Abstract
Noninvasive molecular biomarkers are becoming attractive tools to monitor disease progression, aid drug development programs and use as surrogate outcome measures in clinical trials. Cutting edge proteomic methods to assay biomarkers in body fluids have been developed in the past few years, but transitioning them to clinical practice has been slow and depends on the qualification of both the method and the biomarker.
Major advances in biomarker research have occurred recently; thanks to recent developments in proteomic methods, a number of valuable biomarkers have been discovered and continue to be discovered for different diseases and conditions Citation[1–4]. Historically, effort to discover novel biomarkers in body fluids, especially in plasma or serum, have been challenging owing to the large dynamic range of the proteome and the moderate resolving power of early instruments Citation[5]. But with development of fractionation methods coupled with high precision and high resolution mass spectrometers Citation[6] and advances in affinity array assays Citation[7,8], measuring biomarkers in complex samples is becoming a relatively routine task and has attracted several investigators and biopharmaceutical companies into a race for biomarker discovery and validation.
Potential biomarkers include metabolites, miRNAs, and proteins. Proteins, in particular, seem to be more attractive candidates because they are specific and more informative about a condition than miRNAs and metabolites.
Biomarkers as applied to a given disease have utility in one or more of the following areas:
Can act as a diagnostic tool. A panel of biomarkers is more specific and reliable than a single biomarker.
Can inform us about disease pathobiochemical pathways that can lead to development of novel therapeutic targets.
Can act as pharmacodynamic biomarkers to monitor safety and efficacy of a new treatment.
Can act as surrogate biomarkers to anticipate or predict later clinical benefit of an intervention.
Molecular biomarkers are unbiased compared with questionnaires and other tests imposed on patients, especially in pediatric patients.
Biomarker discovery and validation studies can be performed using two main technologies: mass spectrometry (MS)-based proteome profiling and affinity multiplexing assays. Both technologies have advantages and disadvantages as discussed below.
MS-based serum or plasma proteome profiling using data dependent acquisition (DDA) is ideal for discovery of novel biomarkers but lacks throughput for validation phase when hundreds to thousands of samples need to be analyzed. The workflow in MS approach consists of several steps, including sample fractionation to simplify sample complexity followed by analysis of each fraction by liquid chromatography tandem mass spectrometry (LC-MS/MS). Sample preparation for MS remains a time consuming bottleneck. Furthermore, LC-MS/MS analysis of each fraction can take an hour or more, resulting in several hours to a day to complete analysis of a single sample. This has tremendously limited the throughput needed for biomarker studies. Another issue with DDA is the poor repeatability in detecting and quantifying the same set of proteins, especially low abundant proteins, across all samples Citation[9]. To overcome this issue, investigators often perform wide proteome profiling using untargeted DDA approach on a subset of representative samples or pooled samples for a discovery phase, then targeted MS, also known as data-independent acquisition approach, for validation across biological replicates Citation[10].
Targeted quantification of proteins by MS is gaining popularity and was selected as the method of the year in 2012 Citation[11]. This method is suitable for validation of biomarkers and relies on spiking biological sample with stable isotope-labeled internal standard peptide(s) or protein(s) that share 100% sequence homology with target peptides or proteins in the sample. Spiked samples are then processed and analyzed by LC-MS/MS. Labeled and unlabeled peptides are selected for MS/MS and the transition ions, often ‘y’ ion series, are used to extract the ion chromatogram and determine quantitative ratios. The method is highly specific, accurate and linear over a wide dynamic range Citation[12]. It is particularly applicable when antibodies for the target proteins are of poor quality or not available Citation[13]. But simultaneous targeting of several biomarkers in multiple samples is still not routine. It requires tremendous optimization of the method, including selection of best ionizing peptides, testing different amounts of internal standard to spike into samples, establishing calibration curves to ensure linearity and quantitative dynamic range coverage. To overcome some of these challenges, an approach using a wider MS/MS window termed Sequential Window Acquisition of all THeoretical Mass Spectra (SWATH) strategy was recently tested and consistently identified and quantified 15,000 peptides corresponding to 2500 proteins across 18 samples Citation[14]. While the number of reproducibly quantified proteins across all 18 samples is impressive, the study was done on yeast extract amenable to metabolic labeling, thus facilitating quantification. This technique might prove challenging when dealing with real clinical samples for which several targets with different abundances need to be accurately quantified.
Affinity proteome profiling with high multiplexing and parallel quantification capabilities are emerging as very attractive methods for biomarker studies. In the past 5 years, two useful approaches have matured. Affinity-based protein profiling using antibody bead arrays totaling 4608 antibodies that target 3450 genes Citation[1] and SomaScan technology that uses modified aptamers or SOMAmers targeting 1129 proteins in a single run Citation[8]. Both methods were found to be highly sensitive, requiring only small volume of serum or plasma, with no need for albumin depletion and capable of measuring multiple targets over a wide dynamic range. Antibody bead arrays are dependent on antibody quality to recognize specific epitopes but some of the antibodies might cross-react with multiple proteins rendering validation challenging. Furthermore, the method requires rigorous and laborious sample preparation to optimize dilution of individual antibodies. The method is often implemented in two steps. First, a discovery phase using all available antibodies to screen a subset of case and control samples, then the best reacting antibodies that discriminate between the groups are selected for further validation using larger sample size Citation[1]. SOMAmers on the other hand recognize specific epitopes on native folded proteins with high specificity and high sensitivity Citation[15]. Currently, there are 1129 well-characterized SOMAmers and each SOMAmer is tagged with a unique small DNA sequence than can be used to deduce the concentration of the bound SOMAmer using high throughput microarray analysis. While this new technology is very promising and well optimized for plasma and serum, it remains expensive to run and will require optimization with new matrices.
Importantly, to be approved by regulatory agencies, a bio-analytical assay needs to adhere to rigorous guidelines and undergo ‘fit-for purpose’ technical validation. An assay using MS is still in the early stages of qualification Citation[12]. While MS assay is clearly more specific and accurate with coefficient of variations (CVs) that adhere to regulatory guideline Citation[16], the methodology itself needs to undergo analytical validation before being qualified by regulatory agencies. In this context, the Clinical Proteomic Tumor Analysis Consortium has been working closely with the National Cancer Institute and in collaboration with the US FDA to establish a guidance document for targeted MS assay. These include: response curve with a blank and a minimum of six concentrations of the target; testing repeatability; test target peptide(s) selectivity in replicate biological samples; test stability of the method under different conditions, temperature, freeze–thaw cycles and shelf time; and finally examine intra- and inter-variability of the entire procedure over time. But one of the contributing factors to variation seen in targeted MS assay is the spiking of the stable isotope labeled standards as peptides after the proteolysis step. This has given rise to concerns about reproducibility of the assay in respect to sample handling, trypsin digestion efficiency, and batch to batch repeatability. One way to overcome these challenges is to use full length stable isotope-labeled proteins as internal standards before digesting and processing the sample. This strategy was recently implemented to measure dystrophin protein in human muscle extracts and was found to be highly reproducible with CVs <10% when compared with post-digestion spiking approach Citation[17]. But this requires an abundant and renewable source of full length-labeled targets. Efforts toward making full length-labeled standard proteins are underway but at high cost. Biomarker measurement using affinity approaches on the other hand are easily adaptable to clinical settings if they adhere to guidelines set by regulatory agencies (e.g., high repeatability, linearity, and low CVs).
Overall, mature technologies are available today for biomarker studies. The two affinity-based proteome techniques are reliable, highly reproducible and can analyze hundreds to thousands of samples in a short period of time but will not detect potential biomarkers for which no antibodies or SOMAmers exist. The MS-based proteome profiling approach is considerably more labor intensive and time consuming, but highly accurate and can discover new biomarkers, including truncated proteins that are often overlooked when using affinity-based proteomics. Great strides have been made and we can anticipate continued improvements and cost reductions in the future. However, well-standardized protocols for serum collection and well-controlled cohorts with balanced controls and cases, including age, gender and ethnicity, are needed for discovery and validation of reliable biomarker.
Acknowledgements
The author would like to thank Dr. Kristy Brown and Dr. Eric P Hoffman for their help with proof-reading this editorial.
Financial & competing interests disclosure
Y Hathout was supported by National Institute of health grants (R01AR062380, P50AR060836, R24HD050846 and P30HD040677). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
References
- Bystrom S, Ayoglu B, Haggmark A, et al. Affinity proteomic profiling of plasma, cerebrospinal fluid, and brain tissue within multiple sclerosis. J Proteome Res 2014;13(11):4607-19
- Hathout Y, Marathi RL, Rayavarapu S, et al. Discovery of serum protein biomarkers in the mdx mouse model and cross-species comparison to Duchenne muscular dystrophy patients. Hum Mol Genet 2014;23(24):6458-69
- Mehan MR, Williams SA, Siegfried JM, et al. Validation of a blood protein signature for non-small cell lung cancer. Clin Proteomics 2014;11(1):32
- Quon BS, Dai DL, Hollander Z, et al. Discovery of novel plasma protein biomarkers to predict imminent cystic fibrosis pulmonary exacerbations using multiple reaction monitoring mass spectrometry. Thorax 2015. [Epub ahead of print]
- Anderson NL, Anderson NG. The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002;1(11):845-67
- Pernemalm M, Lehtio J. Mass spectrometry-based plasma proteomics: state of the art and future outlook. Expert Rev Proteomics 2014;11(4):431-48
- Ayoglu B, Haggmark A, Neiman M, et al. Systematic antibody and antigen-based proteomic profiling with microarrays. Expert Rev Mol Diagn 2011;11(2):219-34
- Gold L, Walker JJ, Wilcox SK, Williams S. Advances in human proteomics at high scale with the SOMAscan proteomics platform. N Biotechnol 2011;29(5):543-9
- Tabb DL, Vega-Montoto L, Rudnick PA, et al. Repeatability and reproducibility in proteomic identifications by liquid chromatography-tandem mass spectrometry. J Proteome Res 2010;9(2):761-76
- Jaffe JD, Keshishian H, Chang B, et al. Accurate inclusion mass screening: a bridge from unbiased discovery to targeted assay development for biomarker verification. Mol Cell Proteomics 2008;7(10):1952-62
- Marx V. Targeted proteomics. Nat Methods 2013;10(1):19-22
- Carr SA, Abbatiello SE, Ackermann BL, et al. Targeted peptide measurements in biology and medicine: best practices for mass spectrometry-based assay development using a fit-for-purpose approach. Mol Cell Proteomics 2014;13(3):907-17
- Harlan R, Zhang H. Targeted proteomics: a bridge between discovery and validation. Expert Rev Proteomics 2014;11(6):657-61
- Selevsek N, Chang CY, Gillet LC, et al. Reproducible and Consistent Quantification of the Saccharomyces cerevisiae Proteome by SWATH-mass spectrometry. Mol Cell Proteomics 2015;14(3):739-49
- Lollo B, Steele F, Gold L. Beyond antibodies: new affinity reagents to unlock the proteome. Proteomics 2014;14(6):638-44
- Addona TA, Abbatiello SE, Schilling B, et al. Multi-site assessment of the precision and reproducibility of multiple reaction monitoring-based measurements of proteins in plasma. Nat Biotechnol 2009;27(7):633-41
- Brown KJ, Marathi R, Fiorillo AA, et al. Accurate quantitation of dystrophin protein in human skeletal muscle using mass spectrometry. J Bioanal Biomed 2013;(Suppl 7)