Comparative analysis of differentially abundant proteins quantified by LC–MS/MS between flash frozen and laser microdissected OCT-embedded breast tumor samples

Lori A. Sturtz, Guisong Wang, Punit Shah, Richard Searfoss, Praveen Kumar Raj-Kumar, Jeffrey A. Hooke, J. Leigh Fantacone-Campbell, Brenda Deyarmin, Mary Lou Cutler, Rangaprasad Sarangarajan, Niven R. Narain, Hai Hu*, Michael A. Kiebish*, Albert J. Kovatich*, Craig D. Shriver*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Background: Proteomic studies are typically conducted using flash-frozen (FF) samples utilizing tandem mass spectrometry (MS). However, FF specimens are comprised of multiple cell types, making it difficult to ascertain the proteomic profiles of specific cells. Conversely, OCT-embedded (Optimal Cutting Temperature compound) specimens can undergo laser microdissection (LMD) to capture and study specific cell types separately from the cell mixture. In the current study, we compared proteomic data obtained from FF and OCT samples to determine if samples that are stored and processed differently produce comparable results. Methods: Proteins were extracted from FF and OCT-embedded invasive breast tumors from 5 female patients. FF specimens were lysed via homogenization (FF/HOM) while OCT-embedded specimens underwent LMD to collect only tumor cells (OCT/LMD-T) or both tumor and stromal cells (OCT/LMD-TS) followed by incubation at 37 °C. Proteins were extracted using the illustra triplePrep kit and then trypsin-digested, TMT-labeled, and processed by two-dimensional liquid chromatography-tandem mass spectrometry (2D LC–MS/MS). Proteins were identified and quantified with Proteome Discoverer v1.4 and comparative analyses performed to identify proteins that were significantly differentially expressed amongst the different processing methods. Results: Among the 4,950 proteins consistently quantified across all samples, 216 and 171 proteins were significantly differentially expressed (adjusted p-value < 0.05; |log2 FC|> 1) between FF/HOM vs. OCT/LMD-T and FF/HOM vs. OCT/LMD-TS, respectively, with most proteins being more highly abundant in the FF/HOM samples. PCA and unsupervised hierarchical clustering analysis with these 216 and 171 proteins were able to distinguish FF/HOM from OCT/LMD-T and OCT/LMD-TS samples, respectively. Similar analyses using significantly differentially enriched GO terms also discriminated FF/HOM from OCT/LMD samples. No significantly differentially expressed proteins were detected between the OCT/LMD-T and OCT/LMD-TS samples but trended differences were detected. Conclusions: The proteomic profiles of the OCT/LMD-TS samples were more similar to those from OCT/LMD-T samples than FF/HOM samples, suggesting a strong influence from the sample processing methods. These results indicate that in LC–MS/MS proteomic studies, FF/HOM samples exhibit different protein expression profiles from OCT/LMD samples and thus, results from these two different methods cannot be directly compared.

Original languageEnglish
Article number40
JournalClinical Proteomics
Volume17
Issue number1
DOIs
StatePublished - 1 Dec 2020
Externally publishedYes

Keywords

  • Breast tumor
  • Flash-frozen
  • Laser microdissection
  • Optimal cutting temperature compound
  • Proteomics
  • Stroma

Fingerprint

Dive into the research topics of 'Comparative analysis of differentially abundant proteins quantified by LC–MS/MS between flash frozen and laser microdissected OCT-embedded breast tumor samples'. Together they form a unique fingerprint.

Cite this