Background. Only a futuristic vision in the era of bulk RNA sequencing, single-cell transcriptomics (scT) has revolutionized the analysis of cellular heterogeneity as it allows the development of comprehensive reference maps based on reconstruction of cellular developmental pathways and modelling of transcriptional dynamics.
While scT helps to gain insights into various cellular functions, the currently used methods lack spatial information and alter cell type proportions, thereby removing critical information essential for the understanding of the cellular microenvironment, intercellular communication and their influence on tissue architecture and function.
Challenge. Current spatial transcriptomics (ST) methods, developed to overcome these challenges, are based on imaging and sequencing techniques, or a combination thereof. However, depending on the research question, several parameters such as spatial resolution, detection limit, screening area, accessibility, compatibility with existing workflows and costs, have to be balanced against each other to select the most suitable method for a given experiment. One solution to this dilemma was the previous development of Deterministic barcoding in tissue (DBiT-seq), which uses microfluidic channels to barcode tissue sections by DNA oligonucleotides and allows the integration of multi-omics information. Despite being a major step forward, DBiT-seq does however not allow for higher throughput applications. Therefore, Johannes Wirth, Ph.D. student in the laboratory of Matthias Meier at the Helmholtz Pioneer Campus developed an advanced workflow providing a multiplexing method for ST paired with high-quality imaging (called xDBiT), thereby significantly advancing the previously reported DBiT-seq method.
Solution. On the technical side, he changed the chemical composition of the initial reaction protocols and implemented a serpentine channel design allowing to analyze nine tissue sections in parallel. Through various validation steps in several tissues, Johannes demonstrated that there is no leakage between individual wells and only neglectable cross-contaminations from neighboring samples. Most importantly, compared to the original method, these optimizations increased the reads and genes per spot by 6 and 12-fold, respectively. In addition, the read depth within xDBiT in combination with deconvolution tools, turned out to be sufficient to resolve even rare cell types, such as podocytes in the glomeruli of the kidney. On top of the technical advances, the team also provided an open-source analysis pipeline to generate xDBiT datasets and make the method easily accessible.
Conclusion. xDBiT enables multiplexing, increases sequencing depth, and improves the image data quality – thus, providing a robust and accurate analysis tool for spatial gene expression patterns. As xDBiT is suitable for a variety of tissues, it will facilitate studies on complex diseases and multi-organ (dys-)function.
Overall, xDBiT advances the toolbox of spatial transcriptomic methods towards higher throughput measurements and bears the potential for a wide range of applications in biological and clinical research.