Reproducible proteomics sample preparation for single FFPE tissue slices using acid-labile surfactant and direct trypsinization

Melanie Christine Föll, Matthias Fahrner, Victor Oginga Oria, Markus Kühs, Martin L. Biniossek, Martin Werner, Peter Bronsert, Oliver SchillingClinical 2018 · DOI

TL;DR

This paper introduces and validates a streamlined proteomics sample preparation method (Direct Tissue Trypsinization, DTR) for formalin-fixed, paraffin-embedded (FFPE) tissues. The DTR protocol offers comparable proteomic data quality and reproducibility to the established Filter-Aided Sample Preparation (FASP) method, while being faster and requiring less sample manipulation, making it ideal for precious archival tissues.

What Problem Does This Paper Address?

Clinical research heavily relies on analyzing human tissues to understand disease mechanisms and discover biomarkers. Formalin-fixed, paraffin-embedded (FFPE) tissues are the standard for long-term preservation in pathology archives, offering a rich source of historical patient data. However, the crosslinking caused by formalin fixation makes these tissues challenging for proteomic analysis. Traditional methods like Filter-Aided Sample Preparation (FASP) are effective but involve multiple centrifugation steps, leading to potential sample loss and increased processing time, which is a significant drawback when dealing with limited or precious FFPE samples, such as single tissue slices or those obtained from minimally invasive biopsies. This paper addresses the need for a more efficient and robust sample preparation method for FFPE tissues that maintains high data quality and reproducibility.

Methods

The study compares a novel Direct Tissue Trypsinization (DTR) protocol with the established Filter-Aided Sample Preparation (FASP) method.

  • DTR Protocol: This method utilizes an acid-labile surfactant, RapiGest, in the lysis buffer. RapiGest aids in protein extraction and solubilization without interfering with subsequent trypsin digestion. The key advantage is that it bypasses the need for SDS removal steps, directly proceeding from lysis and heat-induced antigen retrieval (HIAR) to trypsinization. RapiGest is later removed by acidification.
  • FASP Protocol: This standard method uses a buffer containing Sodium Dodecyl Sulfate (SDS) for lysis and HIAR. SDS is a strong detergent but inhibits trypsin digestion, necessitating its removal using centrifugal filter units (either 10 kDa or 30 kDa cut-off) before digestion. This multi-step process involves several centrifugation and washing steps.

The researchers also investigated the compatibility of the DTR protocol with common histological stainings (Hematoxylin, H&E, Hemalaun) to assess if these pre-analytical steps impact proteomic analysis.

Datasets

The study utilized:

  • Murine kidney FFPE tissue: Four single, deparaffinized tissue slices were used for each of the three protocols (DTR, FASP 10k, FASP 30k) to compare their performance.
  • Human tonsil FFPE tissue: Four single slices from each of two patients were used for each staining condition (unstained, Hematoxylin, H&E, Hemalaun) to assess staining compatibility.

No new benchmark datasets were introduced; rather, existing FFPE tissue samples were used to evaluate the protocols.

Key Insights and Ideas

  • Efficiency Matters: The DTR protocol is significantly faster and requires fewer manual steps than FASP, making it more amenable to high-throughput analysis and precious samples.
  • Surfactant Choice is Key: The use of RapiGest in DTR circumvents the need for SDS removal, a major bottleneck and source of sample loss in FASP.
  • Staining Doesn't Have to Be a Showstopper: Histological staining, a common practice for FFPE tissues, does not fundamentally hinder DTR-based proteomics. While H&E staining caused a slight decrease in protein identification, Hemalaun and Hematoxylin showed minimal impact.
  • Reproducibility is Comparable: Despite a slight increase in missed tryptic cleavages with DTR, the overall quantitative and qualitative reproducibility of the DTR protocol is on par with FASP.

Key Results

  • Protein Identification: The DTR protocol identified an average of 1841 proteins per sample, comparable to FASP 10k (1857) and FASP 30k (1970).
  • Reproducibility: Pearson correlation coefficients for label-free quantitation (LFQ) intensities were high across all protocols (DTR: 0.98-0.99, FASP 10k: 0.96-0.98, FASP 30k: 0.99).
  • Missed Cleavages: DTR showed a reproducible 15% increase in missed tryptic cleavages compared to FASP, but this did not negatively impact quantitation.
  • Staining Compatibility: Unstained, Hematoxylin, and Hemalaun stained tissues yielded comparable proteomic results with DTR. H&E staining resulted in a ~13% decrease in protein identification.
  • Physicochemical Properties: Proteins identified by DTR and FASP showed similar distributions in terms of molecular weight, isoelectric point, and hydrophobicity, indicating no major bias introduced by the different detergents.

Limitations and Future Work

  • Missed Cleavages: While not impacting quantitation in this study, the increased rate of missed cleavages in DTR could potentially affect downstream analyses that rely on precise peptide identification or specific cleavage sites.
  • Tissue Amount: The study used relatively large tissue slices (0.47-0.6 mm³) to ensure compatibility with both protocols. Further optimization might be needed for extremely small samples (e.g., single cells or very small biopsies).
  • Staining Optimization: While compatible, the study suggests performing a small pilot experiment when using different tissues and staining methods to find optimal conditions. The impact of shortened staining times was not explored in detail.
  • Future Work: The authors highlight the suitability of DTR for small FFPE tissues and its potential as a viable alternative to FASP for clinical research, particularly for biomarker discovery and retrospective studies. Further validation on diverse FFPE tissue types and clinical samples would be beneficial.

Summary updated

Abstract

In essence, the DTR protocol can keep up with the FASP protocol in terms of qualitative and quantitative reproducibility and performed almost as well in terms of proteome coverage and missed cleavages. We highlight the suitability of the DTR protocol as a viable and straightforward alternative to the FASP protocol for proteomics-based clinical research.
BibTeX
@inproceedings{foll2018reproducible,
  author = {Melanie Christine Föll and Matthias Fahrner and Victor Oginga Oria and Markus Kühs and Martin L. Biniossek and Martin Werner and Peter Bronsert and Oliver Schilling},
  title = {{Reproducible proteomics sample preparation for single FFPE tissue slices using acid-labile surfactant and direct trypsinization}},
  booktitle = {Clinical},
  year = {2018},
  doi = {10.1186/s12014-018-9188-y},
}
Figure