Precision Biomarker Laboratory (PBL) Services
PBL provides liquid chromatography mass spectrometry (LC-MS) based analysis of plasma and dried blood samples, which allows reproducible and accurate assays on several hundreds of proteins simultaneously.
Discovery proteomics methods capable of quantifying thousands of proteins now make it feasible to obtain protein signatures that reflect an individual’s overall health or disease status. Since the proteome is dynamic, the signature we measure is a composite of an individual’s genetic and health status, which can reflect complex underlying disease factors such as diabetes, fragility, obesity, and the impact of various environment stimuli. Proteomic signature changes over time can inform responses to interventions and treatments on a short- and long-term time frame. Novel protein biomarker discovery - alone or in combination with other OMICS such as metabolomics - provides insights into the details of microenvironmental changes.
Changes in cellular protein concentrations and isoform expression are dictated by transcriptional, translational, and protein degradation rates. Post-translational modifications (PTMs), especially those related to disease processes, add layers of functional modulation. Blood components include resident plasma proteins, as well as proteins from tissue leakage or which circulate in response to systemic stimuli, and include innate immunity (e.g. cytokine and others, Fig. 1). With large scale proteomic analysis of hundreds of thousands of samples, the biological, environmental, and genetic impact on health and disease signatures should underscore mechanistic and biological variability.
LC-MS identifies and quantifies proteins at the peptide level - in this case, tryptic digested peptides. PBL has optimized and scaled the proteome technical pipeline steps (see Fig. 2), starting with sample preparation, which requires consistent protein denaturation, reduction and alkylation of cysteine residues to help minimize charge state variance, protein digestion, spike-in of standards, and sample desalting. The second step is data acquisition by LC-MS of the digested plasma or dried blood samples.
LC-MS analysis provides peptide accurate mass and a peak area that is portioned to concentration. Simultaneously, the MS instrument breaks apart those peptides into characteristic daughter fragments that specifically reveal the amino acid sequence allowing for definitive identification of the protein from which it originated. Identification of each protein is based on matching of the amino acid sequence of two or more peptides that are unique to the protein. Note that a protein isoform is identified only if a peptide is observed that comprises a unique amino acid sequence specifically corresponding to the isoform, otherwise the parent protein is used.
Data-independent acquisition MS (DIA-MS) is amenable to large numbers of repetitive samples from complex longitudinal studies, and consists of two steps: 1) creation of a sample-specific peptide library that experimentally encompasses every protein to be quantified in the experiment and 2) DIA-MS analysis of each individual sample, whose constituent peptides can be analytically defined by the pre-created human protein peptide library.
- Specificity and Proteome Breadth: The direct measurement inherent to MS provides unparalleled specificity with no cross-reactivity between analytes. LC-MS measures and quantifies based on two or more peptides representing different regions of the protein. LC-MS is uniquely capable of differentiating closely related proteins (a.k.a. proteoforms), including known isoforms or PTMs, or single nucleotide polymorphisms (SNPs) are confirmed based on unique amino acid sequence. This level of proteome breath can provide additional mechanistic understanding.
Sample Quantity and Quantitative Precision: Compared to other analytical techniques, LC-MS can accommodate reduced sample volume (10 ul) requirements, while providing accurate fold change and precise quantitative measurements, especially if stable-isotope (N15) labeled peptides are included as internal standards. This is often done in discovery if there are specific analytes of interest, and always done with targeted assays. With respect to reproducibility and method robustness, LC-MS methods can analyze thousands of samples with a reproducibility level equivalent to or near clinical chemistry grade assay (%CV of 20%). As well, we have established linearity and lower limits of detection and quantitation as one does with high quality capture assays like ELISA.
Transferability: Due to this rigor, MS assays can be transferable through communication portals and training across a multitude of MS-equipped centers. We have extensive automation involved in our workflow, including sample preparation, QC standardization, data acquisition and automated data processing. This provides a standardized path from assay development to validation of candidate biomarkers on the way to a clinical assay, if desired.
Cost effectiveness: The automation of all aspects of the proteomic pipeline has improved throughput and coupled with efficiency from scaling has reduced the overall cost for running plasma or dried blood samples.
Workflow ||Quantity required||
- Ultimate 300-Triple TOF
|Depleted||10 ul||TFE/Trypsin||Ultimate 300/Exploris|
|Dried Blood||Single||||10 ul tip*||TFE/Trypsin||Evosep/Exploris|
Plasma samples are analyzed using two complementary MS-workflows (naïve and depleted of top 14 abundant proteins) collectively called the “dual workflow” to maximize proteome coverage. (See Table 1) The dual workflow allows precise quantification of 800-1000 proteins representing broad functional groups including inflammation, coagulation, innate and humoral immunity, vascular and endothelial status, lipid metabolism, and tissue leakage proteins indicative of heart, lung, brain and kidney damage, to name a few. (see Fig. 2)
Dried blood samples, such as provided by the Mitra device (Neoteryx) which wicks up 10 or 20 ul of blood from a finger stick, are analyzed in a single run. For proteomics, one 10 ul aliquot is sufficient for detection of over 550 unique proteins.
MS data is acquired on either harmonized 6600 Triple TOF (Sciex) or harmonized Orbitrap Exploris (Thermo). MS instruments are operated in DIA‐MS mode with iRT Standards and sample processing control peptides (beta gal) which are added to each sample before acquisition. Exploris mass spectrometers are equipped with an EasySpray ion source and connected to Ultimate 3000 nano LC system (Thermo) or Evosep System (Evosep). DIA-MS is performed using 12 Dalton (50 scan events) windows over the precursor mass range of 400-1000 m/z and the MS2 mass range is set from 100-1500 m/z. The 6600 Triple TOF (Sciex) is connected to Eksigent 415 LC system that is operated in micro-flow mode. DIA MS1 scans are acquired using a dwell time of 250 ms in the mass range of 400-1250 m/z using a 4Da fixed window over a precursor range of 400-1000 m/z with a dwell time of 22msec.
Quality control: PBL employs several levels of QC are employed, including plate replicates for assessment of sample preparation, and MS batch replicates which are digested pools run before and after each MS batch. We also include IRT peptides standards in each sample. We continuously conﬁrm stable system performance across patient sample analyses with “QuiC,” a new monitoring tool that generates liquid chromatography and mass spectrometer QC readouts automatically from raw data files. We add synthetic peptide standards to all individual samples prior to analysis and confirm they can be detected reproducibly across the entire mass spectrometry run (Fig. 3.)
Data analysis: Data analysis is provided for peptide and protein quantification and corresponding information from Uniprot on cell specificity and functionality for all unambiguous proteins observed. PTMs including phosphorylation, methylation and acetylation can be provided, where detected. Protein changes over time, both for individual patients and collectively, will be provided upon disclosure of time series.
Assay libraries will be generated. DDA files are converted to mzXML and searched through the Trans Proteomic Pipeline (TPP) using multiple search engines, i.e. Comet, X!Tandem Native scoring, and X!Tandem K‐scoring against Uniprot database of interest along with randomized decoy proteins. In cases of isoform identification, Uniprot database with canonical and isoform specific sequences will be included. The raw spectral libraries will be generated from all valid peptide spectrum matches and then refined into the non-redundant consensus libraries and, for each peptide, retention time will be mapped into the iRT space with reference to a linear calibration constructed for each run. These spectral libraries will then be filtered to generate peptide assay libraries by selecting the top twenty most intense b- or y-ion fragments. Protein isoforms will only be included in the library if a peptide comprising an amino acid sequence unique to the isoform is identified.
The resulting file is imported into Skyline for DIA data visualization and simultaneously converted into a library for OpenSWATH quantitation analysis. DIA-MS analysis is done using our Optra Platform, which is a powerful AI-based technology with distributed architecture to enable highly customizable DIA workflows that derive meaningful biological insights via automated data processing, high-throughput data analytics and standardized reporting. DIA-MS.wiff files are first converted to profile mzML using Sciex Converter and then analyzed using OpenSWATH which utilizes a target-decoy scoring system (e.g. mProphet) to estimate the identification of FDR. Next, to obtain high-quality quantitative data at the protein level, proteins whose peptides are shared between multiple different proteins (non-proteotypic peptides) are discarded. Data pre-processing and statistical analysis of MS runs into quantitative data are performed using mapDIA.
Proteomic Biomarker Discovery - Select Publications
- 1. Takimoto E, Champion HC, Li M, et al. Chronic inhibition of cyclic GMP phosphodiesterase 5A prevents and reverses cardiac hypertrophy. Nat Med. 2005;11(2):214‐222. doi:10.1038/nm1175.
- 2. Lee DI, Zhu G, Sasaki T, et al. Phosphodiesterase 9A controls nitric-oxide-independent cGMP and hypertrophic heart disease. Nature. 2015;519(7544):472‐476. doi:10.1038/nature14332.
- 3. Texas Brain Injury Alliance [Internet]. Austin (TX): Texas Brain Injury Alliance. Brain Injury Statistics; [cited 2020 June 1]. Available from: http://www.texasbia.org/about-brain-injury/brain-injury-statistics/
- 4. BRAINBox Solutions [Internet]. Richmond (VA): BRAINBox Solutions. Company Also Announces Commencement of International Pivotal Clinical Study; [cited 2020 June 1]. Available from: https://brainboxinc.com/brainbox-solutions-receives-breakthrough-device-designation-from-fda-for-first%E2%80%90of%E2%80%90kind-device-to-aid-in-concussion-diagnosis-and-prognosis/
- 5. Everett AD, Van Eyk J, Korely F, & The Johns Hopkins University. 2014 July 17. Multi-protein biomarker assay for brain injury detection and outcome. US 10534003.
- 6. Van Eyk J, Everett AD, Jin Z, & The Johns Hopkins University. 2017 June 28. Citrullinated brain and neurological proteins as biomarkers of brain injury or neurodegeneration. US 10365288.
- 7. Van Eyk J, Everett AD, Jin Z, & The Johns Hopkins University. 2013 Mar 13. Citrullinated brain and neurological proteins as biomarkers of brain injury or neurodegeneration. US 9709573.
- 8. Van Eyk J, Everett AD, Jin Z, & The Johns Hopkins University. 2017 June 28. Citrullinated brain and neurological proteins as biomarkers of brain injury or neurodegeneration. US Application 15/636076.