Introduction
The objective of this paper is to introduce a new topic in protein analyses and profiling in general, and to discuss on the potential practical implementation in detecting neurological diseases. A need for such a research occurs due to the arguable fact that new techniques and methods for protein analyses have been developing, and only a few have actually been implemented in routine clinical practice. This study is a part of a larger effort to develop methods in order to understand protein diversity and specificity of a human proteome in health and disease. Primarily, we will introduce and discuss some standard and already implemented techniques in clinical practice, and further we will focus on the new methods for cerebrospinal fluid (CSF) analysis as a primary biological media for study of neurological diseases.
Proteins in neurological diseases
Neurological conditions in patients occur as a result of a structural or biochemical abnormality in the brain, spinal cord or in the nerves leading to and from them. After the onset of symptoms, additional parameters are analyzed in order to distinguish between different, so called, patterns and contribute to the diagnosis of a disease. Protein profiling has been introduced in clinical practice 100 years ago (W. Bruno 1956; Vesterberg 1989). Cerebrospinal fluid is the media used for such investigations, due to its close proximity to the brain; therefore being a biological fluid that can potentially obtain first-hand information about the potential causes of the induced change. Changes usually occur either in the CSF physical appearance, or can be manifested through the intrinsic characteristics – mainly the protein content and composition. Protein profiling in neurological diseases, therefore,contributes to the overall clinical diagnosis. However, in order to acquire the complete clinical state in a patient one must include comparative protein analysis in the other biological fluids (serum, plasma or urine). Since proteins in CSF are derived from the plasma, changes in the composition or concentration in either of these media will result in changes in the other (Andersson, Alvarez-Cermeno et al. 1994). Total protein concentration reveals only little about the function of blood-brain barrier and exploits changes only after the onset of symptoms related to a neurological disorder. Also, there are neurological disorders, among others multiple sclerosis, which are characterized with normal or moderately increased total protein concentration (although changes in the individual protein concentration occurs) (Link and Muller 1971). This serious neurological disorder, and many others, mostly related to demyelization processes in the CNS; do not exhibit changes in total protein concentration (Hein Nee Maier, Kohler et al. 2008). Other neurodegenerative disorders such as Alzheimer disease (AD), Parkinson disease (PD), or amyotrophic lateral sclerosis (ALS) have different causes and affect different regions of the nervous system, and are characterized by abnormal accumulation of protein aggregates and organelles along the axon due to disruption of axonal transport. As a consequence, an early pathogenic event leading to the demise of neurons occurs (Chevalier-Larsen and Holzbaur 2006; De Vos, Grierson et al. 2008). This remark only contributes to the importance of analyzing proteins and protein profiles in such diseases.
In order to achieve our goal and to be able to correlate proteins to certain disease, we must first address to understanding the origin of the complexity of the proteome in human cells and body fluids. The importance of studying cells at the proteome level is underscored by the difficulty in predicting protein characteristics from genomic sequence data alone. These characteristics include post-translational modifications, subcellular distribution, stability, biomolecular interactions, and function. In contrast to DNA and RNA, proteins can be modified by phosphorylation, glycosylation, acetylation, nitrosation, poly(ADP ribosylation), ubiquitination, farnesylation, sulfation, linkage to glycosyl-phosphatidylinositol anchors, and SUMOylation. In total, there are about 300 different posttranslational modifications that have been reported. These modifications can profoundly affect protein conformation, stability, localization, binding interactions, and function (Aebersold and Goodlett 2001). Separate media further complicate the application of protein analyses. Cerebrospinal fluid is a hurdle media for protein analyses for one basic reason – limited amount of starting material that is often insufficient to carry out all the analysis required. Additional reasons for the difficult analysis of body fluid proteome is their large dynamic range reflected by the presence of very abundant proteins like albumin, and in the case of CSF, minute quantities of brain-derived proteins (Anderson and Anderson 1998). Further complicating the analysis of CSF is the possible infiltration of serum proteins that is caused by a leaky blood-brain barrier that is especially pronounced in patients with brain disorders. As a consequence, it is often impossible to know if a protein that is found in CSF is derived from the brain or serum. And not only proteins fitting in a specific family, but the separate protein fractions, and moreover, the individual proteins and even protein variants, produce useful information about the presence of all types of disorders. Considerable fact is that additional modifications in proteins occur after translation of the information written in the genes. Having this remark in mind, it is obvious that the proteome is practically more related to the phenotype of an individual, and hence, protein profiling will result in the most precise understanding of disease mechanisms as well as the molecular effects of drugs (Turck, Maccarrone et al. 2005). The ultimate goal of proteomics in medicine, and its sub disciplines is therefore, to provide quantitative and qualitative data of sample proteins that reflect a certain phenotype, disease state or a response to disease treatment.
Methods for protein analysis
Analyzing proteins as a diagnostic parameter in clinical practice was introduced with the development of the traditional techniques that combined methods based on protein precipitation and colored reactions. Since the introduction of electrophoresis in routine protein profiling in patients with neurological diseases, the main goal is developing novel methods for protein analyses that will offer more information and onset into the intrinsic protein characteristics. In recent years, almost every approach used to describe a method or technique applied to study protein diversity is termed "proteomics" (Nedelkov 2005). Analyses in proteomics, especially when human plasma proteome is considered, are technically challenging because the circulating proteome is a complex mixture of diverse proteins that spans approximately 10 orders of magnitude in concentration (Anderson and Anderson 2002). Routine techniques in protein profiling include mainly electrophoretic methods and can be distinguished as: gel-based proteomics approaches, fractionation or separation with other media and support material, which are based toward the identification of abundant proteins. For these types of assays, experimental designs need to involve enrichment strategies such as immunoisolation of protein complexes of interest to reduce sample complexity and increase sensitivity of detection.
The development of non-gel-based approaches for quantitative proteomics, together with advances made to detect posttranslational modification guide progress toward delineating the mechanisms involved in nerve regeneration and degeneration dysfunctions. Such neuroproteomics approaches lay the foundation for further detailed functional studies (Sun and Cavalli 2010). Both of these approaches relay on a two stage mechanism of protein analysis: protein separation followed by identification and analysis (Turck, Maccarrone et al. 2005). Classical proteomic approaches employ fractionation on the protein level with the help of 2D-PAGE. This technique produces high resolution protein separations resulting in the display of potentially thousands of protein spots. Identification of these spots is rather difficult and requires additional informatics-based approaches in order to derive relevant conclusions. Some of the limitations of 2D-PAGE have currently been overcome with the implementation of 2D-DIGE (Difference Gel Electrophoresis) where fluorescent dyes are used to distinguish between proteins from control samples and one from individuals with certain disease (Alban, David et al. 2003). It is therefore evident that at this point in proteomic technology development the broadest proteome coverage comes from a combination of multidimensional fractionation, advanced instrumentation and additional computational techniques.
Introducing mass spectrometry in protein analysis contributes towards overcoming several of the disadvantages common for standard techniques. The most efficient and most widely used protein identification method in proteomics are MALDI-TOF-MS, SELDI-TOF-MS, ESI-MS/MS and other methods (Lahm and Langen 2000; Issaq, Veenstra et al. 2002). Using shotgun mass spectrometry approach proteins are digested by specific enzymes into small peptides and analyzed on-line by MS. This technique allows low abundant proteins to be identified despite the presence of high abundant proteins in the samples. Further advantage was done by Aebersold and co-workers (Gygi, Rist et al. 1999). They have developed a novel isotope-coded affinity tag (ICAT) strategy that permits the stable-isotope labeling of cysteine residues in proteins, thus facilitating a quantitative global analysis of differences in protein expression. A more common approach is combining chromatography separation (usually liquid chromatography, since proteins are high molecular mass compounds) with mass spectrometry detection. Chromatographic methods reduce the complexity of protein mixtures on the basis of different binding principles, and every approach adds a unique resolving power. The proteins are usually separated on the basis of affinity, charge, hydrophobicity, or size (Fountoulakis, Takacs et al. 1999; Takacs, Rakhely et al. 2001). The choice of the chromatographic method best-suited to fulfill the experimental requirements is essential for the success of the experiment (Fountoulakis and Takacs 1998). There are several methodologies that use liquid chromatography (mainly HPLC) as a tool for protein separation. Basic LC-MS/MS techniques generally employ ion-pair reversed-phase chromatography followed by electrospray ionization and mass spectrometry analysis. Depending on the ionization technique, LC -MS/MS coupling can be done either "on-line" (where HPLC is directly coupled to the ESI source of the tandem MS) or "off-line" (where, after being resolved in the HPLC process, samples are deposited directly onto a MALDI target, mixed with corresponding matrix and analyzed via MALDI-TOF MS) (Mallick and Kuster 2010). In the last 15 years combined approaches coupling immunoaffinity separation and mass spectrometry detection have been implemented (Nedelkov 2006). Also, a new approach that is considered promising in the efforts of proteome analysis is the use of aptamers as an alternative class of reagents to use for highly multiplexed protein measurements (Brody, Gold et al. 2010).
All of the above mentioned protein composition studies can be extended to two main approaches – analyzing either individual proteins or obtaining protein profile ("group" protein analysis).
Protein profiling approach
In the "group" protein analysis, termed as protein profiling, our main focus is to detect and identify as many proteins as possible, perform qualitative analyses, and obtain protein pattern for each sample analyzed (Fig.1/a).
Electrophoresis techniques have been the hub of laboratory testing and a component of almost all diagnostic panels (McCudden, Voorhees et al. 2010). A rather intriguing fact is however, that electrophoresis has come only a short way in advancing during the years, when used for protein profiling in patients with neurological diseases. Since it was implemented as a method in routine clinical practice, electrophoresis has come only a short way in actually progressing. In routine clinical practice discontinuous polyacrylamide gel electrophoresis (DISC-PAGE) is usually applied for determining the protein profile (Monteoliva and Albar 2004). More detailed information can be obtained using denaturing sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) (Hu, Huang et al. 2004). In diagnosing some specific neurological diseases, isoelectric focusing (IEF) and immunofixation can be used to confirm the presence of specific oligoclonal, monoclonal or polyclonal immunoglobulin. Determination of these protein components is in many cases necessary to delineate between differential diagnoses. By combining IEF with gel electrophoresis, two dimensional gel electrophoresis is created (2D-GE), a technique, which is probably the most frequently used methodology for protein profiling. When applied into clinical practice, this methodology is used to analyze the complete proteome present in tissue samples, biological fluids and cell lysates. The result is an electrophoregram (for 1D-GE techniques) or pattern plot (for 2D-GE) fingerprinting the specific protein composition in the analyzed sample.
Fig. 1. Methods for protein analyses a) protein profiling using microchip electrophoresis, b) single protein analysis using mass spectrometry immunoassay (MSIA)
The general goal when using these types of qualitative group approaches is obtaining complete protein profiles both, from control samples and samples from patients with different (neurological) diseases. By comparing the obtained protein profile from the control subject (normal reference sample) to a profile originating from a patient we are able to construct a certain pattern based on the principal of similarity analysis (Trenchevska, Aleksovski et al. 2009). In our work, using this approach, we have been able to implement microchip electrophoresis method to distinguish a specific pattern for several neurological diseases (multiple sclerosis, blood-brain barrier dysfunction diseases, inflammatory conditions and autoimmune disorders) which will be discussed further in details.
Single protein assays
Although, the classical proteomics efforts are geared towards the analysis of the protein constitutes in both CSF and serum (Maccarrone, Birg et al. 2004; Maccarrone, Milfay et al. 2004), alternative strategies have been established for the identification of novel disease markers for neurological disorders. When it comes to protein analyses, often the concentration plays a significant role in contributing to diagnosis (example – determination of IgG concentration in multiple sclerosis). However, quantitation of all proteins in a sample is a Herculean task itself, due to the fact that the outcome of a proteome profiling experiment results in a content-rich and convoluted data that require powerful approaches to discover the subtle differences between the healthy and diseased samples (Figure 1a). In this regard, mass spectrometry has found its way as a sensitive technique for quantitative protein analyses. Although mass spectrometry has been a golden standard for proteomics from the very beginning, its complexity, cost, maintenance and other analytical characteristics are the underlying reason why this technique is rarely used in routine clinical practice (Rabilloud 2002). A step forward has been done with introducing immunoaffinity approaches (Engvall and Perlmann 1971; Ritchie 1999; Craig, Ledue et al. 2001; Yan, Lee et al. 2004). The pilot work has been done with introducing ELISA as a technique for quantification single proteins from biological samples. This technique is currently the most common abandoned method used in routine screening of known proteins and is commonly used as a laboratory test for detecting antibodies or rapidly screen and quantify antigens in biological samples. Several advantages have made this technique favorable; it follows simple sample preparation, it does not require specific conditions and data acquisition is relatively simple (Dufva and Christensen 2005). However, it can only be applied in analyzing known proteins, and, what is more, a suitable antibody must exist for retrieving the antigens from the sample. Also, ELISA based methods determine the total protein concentration and are lacking the ability to detect, identify or quantify protein variants, post-translational modifications or point mutations. Therefore, advances in this area are directed in developing assays that use combined techniques in order to obtain more detailed information about each protein intrinsic characteristics. Combined techniques have been conceptualized in the mid-nineties, via the SELDI (Hutchens and Yip, US patent No.5,719,060) (Engwegen, Gast et al. 2006; Poon 2007), and later by MSIA (Nelson et a!., US patent No.6,974,704) (Nelson, Krone et al. 1995; Krone, Nelson et al. 1996) approach. The newly developed methodologies are the platform for biomarker discovery assays. These assays target "individual" proteins and use specific antibody towards the protein of interest. These specific proteins from the proteome are considered to be potential diagnostic markers for certain neurological diseases. MSIA methodology, which will be further discussed in details, bridges between the selectivity that can be obtained with immunoassays and the specificity of mass spectrometry detection (Fig. 1/b).
Electrophoresis in protein profiling
Identification and determination of different types of proteins play an important role in medical diagnosis. Conventional electrophoresis methods are well known for protein detection and analysis in several biological fluids, primarily blood serum and plasma and cerebrospinal fluid. CSF analysis, coupled with other methods, remains the cornerstone of diagnosis of various neurological diseases, including multiple sclerosis (Reiber, Otto et al. 2001; Sindic, Van Antwerpen et al. 2001). Even though electrophoresis is practically the basic technique in protein profiling, both in routine and scientific clinical practice, it provides only semiquantitative information about the concentration of a certain protein. Regarding the protein profiling, these techniques still offer significant amount of information, especially about the abundance of specific protein classes, such as immunoglobulins. This is of great importance in demyelization neurological diseases. Additional criteria for this and other types of neurological diseases are presence of oligoclonal immunoglobulins. For detection of oligoclonal IgG in serum and unconcentrated CSF, several techniques can be used, primarily isoelectric focusing (IEF) combined with polyethylene-enhanced gel immunofixation and silver staining, CSF:serum quotient diagram and body index (Tourtellote, Povin et al. 1980; Sandic, Monteyne et al. 1994; Mitrevski, Stojanoski et al. 2001). Combining IEF and PAGE, 2D-GE methods are developed, optimizing techniques for detailed protein mapping, as mentioned previously. Introducing bioinformatics tools is further implemented in order to obtain more information from the protein profile. This approach necessitates usage of protein standards in each run. Using additional computational programs and following the advances in the field of bioinformatics, more information can be obtained from each protein profile spectra.
