A streamlined mass spectrometry-based proteomics workflow for large-scale FFPE tissue analysis.
Author(s): Coscia F, Doll S, Bech JM, Schweizer L, Mund A, Lengyel E, Lindebjerg J, Madsen GI, Moreira JM, Mann M
Publication: J Pathol, 2020, Vol. 251, Page 100-112
PubMed ID: 32154592 PubMed Review Paper? No
Purpose of Paper
The paper compared different protein extraction methods (sodium dodecyl sulfate, SDS; filter-aided sample preparation, FASP; RapiGest-based extraction; 2, 2, 2-trifluoroethanol (TFE)-based extraction), laser-capture microdissected (LCM) and manually macrodissected FFPE tissue specimens, and macrodissection sampling approaches for potential effects on proteomic profiles using nanoflow LC-M/MS; potential effects associated with the duration of FFPE block storage were also investigated.
Conclusion of Paper
Protein digestion with TFE resulted in a peptide yield that was 2.3-fold higher than extraction with an SDS-based method. Protein digestion with TFE also recovered more identified peptides and proteins than extraction with RapiGest- or FASP-based methods. Notably, very strong proteome correlations were present among TFE-, FASP-, and RapiGest-based extraction methods (r=0.97-0.98), and the percentage of peptides that were lysine methylated did not differ among the protein extraction methods evaluated (4-6%). LCM and manually macrodissected areas yielded similar numbers of peptides and protein groups, and equivalent total ion currents and chromatographic signals. Principal component analysis revealed that clustering occurred by tissue type, not tumor enrichment method. Proteome correlations between LCM samples from the same patient that underwent protein extraction and purification with the same protocol were higher (r=0.94-0.97) than when samples collected from different patients that were processed by identical means were compared (r=0.86-0.88). The variability of the proteome between macrodissected areas of the same section was slightly higher than when the same area was sampled from a consecutive section (%CV of 20.1 versus 18.9, respectively).
FFPE blocks of colorectal adenoma specimens that were stored for 6-9 y had modestly higher numbers of identified peptides and proteins than those stored for 10-13 y (10% for peptides, 4% for proteins) or 14-20 y (by 16% for peptides, 6% for proteins). Blocks stored for the longest duration (14-20 y) also had a slightly higher peptide modification rate for both methionine oxidation (0.6%) and lysine methylation (0.9%) than those stored for 6-9 y. A small proportion (8%) of peptides that were uniquely identified among FFPE blocks stored for 6-9 y were considered to be “low abundance” relative to peptides that were identified across all block storage durations, a trend that was also present among proteins (6%). Notably, very strong proteome correlations were observed among all FFPE block storage durations (r=0.95-0.99). When a consensus clustering approach was applied, there were four main clusters, separated by biological variability, not the duration of block storage.
Studies
-
Study Purpose
This study compared different protein extraction methods (sodium dodecyl sulfate, SDS; filter-aided sample preparation, FASP; RapiGest-based extraction; 2, 2, 2-trifluoroethanol (TFE)-based extraction), laser-capture microdissected (LCM) and manually macrodissected FFPE tissue specimens, and macrodissection sampling approaches for potential effects on the proteome profile using nanoflow LC-M/MS. Macrodissected FFPE tissue specimens were obtained by manually scraping an area of a 10 µm-thick section from a glass slide with a razor blade after a pathologist reviewed the H&E stained section. A macrodissected area (5 x 5 mm area from a 10 µm-thick section) from a single high-grade serous ovarian tumor FFPE specimen was used to compare SDS and TFE protein extraction methods, whereas a single glioma FFPE specimen was used to compare TFE, FASP, and RapiGest extraction methods. Two high-grade serous ovarian cancer FFPE specimens were used to assess between- and within-patient variability. Four colorectal adenoma FFPE specimens were used to assess the reproducibility of microdissection sampling within a section and between sections. Additional details relating to biospecimen collection, preservation, and processing were not provided. For laser-capture microdissection, FFPE sections were mounted to a PEN membrane slide, and an area of approximately 1.5 x 1.5 mm was collected. LCM and macrodissected tissue were placed in individual PCR tubes for protein extraction and digestion. Nanoflow LC/MS-MS analysis was performed on a quadrupole Orbitrap mass spectrometer in conjunction with an EASY nLC 1200 ultra-high-pressure system.
Summary of Findings:
Protein digestion with TFE led to a peptide yield that was 2.3-fold higher than obtained with an SDS-based digestion method when macrodissected areas of a single serous ovarian tumor FFPE specimen were analyzed by nanoflow LC-MS/MS. Of the 5,041 proteins that were identified in total, there was a 92% overlap in proteins identified between the two protein digestion methods (TFE, SDS) and a very strong correlation between their proteomes (r=0.95). The number of peptides and proteins identified in a macrodissected glioma FFPE specimen was highest with TFE-based protein extraction, relative to extraction with RapiGest- or FASP-based methods. Notably, very strong proteome correlations were present among TFE-, FASP-, and RapiGest-based extraction methods (r=0.97-0.98). Further, each of the protein extraction methods evaluated yielded very similar proportions of proteins that originate from different cellular compartments (± 0.4% among ovarian and glioma comparisons). The percentage of peptides that were lysine methylated did not differ among the protein extraction methods evaluated (4-6%). The authors included supplemental data that indicated that acetonitrile (ACN, an organic solvent that is less hazardous than TFE) performed similarly to TFE, with similar peptides and protein yields, and a very strong proteome correlation (r=0.97).
A comparison of LCM and macrodissected tissue from ovarian, glioma, colorectal, and urachal carcinoma specimens revealed similar numbers of peptides and protein groups, and equivalent total ion currents and chromatographic signals. Principal component analysis revealed that clustering occurred by tissue type, not tumor enrichment method. A total of five LCM samples from two high-grade serous ovarian cancer specimens were used to assess between-patient, inter-sectional, and intra-sectional variability. As expected, correlations were stronger for comparisons between samples from the same patient (r=0.94-0.97) and modestly lower for comparisons between patients (r=0.86-0.88).
The variability of the proteome between macrodissected areas of the same section of a colorectal adenoma FFPE specimen was slightly higher than when the same area was sampled from a consecutive section (%CV of 20.1 versus 18.9, respectively); for context, the CV of technical replicates was 7.9%.
Biospecimens
Preservative Types
- Formalin
Diagnoses:
- Neoplastic - Carcinoma
- Neoplastic - Other
- Neoplastic - Benign
Platform:
Analyte Technology Platform Protein LC-MS or LC-MS/MS Peptide LC-MS or LC-MS/MS Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Analyte Extraction and Purification Analyte isolation method FASP
RapiGest
TFE
SDS
Biospecimen Aliquots and Components Cell capture method LCM
Macrodissection
Biospecimen Aliquots and Components Biospecimen heterogeneity Intratumoral sampling (exact positions not specified)
-
Study Purpose
The purpose of this study was to investigate the potential effects of prolonged FFPE block storage on proteomic analysis of colorectal adenoma specimens by nanoflow LC-MS/MS. A total of 118 colorectal adenoma FFPE specimens representing 101 patients were used for analysis. FFPE specimens were stored as FFPE blocks for 6-9 y (43 specimens), 10-13 y (41 specimens), or 14-20 (14 specimens). Additional details relating to biospecimen collection, preservation, processing, and FFPE block and slide storage conditions were not provided. A macrodissected area (5 x 5 mm) was collected from a 10 µm-thick section by scraping an area from a glass slide with a razor blade after a pathologist reviewed the H&E stained section. Specimens were then placed in a PCR tube prior to TFE-based protein extraction and digestion. Nanoflow LC/MS-MS analysis was performed on a quadrupole Orbitrap mass spectrometer in conjunction with an EASY nLC 1200 ultra-high-pressure system.
Summary of Findings:
FFPE blocks of colorectal adenoma specimens that were stored for 6-9 y had higher numbers of identified peptides and proteins than those stored for 10-13 y (10% for peptides, 4% for proteins) and 14-20 y (by 16% for peptides, 6% for proteins). Blocks stored for the longest duration (14-20 y) also had a slightly higher peptide modification rate for both methionine oxidation (0.6%) and lysine methylation (0.9%) than those stored for 6-9 y. Approximately 8% of the peptides that were uniquely identified among FFPE blocks stored for 6-9 y were considered to be “low abundance” relative to peptides that were identified across all block storage durations, a trend that was also present among proteins (6%). Notably, very strong proteome correlations were observed among all FFPE block storage durations (r=0.95-0.99). When a consensus clustering approach was applied, there were four main clusters, separated by biological variability, not the duration of block storage.
Biospecimens
Preservative Types
- Formalin
Diagnoses:
- Neoplastic - Benign
Platform:
Analyte Technology Platform Protein LC-MS or LC-MS/MS Peptide LC-MS or LC-MS/MS Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Storage Storage duration 6-9 y
10-13 y
14-20 y