The effect of Proteinase K treatment on GeoMx digital spatial profiling data quality from formalin-fixed, paraffin-embedded tissue

• Published in: Pathology 2024-12, Vol.56 (7), p.1028-1035
• Main Authors: Hatton-Jones, Kyle M., West, Nicholas P., Barcelon, Jean, Cox, Amanda J.
• Format: Article
• Language: English
• Subjects: Data Accuracy; digital spatial profiling; Endopeptidase K - metabolism; FFPE tissue; Formaldehyde; Gene Expression Profiling; High-Throughput Nucleotide Sequencing; Humans; Nasal Mucosa - metabolism; Nasal Mucosa - pathology; Palatine Tonsil - pathology; Pancreas - pathology; Paraffin Embedding; Spatial transcriptomics; Tissue Fixation - methods; Transcriptome
• ISSN: 0031-3025; 1465-3931; 1465-3931
• DOI: 10.1016/j.pathol.2024.06.004

The emergence of spatial profiling technologies in recent years has accelerated opportunities to profile in detail the molecular attributes of a wide range of tissue pathologies using archival specimens. However, tissue treatment for fixation and storage does not always support generation of high-quality genomic data. The purpose of this study was to investigate the impacts of Proteinase K (ProtK) treatment, as a way to increase target transcript exposure, on downstream sequencing data quality metrics for spatial transcriptomic data using formalin-fixed, paraffin-embedded samples. In a series of four independent assessments using different tissue types (nasal mucosa, tonsil, pancreas), varying concentrations of ProtK (ranging from 0.1 to 1 μg/mL) were used as part of the sample processing workflow to generate transcriptomic data using the Nanostring GeoMx DSP and Illumina NextSeq 2000 platforms. Use of higher concentrations of ProtK was generally found to increase total reads (2–4-fold). However, negative probe counts also tended to be increased (2–12-fold), resulting in reductions in the signal-to-noise ratio (10–70% lower) and the number of genes detected above background (50–80% lower). These effects were not seen in all tissues and impacts of tissue handling and processing, beyond ProtK treatment, on data quality metrics, also require consideration. Regardless, these observations highlight the need for careful consideration of a range of sample processing factors and benefits that may be achieved through the optimisation of sample processing workflows for specific tissues as a way to maximise the generation of quality data using spatial transcriptomic approaches.