AACR: Analytical validation of the multi-nanoparticle Proteograph platform for rapid and deep proteomic profiling | Seer Posters Archives - Seer

AACR Virtual Meeting II – American Association for Cancer Research –  June 2020

Authors

John E Blume1, William C Manning1, Gregory Troiano1, Asim Siddiqui1, Philip Ma1, Robert Langer2, Vivek Farias2, Omid C. Farokhzad1

1Seer, Redwood City, CA, United States
2Massachusetts Institute of Technology, Boston, MA, United States

 

Introduction: Plasma proteins should be useful biomarkers for disease detection, yet few proteins (~120) are FDA approved. Productive biomarker discovery studies are resource-limited due to the complex biochemical fractionation methods used to address the inherent challenges in unbiased plasma profiling such as the large dynamic range. Herein, we describe Proteograph, a novel platform that leverages the nano-bio interactions of nanoparticles (NPs) for deep and unbiased proteomic sampling. NPs reproducibly bind subsets of proteins from biofluids as a function of protein concentration, protein-NP affinity, and protein-protein interactions to form a corona on the NP surface. The corona composition is directly a function of the NPs’ biophysicochemical properties and requires no prior knowledge of the proteins that might be selected by each distinctly engineered NP. With an optimized panel of 10 NPs, we can broadly and deeply interrogate the plasma proteome and rapidly quantify potential biomarkers. The highly parallel NP workflow makes large studies practicable and should improve biomarker discovery and validation.

Methods and Results: We have screened 200+ NPs with different biophysicochemical properties and selected a panel of 10 based on protein detection. A fully automated assay workflow was developed that can process the 10 NP panel across 8 plasma samples in a 7 hr assay. Evaluation of this panel across 16 individual plasma samples detected 2,009 protein groups (1% protein FDR, 84% with 2 or more peptides). Accuracy was demonstrated with spike-recovery experiments using 4 NPs in which CRP, Angiogenin and S100a8/9 were added to plasma at 2X, 5X, 10X, and 100X of measured endogenous levels. Linear model fits for NP corona MS signal vs. ELISA were created with mean slopes of 1.06 ± 0.22 and mean adjusted-rof 0.95 ± 0.05. Precision was demonstrated across 3 NPs using three assay replicates in which the mean of the median CVs for each NP is 24%. The depth of plasma proteome coverage for the 10 NP panel using a pooled plasma sample was determined by comparison of the NP-detected proteins to published MS intensities and spanned nearly the entire reported range. Examining protein annotations (e.g., GO Cellular Compartment and Biological Process, KEGG and Pfam) within each NP corona reveals correlations by 1D-enrichment analysis between protein annotations and NP biophysicochemical properties suggesting specific relationships at the nano-bio surface.

Discussion: We have demonstrated the selection and optimization of a panel of 10 NPs for plasma proteome profiling. We have also demonstrated the breadth and depth of this panel’s ability to accurately and precisely quantify proteins from plasma. We believe the robustness and scalability of the Proteograph platform could enable population-scale deep and unbiased proteomics analysis previously not feasible using existing workflows.

 

View Poster

 

AACR Virtual Meeting II – American Association for Cancer Research –  June 2020

Authors

John E Blume1, William C Manning1, Gregory Troiano1, Asim Siddiqui1, Philip Ma1, Robert Langer2, Vivek Farias2, Omid C. Farokhzad1

1Seer, Redwood City, CA, United States
2Massachusetts Institute of Technology, Boston, MA, United States

 

Introduction: Though early detection of NSCLC greatly improves prognosis, we lack useful clinical tests. Genomics approaches utilizing cell-free DNA provide suitable specificity but moderate sensitivity for early cancer detection. Plasma proteins have the potential to deliver robust panels of biomarkers for early cancer detection that may be complimentary to genomics markers. Complex workflows, which enable deep and unbiased interrogation of plasma proteins that span 10 orders of magnitude, have made it impractical to efficiently perform robust studies, and consequently, comprehensive proteomic data vastly lags other “omics”.

Herein, we report a multi-NP Proteograph platform that rapidly, reproducibly, deeply, and scalably interrogates proteins from biofluids. In a study of 268 subjects, comparing on average 1779 plasma proteins of NSCLC subjects to healthy and pulmonary co-morbid controls, we identified classification panels comprising proteins with known and unknown roles in NSCLC, offering the promise of new biomarkers for early disease detection.

Methods: Subject plasma samples were grouped into NSCLC stages 1,2,3 (early), NSCLC stage 4 (late), or healthy and pulmonary co-morbid controls, for a randomly selected cohort of 288 age- and gender-matched subjects, and interrogated with a panel of NPs in an efficient automated work-flow. Peptides from NP-bound proteins underwent data-independent-acquisition mass spectrometry. Subject samples were also interrogated using conventional Agilent MARS-14 immunodepletion column, which has historically yielded limited clinical value, to determine differences in depth and types protein coverage achieved as compared with panel of NPs.

Results: On average 1,779 proteins were detected from each of the 268 subject samples vs. 413 from depleted plasma. The healthy vs early NSCLC random classification after depleted plasma protein removal achieved an average AUC of 0.90. Classification of healthy subjects to late NSCLC had an average AUC of 0.98. Comparison of the top features of the NSCLC classifiers to the co-morbid classifier indicated clinically significant differences. Among the former were proteins with both known and unknown roles in NSCLC (OpenTargets), underscoring the value of unbiased proteomic analysis.

Conclusions: We demonstrate the utility of the multi-NP Proteograph platform to deeply profile plasma proteins as novel biomarkers. The performance of the healthy vs. early NSCLC classifier confirms the potential of proteins in early disease detection. Our platform enables deep unbiased plasma protein biomarker profiling that matches genomics workflow throughput and suggests feasibility of parallel large-scale complementary studies of proteins and nucleic acids

 

View Poster

AACR Virtual Meeting II – American Association for Cancer Research –  June 2020

Authors

Asim Siddiqui1, John E Blume1, William C Manning1, Gregory Troiano1, Philip Ma1, Robert Langer2, Vivek Farias2, Omid C. Farokhzad1

1Seer, Redwood City, CA, United States
2Massachusetts Institute of Technology, Boston, MA, United States

 

Introduction: Understanding changes in PPI maps from a healthy and diseased state can illuminate our understanding of biological changes and disease processes. PPI maps enable a higher order of information than a simple listing of components by providing functional context, yet existing maps grossly underrepresent the total biological information potential of PPIs. Herein, we describe Proteograph, a novel platform that leverages the nano-bio interactions of nanoparticles (NPs) for deep and unbiased proteomic sampling that can provide insights on PPI across biological samples. Proteograph leverages the protein corona that forms on the surface of NPs as a function of their distinct biophysicochemical properties. NPs reproducibly bind subsets of proteins from biofluids as a function of protein concentration, protein-NP affinity, and protein-protein interactions to form a corona on the NP surface. We have employed Proteograph to quantify known PPIs using a panel of 3 distinct NPs to capture plasma proteins and derive maps of NSCLC and control subjects in order to identify biological changes in interactions, potentially indicative of health and disease.

Method and Results: We collected plasma samples from 288 subjects: healthy (n=82), comorbid (n=81) and NSCLC stages I-IV (n=125). In this initial study, we used three NPs with distinct properties and evaluated the protein corona of plasma samples by mass spectrometry (MS) to quantify 1,235 protein groups (1% FDR). A fully automated assay workflow enabled preparation of 3 NPs’ corona for MS analysis across 288 subjects in approximately 6 days. We mapped the protein groups to a PPI map derived from the STRING database. Partitioning the network into clusters identified 9 interaction clusters with greater than 10 protein members. These clusters enabled us to investigate differences in the PPI networks between NSCLC patients vs. controls. Evaluating the expression of proteins in these groups, we identified interaction clusters that had significant differences between cancer vs. control (t-test, p < 0.01 Bonferroni corrected). Six of the clusters show differential behavior between NSCLC vs. healthy controls (p < 0.01). Two of these clusters show differential behavior between NSCLC vs. healthy and comorbid (p < 0.01). Investigation of these differentially expressed clusters reveals links to known cancer biology with proteins related to the immune system and endocytosis pathways.

Discussion: We have used the Proteograph platform to identify PPI clusters that are differentiated between NSCLC and control individuals. We believe the efficiency of the Proteograph platform applied to sufficiently powered studies may enable comprehensive understanding of known PPIs, and potentially infer and confirm new PPIs, in health and disease.

 

View Poster

ASMS Reboot – 68th Conference on Mass Spectrometry and Allied Topics– June 2020

Authors

John E. Blume1*; Shadi Ferdosi1; Daniel Hornburg1; Matthew E. K. Chang2; Philip C. M. Ma1; Omid C. Farokhzad1; Mark R. Flory2; Patrick A. Everley1

¹Seer, Inc., Redwood City, CA – *jblume@seer.bio
²OHSU – Knight Cancer Institute- Cancer Early Detection Advanced Research Center (CEDAR), Portland, OR

 

Summary

Plasma proteins should be useful biomarkers for disease detection, yet only ~120 are FDA approved. Productive biomarker discovery studies are resource-limited requiring complex workflows that operate with trade-offs in throughput, scalability, coverage, and precision.  Thus, there is a need for new technological solutions for large-scale proteomic studies. Herein, we describe Proteograph, a novel platform leveraging the nano-bio interactions of nanoparticles and their distinct biophysicochemical properties in forming protein coronas for deep and unbiased proteomic sampling. We compare our platform to alternative methods for unbiased proteomics profiling. On all measures, including speed, coverage, and precision, Proteograph is superior, for the first time putting unbiased plasma profiling at the required throughput for large-scale clinical studies and biomarker discovery efforts. Proteograph is the first unbiased platform for large-scale proteome profiling studies quantifying thousands of proteins across large numbers of samples.

 

View Poster

US HUPO 16th Annual Conference – April 2020

 

Authors

Patrick A. Everley¹; Shadi Ferdosi¹; Daniel Hornburg¹; William C. Manning¹;  Asim Siddiqui¹; Greg Troiano¹; Omid C. Farokhzad¹; Matthew E. K. Chang²; Mark R. Flory²; John E. Blume¹*

¹Seer, Inc., Redwood City, CA – *jblume@seer.bio ²OHSU – Knight Cancer Institute- Cancer Early Detection Advanced Research Center (CEDAR), Portland, OR

 

Summary

The proteomic complexity of biological samples limits the scale and speed of robust biomarker discovery studies. Complex workflows for deep profiling of the plasma proteome have limited study size and compromised validation and replication potential, while necessitating trade-offs among depth and breadth of coverage, efficiency, precision, and accuracy.

Herein we describe Proteograph, an efficient, automated multinanoparticle (NP) platform for deep and unbiased proteomic profiling. NPs reproducibly form protein coronas driven by their engineered biophysicochemical properties, and those protein signatures provide insights into protein-protein interaction (PPI)in a biological sample. Specific and reproducible protein-NP binding is the product of binding affinities and a sample’s protein concentrations. The result is compression of concentration range that improves detection of low-abundance proteins without loss of linearity (demonstrated by spike-recovery experiments) or measurement precision. As proof-of-concept, we used our panel of 10 specifically engineered superparamagnetic NPs to identify >2,000 protein groups (1% protein and peptide FDR, 84% with 2 or more peptides) from 16 individual plasma samples.

Proteograph allows parallel processing of 80 NP-sample combinations, from sample to purified peptides, in less than 8 hours. We also compared Proteograph to a typical workflow including high-pH reverse-phase  fractionation, and found higher protein group detection (1,949 vs 658) with better precision (27% vs 41% CVs). In mapping the 1949 and 658 protein groups to a PPI map derived from the STRING database, we identified, respectively, 15/21 and 1/21 interaction clusters with >10 protein members where >10% of the members were covered.

Compared to typical fractionation, the Proteograph platform has superior coverage, depth, precision, and accuracy in unbiased proteomic profiling. Furthermore, this NP-based system has the potential to discover PPIs at a level not possible using conventional fractionation. Proteograph may enable the large-scale studies necessary for protein biomarker discovery and yield new biological insights into the human proteome.

 

View Poster 

First name Last name

Title goes here

Philip Ma is the Chief Business Officer for Seer. Previously, he was Vice-President for Digital Health Technologies and Data Sciences at Biogen, a group that he established in 2015 to discover insights and create value for Biogen and its patients. Prior to joining Biogen, Philip was Senior Partner at McKinsey & Company, where he served global leaders in the pharmaceutical and biotech sectors, leading the West Coast Healthcare Practice and global Personalized Medicine practice for the Firm. Active in community affairs, Philip also serves on the Board of Committee of 100, a group of Chinese-American business and community leaders focused on U.S. China relations and the Asian-American experience. Before McKinsey, Philip was a macromolecular crystallographer in the lab of Dr. Carl O. Pabo at Massachusetts Institute of Technology, where he received his Ph.D. in Biology. Philip also has degrees in Biochemistry (A.B. from Harvard College) and in Economics (MPhil from Oxford University, where he was a Rhodes Scholar).