Serum Proteomics in Biomedical Research: A Systematic Review

Proteins that are important indicators of physiological or pathological states may contribute to the early diagnosis of disease, which may provide a basis for identifying the underlying mechanism of disease development. Serum, contains an abundance of proteins, offers an easy and inexpensive approach for disease detection and possesses a high potential to revolutionize the diagnostics. These differentially expressed proteins in serum have become an important role to monitoring the state for disease. Availability of emerging proteomic techniques gives optimism that serum can eventually be placed as a biomedium for clinical diagnostics. Advancements have benefited biomarker research to the point where serum is now recognized as an excellent diagnostic medium for the detection of disease. Comprehensive proteome of human serum fluid with high accuracy and availability has the potential to open new doors for disease biomarker discovery and for disease diagnostics, providing insights useful for future study. Thus, this review presents an overview of the value of serum as a credible diagnostic tool, and we aim to summarize the proteomic technologies currently used for global analysis of serum proteins and to elaborate on the application of serum proteomics to the discovery of disease biomarkers, and discuss some of the critical challenges and perspectives for this emerging field.     

New method for prefractionation of plasma for proteomic analysis

The depth of proteome analysis is severely limited in complex samples with a wide dynamic range of protein abundance such as plasma. Removal of high‐abundance proteins should reveal the signal of lower abundance plasma proteins. However, smaller proteins may be part of larger protein complexes and hence the removal of proteins involved in complexes with high‐abundance proteins such as albumin may inhibit the search for disease biomarkers. Prefractionation of a sample divides it into fractions of reduced complexity, allowing improved detection of lower abundance proteins. Using a prefractionation device, which employs Gradiflow? technology, we were able to separate small volume plasma samples into multiple fractions based on the molecular weight and/or charge. The resulting samples of reduced complexity were directly compatible with 2‐DE. The use of this prefractionation machine may therefore be useful in the hunt for disease biomarkers.     

Tumor pleural effusion proteome profiling for ovarian cancer biomarkers mining

Mass spectrometric proteome profiling of tumor pleural effusion (TPE) liquid fraction from ovarian cancer patients was performed to identify the potential biomarkers of the disease. The methodology of analysis of the TPE protein composition included the removal of high-abundant proteins by affinity chromatography, additional fractionation of the low-abundant proteins based on their lipophilicity, and high-resolution mass spectrometric analysis. As a result, 190 proteins were indentified, 49% of them belonging to the groups of extracellular and membrane proteins. The application of several criteria to data analysis allowed us to generate a group of 26 proteins that are promising candidates for testing as ovarian cancer biomarkers.     

Altered protein profiles in human umbilical cords with preterm and full‐term delivery

Several biomarkers are routinely used clinically for predicting preterm labor; however, these factors are either nonspecific or detected too late. Here, we performed protein profiles in preterm‐ and term‐derived human umbilical cord by using 2DE. Approximately 200 different proteins were identified between preterm‐ and term‐delivered umbilical cords. Among them, 48 proteins were identified. A comparison of preterm proteome to that of term proteome revealed potential candidates for biomarkers, such as hypoxia‐inducible proteins, phosphorylated heat‐shock protein 27 (HSP27), transgelin, vimentin, and transferrin that are specific to preterm umbilical cords. Especially, HSP27 in preterm‐derived umbilical cords shows a significant increase in the mono‐ and tetra‐phosphorylation. The real importance of all of HSP27 phosphorylation as well as hypoxia‐inducible factor 1alpha, and glyceraldehyde 3‐phosphate dehydrogenase require further validation in vitro and in vivo; nevertheless, we believe that they could represent promising diagnostic targets for detection of sudden early delivery. In conclusion, the results of the current study may provide important insights into the molecular mechanisms underlying umbilical cord development.     

Gene Ontology (GO)-Driven Inference of Candidate Proteomic Markers Associated with Muscle Atrophy Conditions

Skeletal muscle homeostasis is essential for the maintenance of a healthy and active lifestyle. Imbalance in muscle homeostasis has significant consequences such as atrophy, loss of muscle mass, and progressive loss of functions. Aging-related muscle wasting, sarcopenia, and atrophy as a consequence of disease, such as cachexia, reduce the quality of life, increase morbidity and result in an overall poor prognosis. Investigating the muscle proteome related to muscle atrophy diseases has a great potential for diagnostic medicine to identify (i) potential protein biomarkers, and (ii) biological processes and functions common or unique to muscle wasting, cachexia, sarcopenia, and aging alone. We conducted a meta-analysis using gene ontology (GO) analysis of 24 human proteomic studies using tissue samples (skeletal muscle and adipose biopsies) and/or biofluids (serum, plasma, urine). Whilst there were few similarities in protein directionality across studies, biological processes common to conditions were identified. Here we demonstrate that the GO analysis of published human proteomics data can identify processes not revealed by single studies. We recommend the integration of proteomics data from tissue samples and biofluids to yield a comprehensive overview of the human skeletal muscle proteome. This will facilitate the identification of biomarkers and potential pathways of muscle-wasting conditions for use in clinics.     

Identification of Peptide tumor markers in a tumor graft model in immunodeficient mice

The medical demand for useful biomarkers is large and still increasing. This is especially true for cancer, because for this disease adequate diagnostic markers with high specificity and sensitivity are still lacking. Despite advances in imaging technologies for early detection of cancer, peptidomic multiplex techniques evolved in recent years will provide new opportunities for detection of low molecular weight (LMW) proteome biomarker (peptides) by mass spectrometry. Improvements in peptidomics research were made based on separation of peptides and/or proteins by their physico-chemical properties in combination with mass spectrometric detection, respectively identification, and sophisticated bioinformatic tools for data analysis. To evaluate the potential of serological tumor marker detection by differential peptide display (DPD) we analyzed plasma samples from a tumor graft model. After subcutaneous injection of HCT-116 cells in immunodeficient mice and their growth to a palpable tumor, plasma samples were analyzed by DPD. The comparison of obtained mass spectrometric data allows discovery of tumor specific peptides which fit well into the biological context of cancer pathogenesis and show a strong correlation to tumor growth. The identified peptides indicate events associated with hyper-proliferation and dedifferentiation of cells from an epithelial origin, which are typical characteristics of human carcinomas. We conclude that these findings are a "proof of principle" to detect differentially expressed, tumor-related peptides in plasma of tumor-bearing mice.     

Fractionation of the human plasma proteome for monoclonal antibody proteomics‐based biomarker discovery 2: Antigen identification by dot‐blot array screening

Immunization with complex mixtures, like the human plasma resulted in the generation of cloned mAb libraries (PlasmaScan? and QuantiPlasma? libraries, with >1000 individual mAbs) reacting with a nonredundant set of antigenic epitopes. mAb proteomics refers to quasi‐hypothesis‐free profiling of plasma samples with nascent or cloned mAb libraries for the discovery of disease‐specific biomarkers. Once mAbs with biomarker potential have been identified, the next task is the determination of cognate antigens recognized by the respective mAbs. To determine the cognate protein antigen corresponding to each individual mAbs in the cloned mAb libraries, we have separated human plasma by consecutive steps of desalting and various chromatography procedures. The process resulted in 783 fractions, which we termed “Analyte Library” (AL). The AL represents the human plasma proteome in relatively low‐protein complexity fractions. Here, to determine the utility of the AL, we selected ten plasma proteins and checked for their presence in the fractions. Among the ten cases, the distribution of four selected plasma proteins matched expectations, as these proteins were present only in a few fractions corresponding to their physical, chemical, and biochemical properties. However, in six cases, we observed “smear” ‐like distribution or complete absence of the proteins, suggesting that protein–protein interactions or protein variants may alter the observed plasma distribution profiles. Nevertheless, we conclude that the AL is an efficient, high throughput tool to complement the mAb biomarker discovery process with cognate protein antigen identification for each mAbs.     

Quantitative analysis of the low molecular weight serum proteome using 18O stable isotope labeling in a lung tumor xenograft mouse model

With advancements in the analytical technologies and methodologies in proteomics, there is great interest in biomarker discovery in biofluids such as serum and plasma. Current hypotheses suggest that the low molecular weight (LMW) serum proteome possesses an archive of clipped and cleaved protein fragments that may provide insight into disease development. Though these biofluids represent attractive samples from which new and more accurate disease biomarkers may be found, the intrinsic person-to-person variability in these samples complicates their discovery. Mice are one of the most extensively used animal models for studying human disease because they represent a highly controllable experimental model system. In this study, the LMW serum proteome was compared between xenografted tumor-bearing mice and control mice by differential labeling utilizing trypsin-mediated incorporation of the stable isotope of oxygen, 18O. The digestates were combined, fractionated by strong cation exchange chromatography, and analyzed by nanoflow reversed-phase liquid chromatography coupled online with tandem mass spectrometry, resulting in the identification of 6003 proteins identified by at least a single, fully tryptic peptide. Almost 1650 proteins were identified and quantitated by two or more fully tryptic peptides. The methodology adopted in this work provides the means for future quantitative measurements in comparative animal models of disease and in human disease cohorts.     

Murine fecal proteomics: A model system for the detection of potential biomarkers for colorectal cancer

Tumor related products shed into the feces offer a potential source of biomarkers for the detection of colorectal cancer (CRC). Using SDS-PAGE followed by nanoflow reversed-phased LC–MS/MS to analyse fecal samples from Apc^Min/+ mice (that develop spontaneous multiple intestinal neoplasia with age) we have identified 336 proteins (115 proteins of murine origin, 201 from fecal bacteria, 18 associated with food intake and 2 of apparent parasitic origin). 75% of the murine proteins identified in this study are predicted to be extracellular or associated with the cell plasma membrane. Of these proteins, a number of the murine homologues of colorectal cancer associated proteins (CCAP) such as hemoglobin, haptoglobin, hemopexin, alpha-2-macroglobulin and cadherin-17 have been identified, demonstrating the potential of fecal proteomics for detecting potential biomarkers and paving the way for subsequent MS/MS based biomarker studies on similar human samples.     

A comprehensive and non‐prefractionation on the protein level approach for the human urinary proteome: touching phosphorylation in urine

Increasing attention has been paid to the urinary proteome because it holds the promise of discovering various disease biomarkers. However, most of the urine proteomics studies routinely relied on protein pre‐fractionation and so far did not present characterization on phosphorylation status. Two robust approaches, integrated multidimensional liquid chromatography (IMDL) and Yin‐yang multidimensional liquid chromatography (MDLC) tandem mass spectrometry, were recently developed in our laboratory, with high‐coverage identification of peptide mixtures. In this study, we adopted a strategy without pre‐fractionation on the protein level for urinary proteome identification, using both the IMDL and the Yin‐yang MDLC methods for peptide fractionation followed by identification using a linear ion trap‐orbitrap (LTQ‐Orbitrap) mass spectrometer with high resolution and mass accuracy. A total of 1310 non‐redundant proteins were highly confidently identified from two experiments, significantly including 59 phosphorylation sites. More than half the annotated identifications were membrane‐related proteins. In addition, the lysosomal as well as kidney‐associated proteins were detected. Compared with the six largest datasets of urinary proteins published previously, we found our data included most of the reported proteins. Our study developed a robust approach for exploring the human urinary proteome, which would provide a catalogue of urine proteins on a global scale. It is the first report, to our best knowledge, to profile the urinary phosphoproteome. This work significantly extends current comprehension of urinary protein modification and its potential biological significance. Moreover, the strategy could further serve as a reference for biomarker discovery. Copyright © 2010 John Wiley & Sons, Ltd.     

Protein expression patterns associated with advanced stage ovarian cancer

This is a comparative proteomic study on biopsies from patients with ovarian cancer to identify potential diagnostic/prognostic biomarkers in both healthy and tumor tissue, interstitial fluid (normal interstitial fluid and tumoral interstitial fluid and peritoneal effusion. Protein expression/identification was evaluated by 2-DE and MS analysis: six proteins showed differential expression in tumoral interstitial fluid and tumor tissue compared to normal interstitial fluid and healthy tissue: five were found to be downregulated and identified as galectin 3, glutathione S-transferase A-2, retinol binding protein 1, phosphatidylethanolamine-binding protein and annexin 5, while the calgranulin, was significantly upregulated in all pathological samples, including the ascitic fluid. Validation of S100A8 overexpression in carcinoma tissue was obtained by immunohistochemistry. To our knowledge, this is the first study to report an over-expression of calgranulin by 2-DE associated with MS/MS analysis on surgical biopsy. The reduced expression of galectin 3 and retinol binding protein 1 in cystic fluid and serum of patients with early stage disease is confirmed in this study. The results highlight alterations in proteins that control cell-cycle progression and apoptosis, as well as factors that modulate the activity of signal transduction pathways. Moreover, this study suggests that calgranulin expression may be used as a diagnostic and/or prognostic biomarker.     

基于气相色谱-质谱联用和图模型分析的脂代谢异常患者血浆代谢组学研究

利用气相色谱-质谱联用技术(GC-MS)和图模型分析方法,寻找脂代谢异常患者可能的血浆代谢标志物群。采用GC-MS技术对37例脂代谢异常患者和10例健康人的血浆样品进行分析,得到血浆代谢物的表达谱。偏最小二乘-判别分析(Partial least squares-discriminant analysis,PLS-DA)得分图可区分脂代谢异常患者与健康人,运用PLS-DA载荷图及t检验发现有9个代谢物在两组间存在显著性差异。经NIST谱库检索,它们分别为缬氨酸、甘氨酸、丙氨酸、焦谷氨酸、葡萄糖醛酸、半乳糖、甘露糖、亚油酸和甘油。在脂代谢异常患者血浆中,除甘油浓度显著高于健康人外,其余8种代谢物浓度均明显低于健康人。图模型分析结果发现这些代谢物与脂代谢异常临床常用诊断指标之间具有很好的相关性。它们可能是脂代谢异常疾病早期诊断和预后新的特异性代谢标志物群。     

2-D difference gel electrophoresis of the lung squamous cell carcinoma versus normal sera demonstrates consistent alterations in the levels of ten specific proteins

Most lung cancers are diagnosed too late for curative treatment to be possible, therefore early detection is crucial. Serum proteins are a rich source of biomarkers and have the potential to be used as diagnostic and prognostic indicators for lung cancer. In order to examine differences in serum levels of specific proteins associated with human lung squamous carcinoma, immunodepletion of albumin and five other high-abundant serum proteins followed by 2-D difference gel electrophoresis (DIGE) analysis and subsequent MS was used to generate a panel of proteins found to be differentially expressed between the cancer and normal samples. Proteins found to have increased abundance levels in squamous cell carcinoma sera compared to normal sera included apolipoprotein A-IV precursor, chain F; human complement component C3c, haptoglobin, serum amyloid A protein precursor and Ras-related protein Rab-7b. Proteins found to have lower abundance levels in squamous cell carcinoma sera compared to normal sera included alpha-2-HS glycoprotein, hemopexin precursor, proapolipoprotein, antithrombin III and SP40; 40. The data presented here demonstrate that high-abundant protein removal combined with 2-D DIGE is a powerful strategy for the discovery of potential biomarkers. The identification of lung cancer-specific biomarkers is crucial to early detection, which in turn could lead to a dramatic increase in survival rates.     

Use of magnetic beads with immobilized monoclonal antibodies for isolation of highly pure plasma membranes

In plasma membrane proteome research, contamination of the isolated plasma membrane fraction with proteins from other organelles is still a problem. Even if highly specific isolation methods are used, such as density gradient centrifugation combined with selective extraction, contaminating proteins cannot be completely removed. To solve this problem, a protocol for the isolation of highly pure plasma membrane fractions from rat liver and two different hepatocellular carcinoma cell lines was developed. Magnetic beads with immobilized mAb's against highly expressed membrane proteins were used for specific binding of membrane vesicles and their separation from other organelles. Isolated plasma membranes were further selectively solubilized with different reagents and analyzed by use of different methods, such as Western blotting, 1- and 2-DE, and MS. Purification and further selective solubilization was validated by use of mAb's against the marker integral plasma membrane protein carcinoembryonic antigen cell adhesion molecule 1, and identification of isolated proteins by MS. The method presented here minimizes contamination with other organelles and enables further identification of membrane proteins.     

Dissolving capability difference based sequential extraction: A versatile tool for in-depth membrane proteome analysis

Profiling membrane proteins would facilitate revealing disease mechanism and discovering new drug targets as they play essential roles in cellular signaling, substrate transport, and cell adhesion. However, the analysis of membrane proteins still remains a challenge due to their high hydrophobicity, as well as the suppression effect of high abundant soluble proteins. In this work, to achieve a membrane proteome profiling, a sample preparation strategy based on sequential extraction at the protein level assisted by a range of extraction reagents with different dissolving capabilities, followed by nano-RPLC-ESI-MS/MS analysis was developed and applied for HeLa cell line analysis. It was found that with progressively harsher extraction reagents (i.e., 2 M NaCl, 4 M urea, 0.1 M Na₂CO₃, and 10% 1-dodecyl-3- methyl-imidazolium chloride (C12ImCl) performed, much more high hydrophobic proteins and low abundant proteins were identified. With our developed strategy, 5553 of the identified proteins (4419 gene products) were annotated to be membrane proteins and 2573 proteins (2183 gene products) have at least one transmembrane domain, to our best knowledge, which is the most comprehensive membrane proteome dataset for HeLa cell line. Notably, 110 of the identified membrane proteins were discovered in the “missing proteins” list referred to those in the neXtProt database. All above results indicated that our strategy has great potential to tackle the difficult but relevant task of identifying and profiling membrane proteins.     

Novel multifunctional chitosan-GMA-IDA-Cu(II) nanospheres for high dynamic range characterization of the human plasma proteome

In this study, we describe characterization of the human plasma proteome based on analysis with multifunctional chitosan-GMA-IDA-Cu(II) nanospheres. Chitosan-GMA-IDA-Cu(II) nanospheres with diameters of 20 to 100?nm have unique properties due to multifunctional chemical moieties, high surface area, high capacity, good dispersibility in buffer solution as well as good biocompatibility and chemical stability which improves their specific interaction with peptides and proteins of the human plasma using different binding buffers. Combining these chitosan-GMA-IDA-Cu(II) nanospheres with MS spectrometry results in a novel strategy which should make it possible to characterize the plasma proteome in a single test. Peptides and proteins adsorbed on the nanosphere can be directly detected by MALDI-TOF-MS. The eluted lower molecular weight peptides and proteins are identified by nano-LC-ESI-MS/MS. A total of 842 unique LMW peptides and 1,682 human unredundant proteins IDs were identified in two different binding buffers, which included relatively low-level proteins (e.g., pg/mL of IL3 Interleukin-3) co-distributed with high-abundance proteins (e.g., 35?C55?mg/mL level serum albumin). As such, this nanosphere technique selectively enabled the identification of proteins over a dynamic range of greater than 8 orders of magnitude. Considering this capacity for selective enrichment of peptides and proteins in human plasma, and the large number of LMW peptides and proteins which can be identified, this method promises to accelerate discovery of biomarkers for clinical application.

Mining the human plasma proteome with three-dimensional strategies by high-resolution Quadrupole Orbitrap Mass Spectrometry

Human plasma is a readily available clinical sample that reflects the status of the body in normal physiological and disease states. Although the wide dynamic range and immense complexity of plasma proteins are obstacles, comprehensive proteomic analysis of human plasma is necessary for biomarker discovery and further verification. Various methods such as immunodepletion, protein equalization and hyper fractionation have been applied to reduce the influence of high-abundance proteins (HAPs) and to reduce the high level of complexity. However, the depth at which the human plasma proteome has been explored in a relatively short time frame has been limited, which impedes the transfer of proteomic techniques to clinical research. Development of an optimal strategy is expected to improve the efficiency of human plasma proteome profiling.     

An investigation into the human serum "interactome"

The protein content of human serum is composed of a millieu of proteins from almost every type of cell and tissue within the body. The serum proteome has been shown to contain information that directly reflects pathophysiological states and represents an invaluable source of diagnostic information for a variety of different diseases. Unfortunately, the dynamic range of protein abundance, ranging from > mg/mL level to < pg/mL level, renders complete characterization of this proteome nearly impossible with current analytical methods. To study low-abundance proteins, which have potential value for clinical diagnosis, the high-abundant species, such as immunoglobulins and albumin, are generally eliminated as the first step in many analytical protocols. This step, however, is hypothesized to concomitantly remove proteins/peptides associated with the high-abundant proteins targeted for depletion. In this study, immunoprecipitation was combined with microcapillary reversed-phase liquid chromatography (microRPLC) coupled on-line with tandem mass spectrometry (MS/MS) to investigate the low-molecular-weight proteins/peptides that associate with the most abundant species in serum. By this targeted isolation of select highly abundant serum proteins, the associated proteins/peptides can be enriched and effectively identified by microRPLC-MS/MS. Among the 210 proteins identified, 73% and 67% were not found in previous studies of the low-molecular-weight or whole-serum proteome, respectively.     

A novel technique to specifically analyze the secretome of cells and tissues

The secretome of cells and tissues may reflect a broad variety of pathological conditions and thus represents a rich source of biomarkers. The identity of secreted proteins, usually isolated from cell supernatants or body fluids, is hardly accessible by direct proteome analysis, because these proteins are often masked by high amounts of proteins actually not secreted by the investigated cells. Here, we present a novel method for the specific detection of proteins secreted by human tissue specimen as well as cultured cells and chose liver as a model. The method is based on the metabolic labelling of proteins synthesized during a limited incubation period. Then, the cell supernatant is filtered, precipitated, and subjected to two-dimensional gel electrophoresis. Whereas fluorography detected a large number of proteins derived from residual plasma and dead cells, the autoradiographs selectively displayed genuinely secreted proteins. We demonstrate the feasibility of this approach by means of the secretomes of the hepatocellular carcinoma-derived cell line HepG2 and human liver slices. The selective identification of cell- and tissue-specific protein secretion profiles may help to identify novel sets of biomarkers for wide clinical applications.