Your roadmap to spatial: Exploring the power of spatial transcriptomics
In the Roadmap to Spatial blog series, we provide an educational overview of all things spatial biology, a growing field of research that seeks to better understand molecular and cellular events in the context of anatomy, morphology, and tissue architecture. In Part One, leading scientists share why biology needs a spatial readout and what is now possible with this technology. Read on to explore key publications demonstrating the power of spatial transcriptomics, a method for spatially resolving gene expression patterns in tissue.
Why biology needs a spatial readout
“Almost everything in biology is structure—how the parts are arranged in space, how they fit together.”
George Church, PhD, Professor, Harvard Medical School (Comments from Xperience 2022)
Biology is more than the sum of its parts. Complex molecular pathways take effect in cells that are part of a larger tissue ecosystem. The composition and organization of endogenous cell populations determine the health and proper function of that tissue, as well as impact the outcome of disease. Simultaneously, the architecture of the tissue plays an important role in enabling or impeding cellular interactions and activity.
Thus, to know the parts of a tissue (the cells), but not the larger structure, would be to miss a key relationship between cell identity, localization, and function in the tissue ecosystem—much like trying to piece together a masterful painting by only knowing the colors the painter used.
In order to gain a more comprehensive understanding of tissue complexity, scientists have begun to use various spatial technologies to survey the tissue landscape at high molecular resolution. These tools can typically provide access to spatially resolved RNA content from the cells in the profiled tissue section, and a number of commercial platforms are in active development to incorporate additional analytes such as protein expression.
There are many possibilities with this approach. For example, spatial context can provide insights into cell–cell communication and strategic localization within a structure. This expands on a readout of the molecular content of individual cells, such as that provided by cutting-edge single cell technologies, by tracing cells back to where they’re found in the natural tissue.
“The in situ methods help to coordinate or to put the cells in their context… That will greatly increase your ability to interpret your single cell data.”
Mats Nilsson, PhD, Professor, Principle Investigator, Stockholm University
Dr. Nilsson of Stockholm University, a leading expert in spatial biology, also sees the power of in situ methods in their ability to provide a view of “the topography of cell architecture and the cellular content [in] highly irreproducible tissue structures, such as tumors.”
Indeed, spatial technologies have great potential to reveal how complex and precious tissues work in both health and disease, providing new opportunities for basic discovery and translational research into tissue-resident diseases, such as cancer or infectious disease. Leading researchers see the same potential in spatial techniques:
“Some of the interesting areas that we are using spatial genomics, particularly in human development, is in whole embryo analysis, whereby we can begin to take coronal sections of this precious material and actually profile the gene expression… [We’re] able to locate all of those cells that make up a human embryo in situ. This will be the first complete atlas.”
Muzlifah Haniffa, MD, PhD, Professor, Principle Investigator, Wellcome Sanger Institute
“What still drives my excitement in this field, and motivates me every day to push the boundaries further, is that going spatial, going to tissue-level analysis, will give us a lot more powerful information to try to understand normal and disease conditions…[For example], what I see in front of me is a full body model of a healthy human person. And we would have a detailed understanding of every tissue—how they work normally…Then, [we could] extract features from diseased individuals, persons with some sort of disease of a tissue. You analyze it. And you map it back to the model…And you would see all the things that are deviating.”
Mats Nilsson, PhD, Professor, Principle Investigator, Stockholm University
“[In situ methods] will allow us to identify all the parts and start addressing pathologies, how basic cellular functions go awry in infections…[Or], say, between neurons, where thousands of connections might form between one neuron and its neighbors—distant neighbors, sometimes, centimeters away—that delicate fabric really cries out for a three-dimensional structure. And when it goes awry in neuropsychiatric disorders, this will help us get inside and make changes.”
George Church, PhD, Professor, Harvard Medical School, Founding Core Faculty & Lead, Wyss Institute
Discovering what’s possible with spatial biology
The spatial movement is, in some ways, well underway, and just beginning in others. Scientists have already made important advances in many fields of research using spatial technology, from characterizing the cellular and molecular events that drive healthy development, to disease pathogenesis and therapeutic response. But they are also discovering new applications and possibilities every day with innovative spatial tools that provide access to more biological analytes while maintaining morphological context. [You can explore some of these innovations here, including spatial T-cell receptor sequencing, spatial CRISPR, and, most recently, spatial mapping of viral RNA.]
And, in the following sections, explore a handful of impactful studies conducted using spatial gene expression, also known as spatial transcriptomics, which allows researchers to combine microscopy and molecular biology to map the transcriptome in tissue sections.
Location matters for glioblastoma heterogeneity
Of all organs in the body, the brain may best demonstrate the close relationship between spatial organization of tissue and function. The same is true for cancers affecting the central nervous system. In this study, researchers sought to characterize the influence of the local microenvironment on glioblastoma tumorigenesis, as well as the functional and spatial organization of malignant tumors (1). They first generated an atlas of spatially resolved transcriptomics for twenty-eight glioblastoma patient samples and healthy controls. Integrating these results with spatially resolved proteomics and metabolomics analysis, they identified five distinct regional transcriptional programs shared across a majority of patient samples.
Two of these programs were associated with high expression of glial-related genes: the first showed increased expression of radial-glia-associated genes, while the other showed enrichment of inflammation-associated genes and a significant increase in infiltrating myeloid and lymphoid cells likely involved in immunosuppression. Two other programs were distinguished by their neural or oligodendrocytic cellular origins. The fifth spatially distinct transcriptional program was uniquely characterized by its association with hypoxia-response and glycolytic genes, suggesting metabolic changes and oxygen concentration drive a distinct transcriptional state. Moreover, genomic data revealed increased copy number alterations in cells from this fifth region, providing evidence for increased genomic instability under hypoxia conditions. Taken together, these detailed results offer a framework to understand inter-patient glioblastoma heterogeneity as well as variable therapeutic responses.
Growing a heart
In a developing organism, the smallest molecular events dictate magnificent physiological changes over time. The challenge of developmental biology is how to parse out those molecular events and link them to the layers of biology they influence—from gene expression changes in specific cell types to the tissue-specific spatial gene expression patterns that regulate organogenesis. One team of researchers set out to understand this process for human cardiac development, combining single cell RNA-sequencing (scRNA-seq), spatial gene expression, and in situ sequencing to build a three-dimensional organ-wide atlas of the cells in the human heart at three different time points during development (2).
Their analysis enabled a rigorous molecular annotation of the anatomic regions of the developing heart, pointing to a new paradigm for heart development where regional gene expression differences within the organ are more pronounced than those between developmental time points. Additionally, their multimodal data informed a clearer annotation of regional and temporal cellular heterogeneity in the heart. For example, while two cardiomyocyte subpopulations were known to localize to the atria and ventricles, respectively, the team identified a previously uncharacterized third cardiomyocyte subpopulation that localized to both regions. [Explore their data to learn more.]
An invisible predisposition drives skin inflammation in lupus
Autoimmune conditions remain challenging to diagnose and treat, due to diverse clinical manifestations and limited knowledge of the disease-specific cellular and molecular mechanisms driving pathogenesis. Cutaneous lupus erythematosus (CLE) is an inflammatory skin disease associated with systemic lupus erythematosus (SLE). While 50% of SLE patients respond to systemic directed therapies, many cases show refractory emergence of CLE even when systemic disease is controlled (3).
To better understand the inflammatory composition of CLE, researchers from the University of Michigan performed both scRNA-seq and spatial gene expression analysis of lesional and nonlesional, or “normal-appearing,” skin samples from patients with active CLE and healthy controls. This revealed that normal-appearing skin in lupus patients was actually a type I interferon–rich, prelesional environment. The researchers also observed significant accumulation of CD16+ dendritic cells in both lesional and prelesional CLE environments, suggesting a central role for this cell population to drive cell–cell communication with local innate immune cells and ultimately promote tissue damage and CLE pathogenesis. [Read the publication here.]
These exciting projects represent just a small sample of the impactful and diverse research that’s being done with spatial technology. And they demonstrate the great potential of spatial information to advance both basic biological discovery and crucial clinical translational applications.
More than one path to spatial
The case is clear—biology needs a spatial readout to unravel the complex tissue systems that provide a framework for the cellular and molecular processes vital to our health and dysregulated in disease. Yet, it may not be so clear what path to spatial is the right one. And truly, there isn’t a “right” path, but there are many, and you’ll want to be informed as you begin to explore the spatial tools landscape. What tools are available and how do they work? How can you go about choosing the right spatial solution for your research projects?
We’ll explore these questions in more detail in Part Two of our Roadmap to Spatial blog series, providing a brief history of spatial technological development, and introducing some of the options for spatial technology available to you today. Stay tuned!
References:
- Ravi VM, et al. Spatially resolved multi-omics deciphers bidirectional tumor-host interdependence in glioblastoma. Cancer Cell 40: 639–655.e13 (2022). doi: 10.1016/j.ccell.2022.05.009
- Asp M, et al. A spatiotemporal organ-wide gene expression and cell atlas of the developing human heart. Cell 179: 1647–1660 (2019). doi: 10.1016/j.cell.2019.11.025
- Billi A, et al. Nonlesional lupus skin contributes to inflammatory education of myeloid cells and primes for cutaneous inflammation. Sci Transl Med 14: eabn2263 (2022). doi: 10.1126/scitranslmed.abn2263