Evaluation of Endogenous Plasma Peptide Extraction Methods for Mass Spectrometric Biomarker Discovery Lina P. Aristoteli,† Mark P. Molloy,†,‡ and Mark S. Baker*,†,‡ Australian Proteome Analysis Facility Ltd and Department of Chemistry and Biomolecular Sciences, Macquarie University, Sydney, NSW. 2109 Australia Received June 19, 2006
Peptides have a role in the inflammatory response, tumor biology, and endocrine processes, presenting them as appealing biomarker candidates. However, peptide extraction efficacy for clinical profiling remains a pivotal technological challenge, as maximum coverage of the plasma peptidome is limited by a range of factors including the inherent complexity of human plasma and the lower concentration of peptides compared to abundant proteins. The aim of this study was to evaluate commonly employed peptide extraction methodologies in terms of total number of peptides detected and the mass range of peptides observed by MALDI. Despite showing coelution of proteins, solid-phase extraction (SPE) methods exhibited superior plasma peptide recovery than ultrafiltration, acetonitrile (ACN) precipitation, or size-exclusion chromatography methods under conditions employed in the study. Not surprisingly, in line with studies challenging the veracity of many peptide biomarker studies, the majority of identified peptides eluted from SPE methods corresponded to proteolytic truncations of the most abundant plasma proteins. The prefractionation of plasma with acetonitrile precipitation prior to SPE provided distinct ion signal profiles and is worthy of further study. In conclusion, this study favors the use of SPE in peptide extraction protocols for increased biomarker coverage and diversity from the plasma peptidome. Keywords: plasma • peptidome • peptide • solid-phase extraction • MALDI • biomarker discovery
Introduction The term “peptidomics” was introduced in 2001 to define the quantitative and qualitative analysis of endogenous peptides in biological samples,1 primarily by chromatography or biochip platforms coupled to various forms of mass spectrometry. Characterization of low molecular weight constituents in biological fluids has led to data sets available for circulating peptides and small proteins in plasma and serum.2-4 To decrease the effect of abundant proteins on peptide detection, hemofiltrate as an alternative to plasma has been used for profiling studies.5,6 Here, the use of hemofiltrate and fractionation techniques for peptide profiling reduced the concentration of larger proteins by a factor of 1000 compared to plasma,5 allowing the identification of over 5000 unique peptides with 95% of these identifications from proteins/peptides with molecular weight below 15 kDa.6 In the setting of cancer biology, endogenous plasma peptides could represent the activation of nascent proteolytic machineries or metabolic processes integral to core elements of tumor initiation, progression, and metastasis. Differentially expressed peptides have been identified in peptidomic profiling of Type * To whom correspondence should be addressed. Professor Mark S. Baker, Australian Proteome Analysis Facility Level 4 Building F7B, Research Park Drive, Macquarie University, Sydney NSW 2109 Australia; Tel: +61 2 9850 6916; Fax: +61 2 9850 6200; E-mail:
[email protected]. † Australian Proteome Analysis Facility Ltd. ‡ Department of Chemistry and Biomolecular Sciences. 10.1021/pr0602996 CCC: $37.00
2007 American Chemical Society
2 diabetes,7 estrogen-positive- and estrogen-negative breast carcinoma tissue,8 normal human mammary and breast cancer cell lines,9 and cerebrospinal fluid from patients with Alzheimer’s disease.10 Attempts have been made to identify peptide biomarkers in serum using proteomic patterns detected by SELDI (surface-enhanced laser desorption ionization) discriminatory for cancers such as ovarian11 and breast cancer.12 However, the possible identification of serum artefacts derived from processing of abundant proteins as cancer biomarkers with unusually high specificity and sensitivity has been one criticism of mass spectrometry peptide “ladder” patterns detected by pattern-matching processes like SELDI.13 Recently, this view that low molecular weight peptide signatures arising from fragmentation of abundant blood proteins provide limited diagnostic utility has been challenged. In a study by Villanueva et al., a serum peptide signature predominantly composed of ladder-like truncations from common blood protein fragments was able to discriminate between advanced prostate, breast, and bladder cancers, and to differentiate these cancers from healthy controls.14 The authors proposed that the serum clotting cascade was required to generate peptide fragments that were susceptible to diseasespecific proteolysis from proteases originating from the tumor or its microenvironment, thereby creating cancer-specific diagnostic signatures. This potentially demonstrates the enormous diagnostic potential of the low molecular weight serum peptidome in addition to detection of in vivo peptide fingerJournal of Proteome Research 2007, 6, 571-581
571
Published on Web 12/23/2006
research articles prints arising from cellular proteins, and tissue-derived peptides that may be attached to carrier proteins in blood.15 It is generally agreed, though, that there is a need for standardization of blood collection and sampling protocols for biomarker discovery, in particular at the peptide level. One major obstacle for the use of plasma for peptide profiling under pathological conditions is that peptides usually emanate from degradation processes, making it difficult to distinguish artifact peptides originating during venipuncture collection and/or sample handling conditions from those that are uniquely derived from the disease. The Human Proteome Organization (HUPO) has initiated a pilot study for assessment of blood preanalytical variables on proteomic and peptidomic profiles (HPPP).16 One of the study aims was to raise awareness for the need for standard operating procedure (SOP)-controlled blood collection and processing procedures to be employed globally for the detection of “valid” protein and peptide disease biomarkers. A main, though controversial,14,15 recommendation from the HUPO HPPP trial phase was the consideration of protease inhibitors during blood collection to prevent protein and peptide degradation artefacts.16 Proteolytic cascades may even be affected by blood sampling techniques that have been shown to influence the concentration of markers of active coagulation.17 Artifactual “peptide families” or “peptide ladders” may result from sequential N- and C-terminal cleavages of peptide sequences in plasma and serum that coincide with known inherent plasma carboxy- and amino-proteases activated by proteases like plasmin.4,18 Peptide artifacts can be observed when one compares patterns observed from serum compared to plasma, and when blood centrifugation occurs at a temperature
