Protein biomarker development programs center on the discovery, identification, and validation of patterns of differential expression of proteins and peptides resulting from a disease state, event, or treatment. Examples include proteins released during cardiovascular events, shed by or resulting from tumors, and modulated due to drug response. Scientific and clinical interest in novel protein biomarkers and improved panels for existing biomarkers continues to grow, with a goal of better decision making in disease diagnosis and prognosis, and in drug discovery.

Improvements in the sensitivity, reproducibility, and throughput of biomarker assays are advancing targeted, hypothesis-driven biomarker quantitation and validation. The evolution from standard ELISAs to bead-based, multiplexed assays—as well as the introduction of novel mass spectrometric assays such as multiple reaction monitoring (MRM) and its extension, stable isotope standards and capture by anti-peptide antibodies (SISCAPA)—highlight these advancements. MRM is based on the quantitation of specific secondary ions from selected and fragmented target peptide sequences, and SISCAPA provides the enhanced sensitivity of enrichment by immunocapture. The addition of biomarker candidates into the validation pipeline will be accelerated by improvements in de novo biomarker discovery and identification technologies, as well as mining of novel biological materials. 

	Bottom Up	Top Down
Features	LC/MS-MS	2-DGE-MS/MS	LC-MALDI-TOF/TOF	SELDI-TOF/TOF
Mass Range	>30 kD	>15 kD	<20 kD	<70 kD
Analysis Of	Tryptic Digests	Intact Proteins
Strengths	Direct quantitation and identification Best sensitivity	Retain PTM and cleavage information		Retain PTM and cleavage information High throughput Ease of sample preparation
Weaknesses	Loss of PTM and cleavage information Poor reproducibility Low throughput	Low throughput Poor reproducibility		Purification often required for identification
Features of top-down and bottom-up biomarker discovery methodologies (Source: Bio-Rad)

Advances have been made in recent years for both “top-down” and “bottom-up” profiling approaches to biomarker discovery. In bottom-up methods, biological samples are digested with a protease (typically trypsin) and subjected to a liquid chromatography (LC) separation prior to analysis in a high-resolution mass spectrometer. The major strengths of these approaches are direct quantitation and identification along with reasonable dynamic range of detection; however, the lengthy process limits throughput and reproducibility. In addition, biomarker candidates below 30 kD are challenging for bottom-up approaches, in part because smaller proteins and peptides have fewer proteolytic cleavage sites and often do not generate enough peptides for confident identification. Native small peptides, biologically generated protein cleavages, and posttranslational modifications (PTMs) are better sought through top-down profiling analyses of native proteins and peptides. Proteolytic events and other PTMs are postulated to be relevant in many diseases and other biological processes, and specific variants or fragments of proteins—including those generated from high-abundance proteins—may prove to be important biomarker candidates (for example, phosphoproteins and protein fragments as biomarkers of brain injury¹ and cytokeratin-18 fragments as biomarkers for nonalcoholic steatohepatitis²). Top-down approaches also benefit from simplified sample preparation, potential for higher sample throughput, and are amenable to many different and complex biological samples. Bottom-up methods are generally driven from more specific and limited sample sets. Features of each approach are outlined in the table.

“New” sample sources
Technological advances have enabled the analysis of many more sample sources for biomarker discovery, moving far beyond the first cancer biomarkers discovered by immunization of laboratory animals with tumor tissue lysate preparations.³ Modern imaging techniques—such as positron emission tomography (PET)—are often able to locate abnormalities such as tumors by monitoring uptake of radioactive sugars, but these techniques do not provide a direct means for discovery of new biomarkers. Instrumentation for matrix-assisted laser desorption/ionization (MALDI) mass spectrometric imaging was first introduced in 2004, and these analyses of diseased and adjacent (healthy) tissues provides for in situ differential protein analysis. Improved instrument offerings, including Bruker Daltonics’ (Bremen, Germany) ultrafleXtreme and Shimadzu’s (Columbia, Md.) AXIMA Performance mass spectrometers have promoted the viability and value of this approach.

In addition to being directly analyzed in traditional samples used for biomarker discovery such as serum, plasma, and urine, biomarkers have more recently been sought through analysis of proteins bound to carrier proteins such as albumin, transferrins, and immunoglobulins. Some of the candidate biomarkers discovered this way are smaller peptides and proteins or the fragments of high-abundance proteins that may have been discounted or missed in previous discovery approaches. This is especially true, if immunodepletion of high-abundance proteins was used to investigate soluble species of lower abundance (for example, the fragmentation products of abundant proteins associated with nephritic syndrome⁴ and breast cancer⁵).

Direct analysis of the “secretome”—proteins secreted into culture media by cell lines or tissue samples—has also been facilitated by the recent improvements in sensitivity and sample requirements of new mass spectrometry instruments. Analyses of the secretome are expected to produce biomarker candidates that may also be detectable and differentially expressed in more accessible biological fluids. Recent studies in this area include non-invasive profiling of the human embryonic secretome with the goal of identifying cultured embryos with the highest implantation potential⁶ and identification of factors that influence the biochemical environment of the nervous system and the nervous system response to inflammation⁷.

Technical advances 
The availability of precast 1D and 2D gels has improved ease-of-use and reproducibility for gel-based top-down approaches, but these classic techniques remain time-intensive and sample-limited and suffer from the challenge of poor reproducibility. High-throughput techniques—such as micro-liquid chromatography, capillary electrophoresis, and array-based sample capture—permit analysis of larger sample sets, which is preferred for greater statistical power and higher confidence in potential biomarker value. Some instrumental advances are beneficial to both top-down and bottom-up approaches, including improvements to liquid chromatographic separations such as shorter elution times, reduced sample volume requirements, and diverse support chemistries. 

The wide dynamic range in protein concentration of many valuable proteomic samples remains a challenge for biomarker discovery programs, but improvements in the specificity of depletion techniques—including the Agilent (Santa Clara, Calif.) Multiple Affinity Removal System and GenWay (San Diego, Calif.) Seppro IgY microbeads—and sample fractionation tools—such as Bio-Rad’s (Hercules, Calif.) ProteoMiner combinatorial peptide library—improve proteomic coverage. These resin-based methods for removal or reduction in the concentration of high abundance proteins enhance the number of peptides and proteins that may be detected in subsequent analyses (e.g, LC-MS, 2D gel, MALDI).  The use of higher resolution mass spectrometry instruments, such as the ultrafleXtreme, also increase the likelihood of discovery of differentially expressed (or differentially post-translationally processed) proteins.

Array-based approaches for sample cleanup and fractionation continue to be valuable for biomarker discovery and validation, and antibody arrays coupled with fluorescence detection are frequently utilized for targeted biomarker assay. The early promise of surface-enhanced laser desorption/ionization (SELDI) as a biomarker discovery platform has been recently validated by the US Food and Drug Administration (FDA) clearance of an ovarian cancer test.^8,9 The test combines results from five biomarker immunoassays; four of these markers were discovered and validated by the SELDI technology. Following their acquisition of SELDI technology from Ciphergen in 2006, Bio-Rad recently announced the launch of the Lucid Proteomic System, co-developed with Bruker Daltonics. With this system, chromatographic ProteinChip SELDI arrays are coupled with Bruker’s flex series MALDI TOF/TOF mass spectrometers.  This combination of high-throughput, simplified sample preparation with high-performance mass spectrometry provides a proteomic tool for top-down biomarker research.

Conclusion 
New technologies such as MRM and SISCAPA have grown in popularity and utility for biomarker validation and quantitation in recent years, and both bottom-up and top-down approaches for biomarker discovery have improved the stream of candidates for validation. These discovery approaches are fundamentally different, present unique challenges and opportunities for proteomic researchers, and can be utilized as complementary approaches to solving today’s complex biomarker requirements. Bottom-up (digest based) methods for biomarker workflow are well established, and recent improvements to top-down approaches better enable the discovery of biologically derived “native” proteins, including specific post-translational fragmentation or modification. The synergistic effect of technological advances with the ability to investigate novel sample sources is anticipated to accelerate the pace of protein biomarker discovery applicable to disease diagnosis and prognosis, in drug discovery applications, and in the expansion of personalized medicine.

Enrique Dalmasso received a PhD in biophysical chemistry from the University of California at Berkeley. He manages Lucid applications and collaborations for Bio-Rad Laboratories. Amanda Bulman received a PhD in chemistry from the University of Virginia. Formerly a project manager for pharmaceutical and academic clients within Bio-Rad’s Biomarker Research Center, she is currently serving as application manager. 

References 
1. Siman R, Toraskar N, Dang A, McNeil E, McGarvey M, Plaum J, Maloney E, Grady MS. A Panel of Neuron-Enriched Proteins as Markers for Traumatic Brain Injury in Humans. J Neurotrauma. [Epub ahead of print] Oct 7 2009.
2. Feldstein AE, Wieckowska A, Lopez AR, Liu YC, Zein NN, McCullough AJ. Cytokeratin-18 fragment levels as noninvasive biomarkers for nonalcoholic steatohepatitis: a multicenter validation study. Hepatology. 2009; 50:1072-1078.
3. Ablin RJ. Immunologic studies of normal, benign, and malignant human prostatic tissue. Cancer. 1972; 29:1570–1574.
4. Candiano G, Musante L, Bruschi M, Petretto A, Santucci L, Del Boccio P, Pavone B, Perfumo F, Urbani A, Scolari F, Ghiggeri GM. Repetitive Fragmentation Products of Albumin and α₁-Antitrypsin in Glomerular Diseases Associated with Nephrotic Syndrome. J Am Soc Nephrol. 2006; 17:3139–3148.
5. Gast MC, van Dulken EJ, van Loenen TK, Kingma-Vegter F, Westerga J, Flohil CC, Knol JC, Jimenez CR, van Gils CH, Wessels LF, Schellens JH, Beijnen JH. Detection of breast cancer by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry tissue and serum protein profiling. Int J Biol Markers. 2009; 24:130-141. 
6. Katz-Jaffe MG, McReynolds S, Gardner DK, Schoolcraft WB. The role of proteomics in defining the human embryonic secretome. Mol Hum Reprod. 2009; 15:271-277.
7. Keene SD, Greco TM, Parastatidis I, Lee SH, Hughes EG, Balice-Gordon RJ, Speicher DW, Ischiropoulos H. Mass spectrometric and computational analysis of cytokine-induced alterations in the astrocyte secretome. Proteomics. 2009; 3: 768-782.
8. Quest Diagnostics press release, September 11, 2009.  http://newsroom.questdiagnostics.com/index.php?s=43&item=362.
9. FDA news release, September 11, 2009. http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm182057.htm.