leveraging the potential of spatial transcriptomics, we have successfully developed a unique TLS gene signature that demonstrates a strong correlation with the response to ICI in breast cancer and various other tumor types. This TLS gene signature outperforms other immune biomarkers, underscoring the pivotal role of TLS in eliciting a robust antitumor immune response. The universality of this TLS signature makes it a potent biomarker capable of identifying patients who would benefit from immunotherapy, pending further validation. Citation Format: Xiaoxiao Wang, David Venet, Frédéric Lifrange, Denis Larsimont, Mattia Rediti, Linnea Stenbeck, Floriane Dupont, Ghizlane Rouas, Andrea Joaquin Garcia, David Gacquer, Ligia Craciun, Laurence Buisseret, Nayanika Bhalla, Yuvarani Masarapu, Kim Thrane, Eva Gracia Villacampa, Lovisa Franzén, Sami Saarenpää, Linda Kvastad, Joakim Lundeberg, Françoise Rothé, Christos Sotiriou. Characterization of tertiary lymphoid structures and its association with response to immune checkpoint inhibitors in triple negative breast cancer using spatial transcriptomics [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-25-07.

Abstract PO4-28-09: Upregulation of the immune checkpoint protein B7-H3 is associated with an immune suppressive microenvironment in progression from in situ to invasive lobular breast cancer

Abstract Invasive lobular breast cancer (ILC) is an understudied subtype of breast cancer characterized by late recurrence, metastasis to serosal surfaces including the peritoneum, and poor long-term outcomes. Lobular carcinoma in situ (LCIS) is a non-obligate precursor to ILC but is associated with a 30% increased risk of developing ILC. Therefore, understanding the underlying changes that occur in the invasive lobular cells as well as their tumor microenvironment (TME) during the transition from LCIS to ILC is critical for the development of novel therapeutic targets that could be used in early disease. To address this, we used combined spatial whole genome transcriptomics and a 97-protein proteomics assay to examine the spatial molecular profiles of coexistent LCIS and ILC. PanCK-positive tumor cells and panCK-negative stromal cells were segmented and analyzed separately for both RNA and protein. Despite the close spatial proximity of LCIS and ILC, there were notable differences in gene and protein expression in the tumor cells as well as the tumor microenvironment (TME). RNA profiling revealed significant upregulation of genes encoding essential components of the extracellular matrix, including type I, 3 and 5 collagens, small leucine-rich proteoglycans byglycan (BGN) and lumican (LUM), periostin (POSTN) and secreted protein acidic and cysteine rich (SPARC). Metalloprotease genes MMP2 and MMP11 were also upregulated. Genes highly expressed in the LCIS tumor compartment were KRT5, KRT14 KRT17 and MYLK which are highly expressed in myoepithelial cells at the periphery of the LCIS. One of the most upregulated proteins in ILC cells was B7 homology 3 protein (B7-H3) or Cluster of Differentiation 276 (CD276). B7-H3 is a transmembrane immunoregulatory protein belonging to the B7 family (that also includes PD-L1 and PD-L2). This immune checkpoint protein is overexpressed in many tumors but is barely expressed in normal cells. B7-H3 has been associated with several aspects of cancer progression, such as evasion of tumor immune surveillance and metastasis, and is strongly linked to poor prognosis in cancer. Compared to the LCIS TME, B7-H3 was also the most upregulated protein in the ILC TME along with the tumor-promoting M1 macrophage marker CD68 and the regulatory T-cell marker FOXP3. The lymphocyte marker CD45 was downregulated as well as T-cell proteins CD3, CD40 and granzyme B. Moreover, CD27, a memory B-cell marker was most highly expressed in the TME surrounding LCIS compared to ILC. Overall, these molecular profiles support a transition to a toward a much more immunosuppressive environment for invasive tumor cells compared to LCIS. Based on these data, our working hypothesis is that upregulation of B7-H3 on lobular tumor cells promotes invasion allowing interaction with laminins, integrins and other ECM proteins in the basement membrane and the interstitial membrane as LCIS acquire invasive properties of ILC. Furthermore, upregulation of B7-H3 on CD3+ HLA-DR+ expressing antigen presenting cells also promotes a suppressive immune microenvironment by inhibiting T-cell proliferation and downregulating cytokine production. In sum, these strongly concurrent spatial transcriptome and protein data suggest that B7-H3 and the immune suppressive network of ILC may be a promising target for early stage lobular breast cancer. B7-H3 Inhibitors could be explored in the neoadjuvant setting to evaluate the efficacy of immunotherapy in primary ILC. Citation Format: Lynda Bennett, Sunati Sahoo, Cheryl Lewis, Indu Raman, Candace Frerich, Guanchun Chen, Min Xu, Suzanne Conzen. Upregulation of the immune checkpoint protein B7-H3 is associated with an immune suppressive microenvironment in progression from in situ to invasive lobular breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-28-09.

Discrete placental gene expression signatures accompany diabetic disease classifications during pregnancy

Gestational diabetes mellitus (GDM) affects up to 10% of pregnancies and is classified into subtypes GDMA1 (managed by lifestyle modifications) and GDMA2 (requiring medication). However, whether these subtypes are distinct clinical entities or more reflective of an extended spectrum of normal pregnancy endocrine physiology remains unclear. Integrated bulk RNA-sequencing (RNA-seq), single-cell RNA-sequencing (scRNA-seq), and spatial transcriptomics harbors the potential to reveal disease gene signatures in subsets of cells and tissue microenvironments. We aimed to combine these high-resolution technologies with rigorous classification of diabetes subtypes in pregnancy. We hypothesized that differences between pre-existing Type 2 and gestational diabetes subtypes would be associated with altered gene expression profiles in specific placental cell populations. In a large case-cohort design, we compared validated cases of GDMA1, GDMA2, and type 2 diabetes (T2DM) to healthy controls by bulk RNA-seq (n=54). Quantitative analyses with RT-qPCR of presumptive genes of significant interest were undertaken in an independent and non-overlapping validation cohort of similarly well-characterized cases and controls (n=122). Additional integrated analyses of term placental single-cell, single-nuclei, and spatial transcriptomics data enabled us to determine the cellular subpopulations and niches that aligned with the GDMA1, GDMA2, and T2DM gene expression signatures at higher resolution and with greater confidence. Dimensional reduction of the bulk RNA-seq data revealed that the most common source of placental gene expression variation was the diabetic disease subtype. Relative to controls, we found 2,052 unique and significantly differentially expressed genes (-2<Log2(fold-change)>2 thresholds; q<0.05 Wald Test) among GDMA1 placental specimens, 267 among GDMA2, and 1,520 among T2DM. Several candidate marker genes (CSH1, PER1, PIK3CB, FOXO1, EGFR, IL2RB, SOD3, DOCK5, and SOGA1) were validated in an independent and non-overlapping validation cohort (q<0.05 Tukey). Functional enrichment revealed the pathways and genes most impacted for each diabetes subtype, and the degree of proximal similarity to other subclassifications. Surprisingly, GDMA1 and T2DM placental signatures were more alike by virtue of increased expression of chromatin remodeling and epigenetic regulation genes, while albumin was the top marker for GDMA2 with increased expression of placental genes in the wound healing pathway. Assessment of these gene signatures in single-cell, single-nuclei, and spatial transcriptomics data revealed high specificity and variability by placental cell and microarchitecture types. For example, at the cellular and spatial (e.g., microarchitectural) levels, distinguishing features were observed in extravillous trophoblasts (GDMA1) and macrophages (GDMA2). Lastly, we utilized these data to train and evaluate four machine learning models to estimate our confidence in predicting the control or diabetes status of placental transcriptome specimens with no available clinical metadata. Consistent with the distinct association of perinatal outcome risk, placentae from GDMA1, GDMA2, and T2DM-affected pregnancies harbor unique gene signatures that can be further distinguished by altered placental cellular subtypes and microarchitectural niches.

Integrating single-cell and spatial analysis reveals MUC1-mediated cellular crosstalk in mucinous colorectal adenocarcinoma.

Mucinous colorectal adenocarcinoma (MCA) is a distinct subtype of colorectal cancer (CRC) with the most aggressive pattern, but effective treatment of MCA remains a challenge due to its vague pathological characteristics. An in-depth understanding of transcriptional dynamics at the cellular level is critical for developing specialised MCA treatment strategies. We integrated single-cell RNA sequencing and spatial transcriptomics data to systematically profile the MCA tumor microenvironment (TME), particularly the interactome of stromal and immune cells. In addition, a three-dimensional bioprinting technique, canonical ex vivo co-culture system, and immunofluorescence staining were further applied to validate the cellular communication networks within the TME. This study identified the crucial intercellular interactions that engaged in MCA pathogenesis. We found the increased infiltration of FGF7+/THBS1+ myofibroblasts in MCA tissues with decreased expression of genes associated with leukocyte-mediated immunity and T cell activation, suggesting a crucial role of these cells in regulating the immunosuppressive TME. In addition, MS4A4A+ macrophages that exhibit M2-phenotype were enriched in the tumoral niche and high expression of MS4A4A+was associated with poor prognosis in the cohort data. The ligand-receptor-based intercellular communication analysis revealed the tight interaction of MUC1+ malignant cells and ZEB1+ endothelial cells, providing mechanistic information for MCA angiogenesis and molecular targets for subsequent translational applications. Our study provides novel insights into communications among tumour cells with stromal and immune cells that are significantly enriched in the TME during MCA progression, presenting potential prognostic biomarkers and therapeutic strategies for MCA. Tumour microenvironment profiling of MCA is developed. MUC1+ tumour cells interplay with FGF7+/THBS1+ myofibroblasts to promote MCA development. MS4A4A+ macrophages exhibit M2 phenotype in MCA. ZEB1+ endotheliocytes engage in EndMT process in MCA.

Spatial transcriptome-guided multi-scale framework connects P. aeruginosa metabolic states to oxidative stress biofilm microenvironment.

With the generation of spatially resolved transcriptomics of microbial biofilms, computational tools can be used to integrate this data to elucidate the multi-scale mechanisms controlling heterogeneous biofilm metabolism. This work presents a Multi-scale model of Metabolism In Cellular Systems (MiMICS) which is a computational framework that couples a genome-scale metabolic network reconstruction (GENRE) with Hybrid Automata Library (HAL), an existing agent-based model and reaction-diffusion model platform. A key feature of MiMICS is the ability to incorporate multiple -omics-guided metabolic models, which can represent unique metabolic states that yield different metabolic parameter values passed to the extracellular models. We used MiMICS to simulate Pseudomonas aeruginosa regulation of denitrification and oxidative stress metabolism in hypoxic and nitric oxide (NO) biofilm microenvironments. Integration of P. aeruginosa PA14 biofilm spatial transcriptomic data into a P. aeruginosa PA14 GENRE generated four PA14 metabolic model states that were input into MiMICS. Characteristic of aerobic, denitrification, and oxidative stress metabolism, the four metabolic model states predicted different oxygen, nitrate, and NO exchange fluxes that were passed as inputs to update the agent's local metabolite concentrations in the extracellular reaction-diffusion model. Individual bacterial agents chose a PA14 metabolic model state based on a combination of stochastic rules, and agents sensing local oxygen and NO. Transcriptome-guided MiMICS predictions suggested microscale denitrification and oxidative stress metabolic heterogeneity emerged due to local variability in the NO biofilm microenvironment. MiMICS accurately predicted the biofilm's spatial relationships between denitrification, oxidative stress, and central carbon metabolism. As simulated cells responded to extracellular NO, MiMICS revealed dynamics of cell populations heterogeneously upregulating reactions in the denitrification pathway, which may function to maintain NO levels within non-toxic ranges. We demonstrated that MiMICS is a valuable computational tool to incorporate multiple -omics-guided metabolic models to mechanistically map heterogeneous microbial metabolic states to the biofilm microenvironment.

Identifying spatial domains from spatial transcriptome by graph attention network

Due to the high dimensionality and complexity of the data, the analysis of spatial transcriptome data has been a challenging problem. Meanwhile, cluster analysis is the core issue of the analysis of spatial transcriptome data. In this article, a deep learning approach is proposed based on graph attention networks for clustering analysis of spatial transcriptome data. Our method first enhances the spatial transcriptome data, then uses graph attention networks to extract features from nodes, and finally uses the Leiden algorithm for clustering analysis. Compared with the traditional non-spatial and spatial clustering methods, our method has better performance in data analysis through the clustering evaluation index. The experimental results show that the proposed method can effectively cluster spatial transcriptome data and identify different spatial domains, which provides a new tool for studying spatial transcriptome data.

Cross-modality mapping using image varifolds to align tissue-scale atlases to molecular-scale measures with application to 2D brain sections

This paper explicates a solution to building correspondences between molecular-scale transcriptomics and tissue-scale atlases. This problem arises in atlas construction and cross-specimen/technology alignment where specimens per emerging technology remain sparse and conventional image representations cannot efficiently model the high dimensions from subcellular detection of thousands of genes. We address these challenges by representing spatial transcriptomics data as generalized functions encoding position and high-dimensional feature (gene, cell type) identity. We map onto low-dimensional atlas ontologies by modeling regions as homogeneous random fields with unknown transcriptomic feature distribution. We solve simultaneously for the minimizing geodesic diffeomorphism of coordinates through LDDMM and for these latent feature densities. We map tissue-scale mouse brain atlases to gene-based and cell-based transcriptomics data from MERFISH and BARseq technologies and to histopathology and cross-species atlases to illustrate integration of diverse molecular and cellular datasets into a single coordinate system as a means of comparison and further atlas construction.

Identification of BGN positive fibroblasts as a driving factor for colorectal cancer and development of its related prognostic model combined with machine learning

BackgroundNumerous studies have indicated that cancer-associated fibroblasts (CAFs) play a crucial role in the progression of colorectal cancer (CRC). However, there are still many unknowns regarding the exact role of CAF subtypes in CRC.MethodsThe data for this study were obtained from bulk, single-cell, and spatial transcriptomic sequencing data. Bioinformatics analysis, in vitro experiments, and machine learning methods were employed to investigate the functional characteristics of CAF subtypes and construct prognostic models.ResultsOur study demonstrates that Biglycan (BGN) positive cancer-associated fibroblasts (BGN + Fib) serve as a driver in colorectal cancer (CRC). The proportion of BGN + Fib increases gradually with the progression of CRC, and high infiltration of BGN + Fib is associated with poor prognosis in terms of overall survival (OS) and recurrence-free survival (RFS) in CRC. Downregulation of BGN expression in cancer-associated fibroblasts (CAFs) significantly reduces migration and proliferation of CRC cells. Among 101 combinations of 10 machine learning algorithms, the StepCox[both] + plsRcox combination was utilized to develop a BGN + Fib derived risk signature (BGNFRS). BGNFRS was identified as an independent adverse prognostic factor for CRC OS and RFS, outperforming 92 previously published risk signatures. A Nomogram model constructed based on BGNFRS and clinical-pathological features proved to be a valuable tool for predicting CRC prognosis.ConclusionIn summary, our study identified BGN + Fib as drivers of CRC, and the derived BGNFRS was effective in predicting the OS and RFS of CRC patients.

Spatial transcriptomics reveals segregation of tumor cell states in glioblastoma and marked immunosuppression within the perinecrotic niche

Glioblastoma (GBM) remains an untreatable malignant tumor with poor patient outcomes, characterized by palisading necrosis and microvascular proliferation. While single-cell technology made it possible to characterize different lineage of glioma cells into neural progenitor-like (NPC-like), oligodendrocyte-progenitor-like (OPC-like), astrocyte-like (AC-like) and mesenchymal like (MES-like) states, it does not capture the spatial localization of these tumor cell states. Spatial transcriptomics empowers the study of the spatial organization of different cell types and tumor cell states and allows for the selection of regions of interest to investigate region-specific and cell-type-specific pathways. Here, we obtained paired 10x Chromium single-nuclei RNA-sequencing (snRNA-seq) and 10x Visium spatial transcriptomics data from three GBM patients to interrogate the GBM microenvironment. Integration of the snRNA-seq and spatial transcriptomics data reveals patterns of segregation of tumor cell states. For instance, OPC-like tumor and NPC-like tumor significantly segregate in two of the three samples. Our differentially expressed gene and pathway analyses uncovered significant pathways in functionally relevant niches. Specifically, perinecrotic regions were more immunosuppressive than the endogenous GBM microenvironment, and perivascular regions were more pro-inflammatory. Our gradient analysis suggests that OPC-like tumor cells tend to reside in areas closer to the tumor vasculature compared to tumor necrosis, which may reflect increased oxygen requirements for OPC-like cells. In summary, we characterized the localization of cell types and tumor cell states, the gene expression patterns, and pathways in different niches within the GBM microenvironment. Our results provide further evidence of the segregation of tumor cell states and highlight the immunosuppressive nature of the necrotic and perinecrotic niches in GBM.

High-density generation of spatial transcriptomics with STAGE.

Spatial transcriptome technologies have enabled the measurement of gene expression while maintaining spatial location information for deciphering the spatial heterogeneity of biological tissues. However, they were heavily limited by the sparse spatial resolution and low data quality. To this end, we develop a spatial location-supervised auto-encoder generator STAGE for generating high-density spatial transcriptomics (ST). STAGE takes advantage of the customized supervised auto-encoder to learn continuous patterns of gene expression in space and generate high-resolution expressions for given spatial coordinates. STAGE can improve the low quality of spatial transcriptome data and smooth the generated manifold of gene expression through the de-noising function on the latent codes of the auto-encoder. Applications to four ST datasets, STAGE has shown better recovery performance for down-sampled data than existing methods, revealed significant tissue structure specificity, and enabled robust identification of spatially informative genes and patterns. In addition, STAGE can be extended to three-dimensional (3D) stacked ST data for generating gene expression at any position between consecutive sections for shaping high-density 3D ST configuration.