Abstract 5116: Low abundance protein enrichment by ProteoMiner coupled to iTRAQ-labeled LC-MS/MS for plasma-based quantitative proteomics

Background: Protein biomarkers for breast cancer are desired for early diagnosis, disease prognosis and drug response monitoring. Biomarkers in bodily fluids, such as plasma, allow for non-invasive monitoring and thus have additional value compared to tissue-based markers. Plasma-based biomarker discovery faces a challenge in that the wide dynamic range of protein concentrations prevents the detection of lower abundance proteins. Additionally, large sample sizes are needed for sufficient statistical power. In this setting, experimental design and normalization are important considerations. In this study, we have investigated the use of a novel protein enrichment strategy combined with isobaric label-based LC-MS/MS as well as two experimental designs for the identification of biomarkers of early stage breast cancer. Methods: Plasma from 12 patients with benign breast lesions and 12 with stage I breast cancer were processed using ProteoMiner enrichment followed by on-bead digestion. Two types of standards were investigated: a pooled standard, consisting of equal portions of the 24 plasma digests and a universal standard, consisting of 12 independent benign, stage I, stage II, and stage III breast cancer plasma samples. These designs have opposing strengths and weaknesses: pooled standards minimize variation but fix the sample size; universal standards allow for addition of new samples but introduce additional variation. The study required 4 iTRAQ-8plex sets which were subjected to 2-D LC separation followed by HPLC-Chip/Q-TOF analysis. Proteins were identified and quantified using Spectrum Mill software. Algorithms were developed for normalization, missing data imputation, protein grouping and identification of differential expression. Results: Use of ProteoMiner beads resulted in extraction of sufficient protein for at least 10 technical replicates and cut down preparation time by 80%, as compared to MARS-based immunodepletion. A total of 416 plasma proteins were identified, 96% of which are low abundance plasma proteins and 14 of which were differentially expressed. Expression values normalized using the pooled vs. universal standards were significantly correlated; however the use of the universal standard design yields a larger variation in expression values. Conclusion: This study demonstrated use of the ProteoMiner technology for enrichment of low abundance proteins from plasma. Fourteen plasma-based biomarkers of stage I breast cancer were identified with statistical significance. A number of these proteins (e.g. protocadherin FAT2, flightless-1 homolog) have been linked to breast cancer relevant processes, such as cell migration, adhesion, estrogen receptor signaling and proliferation. Data indicates that the universal standard is an adequate replacement for pooled standards with the caveat that higher variance should be expected. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 5116. doi:10.1158/1538-7445.AM2011-5116