Clinical Quantitation of Prostate-specific Antigen Biomarker in the Low Nanogram/Milliliter Range by Conventional Bore Liquid Chromatography-Tandem Mass Spectrometry (Multiple Reaction Monitoring) Coupling and Correlation with ELISA Tests

Proteomics discovery leads to a list of potential protein biomarkers that have to be subsequently verified and validated with a statistically viable number of patients. Although the most sensitive, the development of an ELISA test is time-consuming when antibodies are not available and need to be conceived. Mass spectrometry analysis driven in quantitative multiple reaction monitoring mode is now appearing as a promising alternative to quantify proteins in biological fluids. However, all the studies published to date describe limits of quantitation in the low μg/ml range when no immunoenrichment of the target protein is applied, whereas the concentration of known clinical biomarkers is usually in the ng/ml range. Using prostate-specific antigen as a model biomarker, we now provide proof of principle that mass spectrometry enables protein quantitation in a concentration range of clinical interest without immunoenrichment. We have developed and optimized a robust sample processing method combining albumin depletion, trypsin digestion, and solid phase extraction of the proteotypic peptides starting from only 100 μl of serum. For analysis, mass spectrometry was coupled to a conventional liquid chromatography system using a 2-mm-internal diameter reverse phase column. This mass spectrometry-based strategy was applied to the quantitation of prostate-specific antigen in sera of patients with either benign prostate hyperplasia or prostate cancer. The quantitation was performed against an external calibration curve by interpolation, and results showed good correlation with existing ELISA tests applied to the same samples. This strategy might now be implemented in any clinical laboratory or certified company for further evaluation of any putative biomarker in the low ng/ml range of serum or plasma.Used for years across a wide range of pathologies, proteomics studies based on semiquantitative mass spectrometry of proteins have already led to the discovery of numerous protein biomarker candidates. Often tens of putative biomarkers have been described for a single disease, but the subsequent phase of clinical evaluation on large cohorts for each candidate is clearly the bottleneck as revealed by the meager number of newly approved biomarkers for clinical use. One of the critical limitations of discovery work flows arises when no antibody is available to initiate an immunoassay because 1–3 years are required to conceive de novo a reliable immunoassay. Such a delay is a serious drawback when tens of putative markers are concerned. Quantitation of small organic molecules by mass spectrometry has been used extensively for years in the field of environmental contaminant analysis or pharmacokinetic profiling of drug candidates during clinical studies. More recently, absolute quantitation of proteins using mass spectrometry by single (SRM)¹ or multiple reaction monitoring (MRM) and stable isotope dilution has thus naturally emerged as an alternative to immunoassays. Basically the absolute quantitation of a protein is provided by the integration of the specific MRM signals corresponding, respectively, to a proteotypic peptide (1) obtained from enzymatic hydrolysis of the target protein (usually by trypsin) and to its synthetic stable isotope-labeled isotopomer (2). The validation criteria of an MS-based method in terms of accuracy and precision are relatively easy to fulfill when addressing small molecules or proteins below 10 kDa in plasma or serum. Indeed they may be easily extracted from the bulk of high molecular mass proteins simply by selective precipitation. The quantitation of higher molecular mass proteins has proven to be more challenging because of the complexity and large dynamic range of proteins in e.g. plasma. In a pioneering study Anderson and Hunter (3) successfully demonstrated generation of a multiplexed assay for proteins covering high (tens to hundreds of micrograms/milliliter) to medium (hundreds of nanograms/milliliter to a few micrograms/milliliter) abundance ranges in plasma when combined with immunodepletion. However, these ranges remain problematic because clinically relevant biomarkers are usually present in plasma or serum in the low nanogram/milliliter range or below. To significantly improve the limit of quantitation (LOQ) of LC-MRM mass spectrometry, Keshishian et al. (4) evaluated a combination of immunodepletion of the most abundant plasma or serum proteins with strong cation exchange (SCX) chromatography for sample preparation prior to LC-MRM analysis. LOQs in the 1–10 ng/ml range were obtained with a coefficient of variation from 3 to 15% for five exogenous non-human proteins and the human prostate-specific antigen (PSA) protein spiked together into immunodepleted plasma from a healthy female donor. Very recently, a new approach using product ion scanning on a linear ion trap was proposed by Diamandis and co-workers (5) that allowed reaching a limit of quantitation of 1 ng/ml for PSA spiked into control plasma. This study marked a gain in sensitivity compared with previous attempts by others on similar instrumentation (6–9) but applied immunopurification of the target protein.Interestingly all the strategies published to date have been dealing with analytical development of work flows for the validation of biomarker candidates using microbore, nanoflow chromatography. Nanoflow is without any doubt appealing over conventional microflow during the proteomics discovery phase when the amount of biological material, for instance from a tumor biopsy, is often limited. Nonetheless this technique inherently still suffers from a lack of robustness and requires skilled personnel to be operational on a daily basis. As a consequence, nanoflow chromatography is not easily adaptable for the high throughput analysis environment encountered in clinical laboratories or good laboratory practice-certified or contract research organization companies where hundreds of samples are handled per day. In such organizations only microflow separations using 1- or 2-mm-internal diameter HPLC columns are compatible with the requirements of robustness and sample throughput.Therefore, the present work was centered on how a simple work flow could, in the near future, enable the large scale verification phase of putative biomarkers in the ng/ml of plasma range by the use of conventional LC equipment, i.e. using a 2-mm-internal diameter HPLC column. To address this question, we have considered that the absolute quantitation of PSA in true clinical samples could represent a challenging model. Combining immunodepletion of serum albumin and peptide fractionation simply by solid phase extraction (SPE), we were able to demonstrate for the first time the absolute quantitation of PSA by LC-MRM mass spectrometry in clinical serum samples of patients with benign prostate hyperplasia (BPH) or prostate cancer (PCa) within concentrations ranging from 4 to 30 ng/ml. Furthermore a good correlation was observed between the clinical ELISA tests and the mass spectrometry-based assays. We believe that these results are an unprecedented demonstration that the clinical relevance of putative biomarkers issued from proteomics investigation may now be confidently evaluated in the ng/ml range by robust coupling between conventional bore LC and mass spectrometry.