Automated identification of the mouse brain's spatial compartments from in situ sequencing data

Crosstalk between tumor and microenvironment: Insights from spatial transcriptomics

Cancer is a dynamic disease, and clonal heterogeneity plays a fundamental role in tumor development, progression, and resistance to therapies. Single-cell and spatial multimodal technologies can provide a high-resolution molecular map of underlying genomic, epigenomic, and transcriptomic alterations involved in inter- and intra-tumor heterogeneity and interactions with the microenvironment. In this review, we provide a perspective on factors driving cancer heterogeneity, tumor evolution, and clonal states. We briefly describe spatial transcriptomic technologies and summarize recent literature that sheds light on the dynamical interactions between tumor states, cell-to-cell communication, and remodeling local microenvironment.

Points2Regions: Fast, interactive clustering of imaging-based spatial transcriptomics data

Imaging-based spatial transcriptomics techniques generate data in the form of spatial points belonging to different mRNA classes. A crucial part of analyzing the data involves the identification of regions with similar composition of mRNA classes. These biologically interesting regions can manifest at different spatial scales. For example, the composition of mRNA classes on a cellular scale corresponds to cell types, whereas compositions on a millimeter scale correspond to tissue-level structures. Traditional techniques for identifying such regions often rely on complementary data, such as pre-segmented cells, or lengthy optimization. This limits their applicability to tasks on a particular scale, restricting their capabilities in exploratory analysis. This article introduces "Points2Regions," a computational tool for identifying regions with similar mRNA compositions. The tool's novelty lies in its rapid feature extraction by rasterizing points (representing mRNAs) onto a pyramidal grid and its efficient clustering using a combination of hierarchical andk $$ k $$ -means clustering. This enables fast and efficient region discovery across multiple scales without relying on additional data, making it a valuable resource for exploratory analysis. Points2Regions has demonstrated performance similar to state-of-the-art methods on two simulated datasets, without relying on segmented cells, while being several times faster. Experiments on real-world datasets show that regions identified by Points2Regions are similar to those identified in other studies, confirming that Points2Regions can be used to extract biologically relevant regions. The tool is shared as a Python package integrated into TissUUmaps and a Napari plugin, offering interactive clustering and visualization, significantly enhancing user experience in data exploration.

Complete spatially resolved gene expression is not necessary for identifying spatial domains

Spatially resolved transcriptomics (SRT) technologies have revolutionized the study of tissue organization. We introduce a graph convolutional network with an attention and positive emphasis mechanism, termed BINARY, relying exclusively on binarized SRT data to accurately delineate spatial domains. BINARY outperforms existing methods across various SRT data types while using significantly less input information. Our study suggests that precise gene expression quantification may not always be essential, inspiring further exploration of the broader applications of spatially resolved binarized gene expression data.

Laser axial scanning microdissection for high-efficiency dissection from uneven biological samples

Fast and efficient separation of target samples is crucial for the application of laser-assisted microdissection in the molecular biology research field. Herein, we developed a laser axial scanning microdissection (LASM) system with an 8.6 times extended depth of focus by using an electrically tunable lens. We showed that the ablation quality of silicon wafers at different depths became homogenous after using our system. More importantly, for those uneven biological tissue sections within a height difference of no more than 19.2 µm, we have demonstrated that the targets with a size of microns at arbitrary positions can be dissected efficiently without additional focusing and dissection operations. Besides, dissection experiments on various biological samples with different embedding methods, which were widely adopted in biological experiments, also have shown the feasibility of our system.

Padlock Probe-Based Targeted In Situ Sequencing: Overview of Methods and Applications

Elucidating spatiotemporal changes in gene expression has been an essential goal in studies of health, development, and disease. In the emerging field of spatially resolved transcriptomics, gene expression profiles are acquired with the tissue architecture maintained, sometimes at cellular resolution. This has allowed for the development of spatial cell atlases, studies of cell-cell interactions, and in situ cell typing. In this review, we focus on padlock probe-based in situ sequencing, which is a targeted spatially resolved transcriptomic method. We summarize recent methodological and computational tool developments and discuss key applications. We also discuss compatibility with other methods and integration with multiomic platforms for future applications.

Spatial tumour gene signature discriminates neoplastic from non-neoplastic compartments in colon cancer: unravelling predictive biomarkers for relapse

BACKGROUND

Opting for or against the administration of adjuvant chemotherapy in therapeutic management of stage II colon cancer remains challenging. Several studies report few survival benefits for patients treated with adjuvant therapy and additionally revealing potential side effects of overtreatment, including unnecessary exposure to chemotherapy-induced toxicities and reduced quality of life. Predictive biomarkers are urgently needed. We, therefore, hypothesise that the spatial tissue composition of relapsed and non-relapsed colon cancer stage II patients reveals relevant biomarkers.

METHODS

The spatial tissue composition of stage II colon cancer patients was examined by a novel spatial transcriptomics technology with sub-cellular resolution, namely in situ sequencing. A panel of 176 genes investigating specific cancer-associated processes such as apoptosis, proliferation, angiogenesis, stemness, oxidative stress, hypoxia, invasion and components of the tumour microenvironment was designed to examine differentially expressed genes in tissue of relapsed versus non-relapsed patients. Therefore, FFPE slides of 10 colon cancer stage II patients either classified as relapsed (5 patients) or non-relapsed (5 patients) were in situ sequenced and computationally analysed.

RESULTS

We identified a tumour gene signature that enables the subclassification of tissue into neoplastic and non-neoplastic compartments based on spatial expression patterns obtained through in situ sequencing. We developed a computational tool called Genes-To-Count (GTC), which automates the quantification of in situ signals, accurately mapping their position onto the spatial tissue map and automatically identifies neoplastic and non-neoplastic tissue compartments. The GTC tool was used to quantify gene expression of biological processes upregulated within the neoplastic tissue in comparison to non-neoplastic tissue and within relapsed versus non-relapsed stage II colon patients. Three differentially expressed genes (FGFR2, MMP11 and OTOP2) in the neoplastic tissue compartments of relapsed patients in comparison to non-relapsed patients were identified predicting recurrence in stage II colon cancer.

CONCLUSIONS

In depth spatial in situ sequencing showed potential to provide a deeper understanding of the underlying mechanisms involved in the recurrence of disease and revealed novel potential predictive biomarkers for disease relapse in colon cancer stage II patients. Our open-access GTC-tool allowed us to accurately capture the tumour compartment and quantify spatial gene expression in colon cancer tissue.

Rolling circle extension-assisted loop-mediated isothermal amplification (Rol-LAMP) method for locus-specific and visible detection of RNA N6-methyladenosine

N6-methyladenosine (m6A) is the most prevalent RNA modification in eukaryotic mRNAs. Currently available detection methods for locus-specific m6A marks rely on RT-qPCR, radioactive methods, or high-throughput sequencing. Here, we develop a non-qPCR, ultrasensitive, isothermal, and naked-eye visible method for m6A detection based on rolling circle amplification (RCA) and loop-mediated isothermal amplification (LAMP), named m6A-Rol-LAMP, to verify putative m6A sites in transcripts obtained from the high-throughput data. When padlock probes hybridize to the potential m6A sites on targets, they are converted to circular form by DNA ligase in the absence of m6A modification, while m6A modification hinders the sealing of padlock probes. Subsequently, Bst DNA polymerase-mediated RCA and LAMP allow the amplification of the circular padlock probe to achieve the locus-specific detection of m6A. Following optimization and validation, m6A-Rol-LAMP can ultra-sensitively and quantitatively determine the existence of m6A modification on a specific target site as low as 100 amol under isothermal conditions. Detections of m6A can be performed on rRNA, mRNA, lincRNA, lncRNA and pre-miRNA from biological samples with naked-eye observations after dye incubation. Together, we provide a powerful tool for locus-specific detection of m6A, which can simply, quickly, sensitively, specifically, and visually determine putative m6A modification on RNA.

TissUUmaps 3: Improvements in interactive visualization, exploration, and quality assessment of large-scale spatial omics data

Background and objectives

Spatially resolved techniques for exploring the molecular landscape of tissue samples, such as spatial transcriptomics, often result in millions of data points and images too large to view on a regular desktop computer, limiting the possibilities in visual interactive data exploration. TissUUmaps is a free, open-source browser-based tool for GPU-accelerated visualization and interactive exploration of 107+ data points overlaying tissue samples.

Methods

Herein we describe how TissUUmaps 3 provides instant multiresolution image viewing and can be customized, shared, and also integrated into Jupyter Notebooks. We introduce new modules where users can visualize markers and regions, explore spatial statistics, perform quantitative analyses of tissue morphology, and assess the quality of decoding in situ transcriptomics data.

Results

We show that thanks to targeted optimizations the time and cost associated with interactive data exploration were reduced, enabling TissUUmaps 3 to handle the scale of today's spatial transcriptomics methods.

Conclusion

TissUUmaps 3 provides significantly improved performance for large multiplex datasets as compared to previous versions. We envision TissUUmaps to contribute to broader dissemination and flexible sharing of largescale spatial omics data.

Deep learning in spatial transcriptomics: Learning from the next next-generation sequencing

Spatial transcriptomics (ST) technologies are rapidly becoming the extension of single-cell RNA sequencing (scRNAseq), holding the potential of profiling gene expression at a single-cell resolution while maintaining cellular compositions within a tissue. Having both expression profiles and tissue organization enables researchers to better understand cellular interactions and heterogeneity, providing insight into complex biological processes that would not be possible with traditional sequencing technologies. Data generated by ST technologies are inherently noisy, high-dimensional, sparse, and multi-modal (including histological images, count matrices, etc.), thus requiring specialized computational tools for accurate and robust analysis. However, many ST studies currently utilize traditional scRNAseq tools, which are inadequate for analyzing complex ST datasets. On the other hand, many of the existing ST-specific methods are built upon traditional statistical or machine learning frameworks, which have shown to be sub-optimal in many applications due to the scale, multi-modality, and limitations of spatially resolved data (such as spatial resolution, sensitivity, and gene coverage). Given these intricacies, researchers have developed deep learning (DL)-based models to alleviate ST-specific challenges. These methods include new state-of-the-art models in alignment, spatial reconstruction, and spatial clustering, among others. However, DL models for ST analysis are nascent and remain largely underexplored. In this review, we provide an overview of existing state-of-the-art tools for analyzing spatially resolved transcriptomics while delving deeper into the DL-based approaches. We discuss the new frontiers and the open questions in this field and highlight domains in which we anticipate transformational DL applications.

Scalable in situ single-cell profiling by electrophoretic capture of mRNA using EEL FISH

Methods to spatially profile the transcriptome are dominated by a trade-off between resolution and throughput. Here we develop a method named Enhanced ELectric Fluorescence in situ Hybridization (EEL FISH) that can rapidly process large tissue samples without compromising spatial resolution. By electrophoretically transferring RNA from a tissue section onto a capture surface, EEL speeds up data acquisition by reducing the amount of imaging needed, while ensuring that RNA molecules move straight down toward the surface, preserving single-cell resolution. We apply EEL on eight entire sagittal sections of the mouse brain and measure the expression patterns of up to 440 genes to reveal complex tissue organization. Moreover, EEL can be used to study challenging human samples by removing autofluorescent lipofuscin, enabling the spatial transcriptome of the human visual cortex to be visualized. We provide full hardware specifications, all protocols and complete software for instrument control, image processing, data analysis and visualization.

A guidebook of spatial transcriptomic technologies, data resources and analysis approaches

Advances in transcriptomic technologies have deepened our understanding of the cellular gene expression programs of multicellular organisms and provided a theoretical basis for disease diagnosis and therapy. However, both bulk and single-cell RNA sequencing approaches lose the spatial context of cells within the tissue microenvironment, and the development of spatial transcriptomics has made overall bias-free access to both transcriptional information and spatial information possible. Here, we elaborate development of spatial transcriptomic technologies to help researchers select the best-suited technology for their goals and integrate the vast amounts of data to facilitate data accessibility and availability. Then, we marshal various computational approaches to analyze spatial transcriptomic data for various purposes and describe the spatial multimodal omics and its potential for application in tumor tissue. Finally, we provide a detailed discussion and outlook of the spatial transcriptomic technologies, data resources and analysis approaches to guide current and future research on spatial transcriptomics.

The emerging landscape of spatial profiling technologies

Improved scale, multiplexing and resolution are establishing spatial nucleic acid and protein profiling methods as a major pillar for cellular atlas building of complex samples, from tissues to full organisms. Emerging methods yield omics measurements at resolutions covering the nano- to microscale, enabling the charting of cellular heterogeneity, complex tissue architectures and dynamic changes during development and disease. We present an overview of the developing landscape of in situ spatial genome, transcriptome and proteome technologies, exemplify their impact on cell biology and translational research, and discuss current challenges for their community-wide adoption. Among many transformative applications, we envision that spatial methods will map entire organs and enable next-generation pathology.

Spatial and temporal heterogeneity in the lineage progression of fine oligodendrocyte subtypes

Background

Oligodendrocytes are glial cells that support and insulate axons in the central nervous system through the production of myelin. Oligodendrocytes arise throughout embryonic and early postnatal development from oligodendrocyte precursor cells (OPCs), and recent work demonstrated that they are a transcriptional heterogeneous cell population, but the regional and functional implications of this heterogeneity are less clear. Here, we apply in situ sequencing (ISS) to simultaneously probe the expression of 124 marker genes of distinct oligodendrocyte populations, providing comprehensive maps of the corpus callosum, cingulate, motor, and somatosensory cortex in the brain, as well as gray matter (GM) and white matter (WM) regions in the spinal cord, at postnatal (P10), juvenile (P20), and young adult (P60) stages. We systematically compare the abundances of these populations and investigate the neighboring preference of distinct oligodendrocyte populations.

Results

We observed that oligodendrocyte lineage progression is more advanced in the juvenile spinal cord compared to the brain, corroborating with previous studies. We found myelination still ongoing in the adult corpus callosum while it was more advanced in the cortex. Interestingly, we also observed a lateral-to-medial gradient of oligodendrocyte lineage progression in the juvenile cortex, which could be linked to arealization, as well as a deep-to-superficial gradient with mature oligodendrocytes preferentially accumulating in the deeper layers of the cortex. The ISS experiments also exposed differences in abundances and population dynamics over time between GM and WM regions in the brain and spinal cord, indicating regional differences within GM and WM, and we found that neighboring preferences of some oligodendroglia populations are altered from the juvenile to the adult CNS.

Conclusions

Overall, our ISS experiments reveal spatial heterogeneity of oligodendrocyte lineage progression in the brain and spinal cord and uncover differences in the timing of oligodendrocyte differentiation and myelination, which could be relevant to further investigate functional heterogeneity of oligodendroglia, especially in the context of injury or disease.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-022-01325-z.

Comprehensive in situ mapping of human cortical transcriptomic cell types

The ability to spatially resolve the cellular architecture of human cortical cell types over informative areas is essential to understanding brain function. We combined in situ sequencing gene expression data and single-nucleus RNA-sequencing cell type definitions to spatially map cells in sections of the human cortex via probabilistic cell typing. We mapped and classified a total of 59,816 cells into all 75 previously defined subtypes to create a first spatial atlas of human cortical cells in their native position, their abundances and genetic signatures. We also examined the precise within- and across-layer distributions of all the cell types and provide a resource for the cell atlas community. The abundances and locations presented here could serve as a reference for further studies, that include human brain tissues and disease applications at the cell type level.

Matisse: a MATLAB-based analysis toolbox for in situ sequencing expression maps

BACKGROUND

A range of spatially resolved transcriptomic methods has recently emerged as a way to spatially characterize the molecular and cellular diversity of a tissue. As a consequence, an increasing number of computational techniques are developed to facilitate data analysis. There is also a need for versatile user friendly tools that can be used for a de novo exploration of datasets.

RESULTS

Here we present MATLAB-based Analysis toolbox for in situ sequencing (ISS) expression maps (Matisse). We demonstrate Matisse by characterizing the 2-dimensional spatial expression of 119 genes profiled in a mouse coronal section, exploring different levels of complexity. Additionally, in a comprehensive analysis, we further analyzed expression maps from a second technology, osmFISH, targeting a similar mouse brain region.

CONCLUSION

Matisse proves to be a valuable tool for initial exploration of in situ sequencing datasets. The wide set of tools integrated allows for simple analysis, using the position of individual reads, up to more complex clustering and dimensional reduction approaches, taking cellular content into account. The toolbox can be used to analyze one or several samples at a time, even from different spatial technologies, and it includes different segmentation approaches that can be useful in the analysis of spatially resolved transcriptomic datasets.