Skip to main content

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

Hatton-Jones KM, Cox AJ, Barcelon J, West NP

Publication: Pathology (2024) • Vol 56 • Page 1028-1035PMID: 39227250View on PubMed

Purpose

This paper compared GeoMx digital spatial transcriptomics data quality control metrics in adjacent formalin-fixed, paraffin-embedded (FFPE) sections that were digested with 0.1, 0.5 or 1 µg/mL Proteinase K. One nasal polyp specimen was used to compare 0.1 and 1 µg/mL, two nasal polyp specimens were used to compare 0.5 and 1 µg/mL concentrations, and a single tonsil and pancreas specimen were used to compare 0.1, 0.5 and 1 µg/mL concentrations of Proteinase K.

Conclusion

When Proteinase K concentrations that were lower than suggested ((0.5 or 0.1 µg/mL Proteinase K versus 1 µg/mL) were used, NGS analysis resulted in fewer raw reads, negative probe counts, and quantile 3 counts of the 18,676 genes; a higher signal-to-noise ratio; and a greater number of genes detected that were above the limit of quantification (LOQ); although these differences were not found in all specimens. Importantly, while some significant differences between the 0.1 and 1 µg/mL Proteinase K concentrations were observed in the single nasal polyp specimen examined, differences between the 0.5 and 1 µg/mL concentrations were observed in only one of the two nasal polyps. Similarly, differences among the 0.1, 0.5 and 1 µg/mL Proteinase K concentrations were found in the tonsil specimen, but no differences were observed in the pancreas specimen. The authors state that potential differences in specimen processing, including fixation time, could contribute to the different effects that were observed among the specimens.

Studies (1)

Purpose

This study compared GeoMx digital spatial transcriptomics data quality control metrics in adjacent formalin-fixed, paraffin-embedded (FFPE) sections that were digested with 0.1, 0.5 or 1 µg/mL Proteinase K. The study used archival clinical specimens that included two FFPE blocks containing biopsies of two nasal mucosa polyps, an FFPE tonsil specimen and an FFPE pancreas specimen.  No further details on specimen collection and processing were provided. For each FFPE block, adjacent sections were processed using the GeoMx DSP manufacturer’s protocol with the following changes: a Leica Bond Rx system for deparaffinization was performed at 72°C and epitope retrieval was at 95°C for 15 min (pancreas) or 20 min (tonsil and nasal polyps); Proteinase K digestion was performed for 15 min at the experimental concentrations of 0.1 and 1 µg/mL (one nasal polyp, sections processed 1 month apart), 0.5 or 1 µg/mL (two nasal polyps), or 0.1, 0.5 and 1 µg/mL (tonsil and pancreas); hybridization to Whole Transcriptome Atlas probes was at 37°C for 18 h in a humidity-controlled chamber; sections were stained for PanCK (all specimens), CD-45 (tonsil and pancreas only), PGP9.5 (pancreas only) and SYTO13 (nuclear all specimens). Regions of interest (4 for tonsil and pancreas; 6, 7, or 11 for nasal polyps) containing >200 nuclei were photocleaved, and hybridization probes were sequenced on a NextSeq 2000 instrument. Sequencing data were analyzed using the GeoMx DSP Data Analysis Suite.

Summary of Findings

When nasal polyps were processed after a month of FFPE block storage using 0.1 µg/mL Proteinase K instead of the suggested 1 µg/mL concentration, there were fewer raw reads, negative probe counts, and quantile 3 counts of the 18,676 genes, but only the decrease in negative probe counts was significant (P=0.01). The decrease in negative probe counts led to a significantly higher signal-to-noise ratio (P=0.01) and a 6-fold increase in the number of genes detected that were above the limit of quantification (P<0.05). Using a 0.5 µg/mL instead of the suggested 1 µg/mL concentration of Proteinase K for nasal polyps led to significantly fewer raw reads (P=0.03), negative probe counts (P=0.03), and quantile 3 counts of the 18,676 genes (P=0.04), and a significantly higher signal-to-noise ratio (P=0.03) in one of the two nasal polyp specimens, but the number of genes detected that was above the limit of quantification was only non-significantly higher, and all metrics were comparable between these Proteinase K concentrations in the other specimen. In the tonsil specimen, decreasing the Proteinase K concentration from 1 µg/mL to 0.5 µg/mL or to 0.1 µg/mL significantly decreased the number of raw reads (P=0.04 and P=0.0008, respectively), negative probe counts (P<0.0001, both) and the quantile 3 probe counts (P<0.05 and P<0.0001, respectively) and increased the signal-to-noise ratio (P=0.001 and P<0.0001, respectively), and the number of genes detected above the LOQ was significantly higher when 0.1 µg/mL Proteinase K was used than when 1 µg/mL was used (P<0.01). Further, use of 0.1 µg/mL Proteinase K in tonsil specimens resulted in significantly lower negative probe counts (P<0.0001) and a higher signal-to-noise ratio (P<0.01) in comparison to 0.5 µg/mL Proteinase K. In contrast to tonsil and some nasal specimens, Proteinase K concentration did not affect any of the metrics in the pancreas specimen. The authors state that potential differences in specimen processing, including fixation time, could contribute to the different effects observed among the specimens.

  • Biospecimens

    TissueNasopharynx
    TissueTonsil
    TissuePancreas
  • Preservation Types

    Formalin
  • Diagnoses

    Not specified
    NeoplasticBenign
  • Platforms

    RNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen AcquisitionBiospecimen location
    Nasal mucosa
    Pancreas
    Tonsil
    Analyte Extraction and PurificationProtein digestion
    0.1 µg/mL Proteinase K
    0.5 µg/mL Proteinase K
    1.0 µg/mL Proteinase K