Tweet
Spatial biology is an exciting field that encompasses a wide range of techniques and technologies aimed at mapping the organization and interactions of various biomolecules in their native environments. As this area of research progresses, new tools and methodologies are being introduced, accompanied by efforts to establish benchmarking standards and drive technological innovation.
3D Genomics
While spatial biology often involves studying proteins and RNAs in their tissue or cellular contexts, it also includes 3D genomics, which investigates the arrangement of DNA inside the nucleus. Mathew Easterday, CEO of Dovetail Genomics, explained that understanding these genomic conformations helps reveal key processes associated with gene regulation and cellular differentiation, while also providing new tools for classifying cell types and studying their unique behaviors.
Dovetail Genomics provides tools for studying 3D genomics. Originally from Ed Green’s lab at UC Santa Cruz, the group began by addressing genome assembly challenges using their Hi-C-based technology. This approach involves fragmenting and re-ligating DNA to capture genomic organization by identifying DNA segments that are physically proximal in three-dimensional space. The first applications at the company were used to provide valuable data for genome assembly. However, the spatial conformation data was initially disregarded.
Over time, Dovetail transitioned to improved versions of Hi-C technology that retain data for 3D genomic structure, and enable numerous applications like genome phasing, assembly, and three-dimensional modeling. “We’ve come up with other versions of Hi-C that are easier, faster, and provide more uniform data,” noted Easterday. Recent developments include restriction enzyme-free protocols, such as the Dovetail® LinkPrep™ Kit. Easterday pointed out that this innovation ensures uniform genome coverage and simplifies workflows, reducing the time required for Hi-C experiments from a 2-3 days down to a single day.
One promising area for these tools is cancer research, where 3D genomic data is being used to identify structural variants associated with tumorigenesis. Hi-C data provides thousands of interaction points, offering evidence for genomic variations that might be missed with traditional sequencing methods. Similarly, the technology holds promise in immunology, particularly in organ transplantation. Dovetail Genomics is currently conducting experiments that focus on phasing the HLA region, a highly variable genomic region critical for immune compatibility. “We think we could play a role in better matching patients for organ donation,” stated Easterday. Current methods struggle to phase this region, but by incorporating Dovetail’s data, this is becomes feasible. The technology is soon to be in the hands of select HLA labs through a technology early access program created in partnership with CareDx, a leading precision medicine solutions company. While the potential impact on reducing organ rejection rates remains uncertain, early results suggest their approach could improve matching accuracy.
As for the future, Easterday expressed optimism about integrating diverse datasets into cohesive analytical frameworks, especially with the rise of AI technologies.“We're very quickly reaching a state where people are going to be able to build the tools that lead all these data to interact with one another,” he emphasized, noting the synergy between 3D genomic data and broader spatial datasets.Easterday also shared how Hi-C data could augment CRISPR-based genome editing by detecting off-target effects or unexpected conformational changes in edited regions. These advancements, Easterday believes, are pivotal for driving future innovation in research and technology applications.
Sequencing-Based Spatial Transcriptomics
Next-generation sequencing is integral to many transcriptomic methods in spatial biology, and its advancements have paved the way for newer, innovative spatial transcriptomics techniques. Recently, a team of researchers sought to democratize spatial transcriptomics by building their own scalable 3D method that was easy to use, high-resolution, and cost-efficient. This led to the development of an open-source experimental and computational resource known as Open-ST1. The system relies on Illumina’s sequencing platforms and converts their patterned flow cells into high-density, barcoded RNA capture areas with subcellular resolution.
The final workflow includes 3D-printable cutting guides, simplified library preparations, and scalability for up to 15 libraries in 3 days. It also requires lower sequencing depths than other platforms to receive comparable transcriptomic data, capturing ∼1,000 unique molecular identifiers (UMIs) per cell from ∼50,000 cells using 400 million reads. Furthermore, Open-ST is optimized for fresh-frozen samples, and its hematoxylin and eosin (H&E) imaging pipeline integrates high-resolution histology with transcriptomics for 2D and 3D analyses.
During benchmarking against published datasets, Open-ST demonstrated high sensitivity and subcellular precision. A proof-of-concept application with human head and neck squamous cell carcinoma profiled tumor heterogeneity and spatial cell-cell interactions across 21 sections. Using a 3D virtual tissue block, Open-ST identified transcriptional signatures like cholesterol biosynthesis and macrophages at tumor-lymphoid boundaries as potential 3D biomarkers. Despite being limited to polyadenylated transcripts and 3’ capture bias, this platform offers open-source protocols for broader applications, including integration with fixed tissues and advanced segmentation methods. Overall, Open-ST represents a promising tool for spatial omics studies that deliver high-resolution transcriptomic insights in diverse biological contexts.
Benchmarking Technologies
The availability of numerous sequencing-based spatial transcriptomics platforms has supported new areas of research. However, there has been a need to compare existing methods, address platform variability, establish reference standards for platform selection, and develop evaluation metrics for future analyses. Since each technology utilizes a distinct spatial indexing strategy, there are key differences in spatial resolution, molecule-capture efficiency, and RNA diffusion. To address these issues, a team of researchers compared 11 commonly used platforms2. The technologies assessed include microarray-based approaches (Visium and DynaSpatial), bead-based methods (Slide-seq V2, Curio Bio, BMKMANU S1000, Slide-tag, HDST), polony-based techniques (Stereo-seq, PIXEL-seq, Salus), and microfluidic platforms (DBiT-seq).
The researchers in this study first developed a library of reference tissues and regions featuring clearly defined histological structures and used them for data generation. Then they analyzed diverse aspects of the data, spanning basic metrics to advanced downstream analyses, including sensitivity, diffusion, clusterability, and marker gene detection. The study found that spatial transcriptomics requires more sequencing to achieve saturation, with current data falling below this threshold. Platforms like Stereo-seq, Slide-tag, and probe-based Visium demonstrated better capture efficiency at raw sequencing depths, while Slide-seq V2, probe-based Visium, and DynaSpatial performed better with normalized sequencing depths. Additionally, a surprising gene-capturing bias was observed on the polyA-based Visium platform, warranting further validation in other tissues.
This extensive benchmarking provides researchers with a framework for evaluating sequencing-based spatial technologies and established metrics for resolution, diffusion, and cell annotation. Although method-specific biases persist,this study is an important development toward reproducibility and standardization in spatial transcriptomics.
Looking Ahead
Spatial technologies are still in their early stages but are rapidly gaining recognition. Nature honored spatial proteomics as the 2024 Method of the Year, showcasing its significant contributions to advancing our understanding of biological complexity in health and disease3. Innovations such as deep visual proteomics, large-scale atlasing efforts, AI, and multiomic integrations promise even greater advancements and discoveries in the near future.
References
- Schott M, León-Periñán D, Splendiani E, et al. OpenST: High-resolution spatial transcriptomics in 3D. Cell. 2024;187(15):3953-3972.e26. doi:https://doi.org/10.1016/j.cell.2024.05.055
- You Y, Fu Y, Li L, et al. Systematic comparison of sequencing-based spatial transcriptomic methods. Nature Methods. 2024;21(9):1743-1754. doi:https://doi.org/10.1038/s41592-024-02325-3
- Method of the Year 2024: spatial proteomics. Nature Methods. 2024;21(12):2195-2196. doi:https://doi.org/10.1038/s41592-024-02565-3