In order to study mitochondrial proteins, mitochondria have to be isolated first and purified from proteins of other cell compartments. Such a sub-fractionation allows the detection of those proteins which would be invisible in total cell lysates for their low-abundance [Lopez et al., 2000].We selected Epstein-Barr-Virus (EBV) transformed lymphoblastoid cell line samples because they have several advantages compared to other biopsy specimens: they are easily obtained from patients and can grow permanently in liquid, non-adherent cell cultures. This allows the cultivation of large cell numbers without excessive work.
The initial step in purifying mitochondria is to rupture the cell membrane. There are various methods to disrupt cells in order to release their mitochondria. They can be ground, subjected to osmotic shock or to ultrasonic vibration or they can be forced through a small orifice. There are two “classic” methods for mitochondrial isolation:
The second step of mitochondrial isolation is to retrieve mitochondria from the mixture of subcellular components. Both differential-velocity centrifugation and density-gradient centrifugation are commonly used for this purpose.
Mitochondrial proteins can be separated into two main groups [Klose, 1999a]: the hydrophilic proteins including the proteins of the matrix and of the inter-membrane space; the hydrophobic proteins such as transmembrane proteins of the inner- and outer-membrane. Since only proteins in solution can be analyzed, several detergents and ultrasonication are used to solubilize the hydrophobic (membrane) proteins [Harvey et al., 1999].
To measure the total protein content of a sample, several protein assay methods are routinely used.
Methodical approaches to separate proteins electrophoretically by two different principles in order to improve resolution can be traced back to 1956 when Smithies and Poulik developed a two-dimensional (2D) electrophoresis technique combining filter paper electrophoresis (first dimension) and starch gel electrophoresis (second dimension) [Smithies et al., 1956]. In the following years, a number of other 2D-electrophoresis methods were developed by combining various electrophoretic techniques. The current modern 2D-electrophoresis technique was developed independently by Klose (1975) and O’Farrell (1975). They combined isoelectric focussing (first dimension) with SDS-polyacrylamide gel electrophoresis (second dimension). This method separates proteins firstly according to their isoelectric points (pI) and secondly according to their molecular weights (MW). Each protein can then be attributed a p I and a [page 13↓] MW. The 2D-electrophoresis method allows the visualization of thousands of protein-spots at a time, even up to total of 10,000-15,000 protein spots in a single large gel [Klose et al., 1995]. Fig. 2-1 shows the principle of the 2D-electrophoresis method.
Throughisoelectric focussing (IEF) proteins are separated according to their isoelectric points (pI). The pI depicts the pH-value at which the net charge of the protein is zero. The mobility of a protein in an electric field depends on the sum of its positive and negative charges. When the net charge of the protein is zero, the protein stops migrating in the electric field. It focusses where the pH of the gel equals the pI of the protein. A pH gradient can be established by adding a mixture of ampholytes with different isoelectric points to a polyacrylamide gel. The protein mixture can then be loaded either on the anodic or on the cathodic end of the gel. Since some very basic proteins may not migrate into the gel if the proteins were loaded on the cathodic end of the gel, we choose to load our samples on the acid side of the IEF-gel [Klose, 1975, 1995].
|Fig. 2-1: Principle of 2D-electrophoresis . In 2D-electrophoresis a complex protein mixture can be separated by two biochemical principles. In the first dimension isoelectric focussing (IEF) the proteins are separated according to their isoelectric points (pI ), e.g. proteins run in an electric field as long as the surrounding pH differs from their pI . If they reach their pI , their net charge is zero and they stop running in the electric field. In the second dimension, proteins are separated according to their molecular weights (MW ) in a SDS-polyacrylamide gel.|
The SDS-polyacrylamide gel electrophoresis (SDS-PAGE) utilizes SDS as an anionic detergent. SDS forms complexes with proteins and dissociates them into their individual subunits. This combination leads to two results: the ratio of SDS/protein remains sufficiently constant [page 14↓](1.4 g SDS per gram protein). Thus the complexes have a highanionic charge density which is much higher than the charge density of an individual protein. Therefore the charge difference between proteins can be ignored. Due to the same charge/protein ratio at pH 8.4, all SDS-protein complexes migrate to the cathode if an electrical field is applied. Secondly, since all SDS-protein complexes have a similar cylindrical form with a constant diameter (about 1.8 nm) but different lengths, the sizes of the proteins are directly proportional to their molecular weights. Thus the electrophoretic mobility of the SDS-protein complexes depends only on their molecular weight, i.e. the mobility of the proteins is little influenced by any individual protein feature such as charge or conformation [Weber et al.,1969; Laemmli, 1970].
To visualize the protein spots on the gel, the gel has to be stained. If the protein abundance is high (i.e. more than 100 ng), the gel can be dyed with Coomassie brilliant blue. For the detection of lesser protein amounts different silver staining protocols have been developed [Rabilloud, 1990 and 1992; Swain et al., 1995; Klose et al., 1995]. Compared to the commonly used Coomassie brilliant blue staining, silver staining is more sensitive and has an improved detection limit of 1-10 ng. Moreover, the sensitivity of silver staining can be improved further by the use of several sensitizers. These sensitizers act via different chemical mechanisms: increasing the binding of silver (sulfosalicylic acid), creating latent images of spots by precipitation of micro-granules of silver sulfide (sodium thiosulfate, dithiothreitol), promoting silver reduction (glutaraldehyde) and complexing free unbound silver cations (chelators). On the other hand, the silver staining techniques treat proteins with the strong oxidizing agent Ag+ that may cause oxidative damage to the proteins. This can lead to chemical modification or destruction, and subsequent protein microanalysis will be rendered impossible. Several sensitizing pre-treatments of the gel with glutaraldehyde, chromic acid, sodium thiosulfate or thimerosal could even result in covalent modifications of the proteins. Shevchenko et al. (1996a) have tried to solve this problem by modifying the ”classic” silver-staining protocols. They omitted the fixation and sensitization treatment with glutaraldehyde that is known to attach covalently to the protein through Schiff base formation with the α- and ε-amino groups. Additionally they carried out the silver nitrate treatment at 4°C in order to minimize oxidation.
Reproducibility means that if the same sample is run on two or more different two-dimensional gels, each spot on one gel must have its corresponding spot on another gel. Reproducibility is influenced by many factors ranging from sample preparation, stability of electrophoresis conditions and temperature to gel staining and drying [Klose, 1975; O’Farrell, 1975]. Occasionally, some single spots can change their positions within well-reproduced patterns. This phenomenon can be caused by some proteins with specific variable sensitive properties. The problem may be solved by using optimum conditions, i.e. running the same sample twice, side by side and by using the same batches of solutions at each step [Klose, 1995]. The reproducibility of the 2D-electrophoresis is reliable enough that it can be used to detect genetic variations by demonstrating the qualitative and quantitative changes of protein spots [Klose, 1995; Klose et al., 2002].
In order to connect the information from proteome analysis to the corresponding genes, it is necessary to identify the protein spots. Mass spectrometry is becoming more and more impor[page 15↓]tant in the field of protein analysis. It is proving useful for identification of proteins separated by two-dimensional protein electrophoresis. The particular advantage of mass spectrometry is that it generally requires a limited amount of material. Sometimes even femtomoles are sufficient [Perrot et al., 1999]. The most commonly used technique for protein identification by mass spectrometry is called “peptide mass fingerprinting”. This involves the generation of peptides from a protein by a proteolytic enzyme such as trypsin. The masses of the ensuing peptides are determined by mass-spectrometry and are matched against a theoretical list of peptide fragments calculated from databases of known protein sequences [Pennisi, 1997; ExPaSy database]. As peptide mass fingerprinting has a higher sample throughput than amino acid sequencing, it is especially suitable for rapid protein identification. (Fig 2-2)
|Fig. 2-2 : Principle of peptide mass fingerprinting by trypsin digestion . Trypsin cleaves at the carboxylic side of arginine and lysine residues. The sizes of the peptide fragments obtained after trypsin digestion, represent the peptide mass fingerprint and are characteristics of each protein.|
Peptide mass fingerprinting by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry is one tool for protein identification [Fernandez et al., 1998]. The principle of this technique is depicted in Fig. 2-3. After in-gel digestion, the protein is cut into several peptide fragments by proteases, such as trypsin or chymotrypsin. This peptide-mixture is mixed with a matrix of 2,5-dihydroxybenzoic acid or α-cyano-4-hydroxycinnamic acid and is let to crystallize. Subsequently the surface of the peptide/matrix mixture is evaporated and ionized by the photons of a high-energy laser beam. The ions are then accelerated in an electric field and fly towards a target. The speed and therefore the time of flight depends on the mass/charge ratio. The time of flight can thus be used to exactly measure the molecular weight of each peptide fragment up to the precision of 0.1 Dalton. At first this method was a rather unspecific identification tool but rapidly improved with the advent of machines that were able to measure also high molecular weight fragments with sufficient accuracy. New matrix preparations and higher sensitivity led to higher sequence coverage [Fernandez et al., 1998]. In favorable cases a mass coverage of over 90% of the peptide fragments can be achieved. The high accuracy in mass determination is made possible by the “delayed extraction” method [Jensen et al., 1996]. Until now, bio-macromolecules with molecular masses up to 300 kDa can be identified by peptide mass fingerprinting [Nielsen et al., 2002]. This technique has been developed to an extent thathigh throughput analyses are possible, and it has a firm place for protein identification in proteomic projects (see section 2.4.3).
|Fig. 2-3 : Principle of MALDI-TOF (matrix-assisted laser desorption/ionization time-of-flight) mass spectrometry . The peptides in the sample are ionized and energized by a laser-beam and are accelerated in an electric field. The time of flight (=TOF) of each peptide fragment towards a target is measured. Since the TOF is proportional to the mass/charge ratio of each peptide, the mass of the peptide can thus be calculated. This way the mass spectra of the peptide fragments of a whole protein can be obtained.|
|Fig. 2-4: Principle of MALDI-QTOF (quadrupole time-of-flight) tandem mass spectrometry . The sample is ionized and energized by a laser-beam and flies into a quadrupole ion guiding cell (Q0 ), where the ions are focussed and cooled. Then a peptide of interest (the parent ion) is selected at the quadrupole Q1 -cell and guided into the quadrupole collision cell (Q2 ). There the parent ion collides with argon atoms and splits into daughter-ions. The masses of these daughter ions are then measured via TOF mass spectrometry.|
The peptide mass data from MALDI-TOF mass spectrometry were used for database searches in order to identify the target protein by peptide mass fingerprinting. Sometimes no positive hit was found or the result did not satisfy the specified stringency criteria, which usually required more than four matched peptides at an accuracy of 0.1 Da. In this case, the peptide sequence of one or two abundant fragments was be determined directly by MALDI-QTOF tandem mass spectrometry. The full amino acid sequence of a peptide provides much more accurate information for further protein identification. It can also be used to confirm the results gained from peptide mass fingerprinting. MALDI-QTOF tandem mass spectrometry uses the MALDI ion source. The quadrupole filters peptides within a selected size range that are later guided into a collision cell to be broken into smaller fragments. The masses of these overlapping fragments are then analyzed by time-of-flight mass spectrometry (Fig. 2-4). Although this method is more complicated to perform, the results are more reliable than peptide mass fingerprinting because the sequence information of one or two peptides usually identifies the protein with high accuracy. Before the analysis with MALDI-QTOF mass spectrometry samples have to be purified by nano-scale reversed chromatography. This procedure removes salts and small chemical compounds of the buffer and thus reduces the chemical noise of the spectra and improves the sensitivity (Annan et al.,1996; Gobom et al., 2001).
The obtained spectra of peptide masses are analyzed further by searching though different databases to find the corresponding protein. Each protein in the databases can be “digested in silico” by trypsin and thus provides a theoretical spectrum of its peptide masses. Comparing the experimental with the bioinformatic data, several candidate proteins with high probability scores can be identified. This task can be performed with the help of several search engines on the internet: MASCOT, ProFound, MS-Fit, PeptIdent, PeptideSearch, and PepSea (see “The list of internet sites”).
Before the search, several parameters have to be set. These include the taxonomy of the specimen, the used protease and the number of accepted missed cleavages, the peptide mass states (usually the monoisotopic mass),the mass deviation tolerance, and possible modifications. Peptide modifications are important since they influence the peptide masses and might be introduced artificially in the preparation process (e.g. oxidation of methionine-residues). The oxidation of an amino acid (e.g. methonine) in a polypeptide increases its mass by 18 Da. The number of 18 Da deviations should therefore correspond to the number of methionine residues in a certain peptide. On the other hand, this “artifact” may serve as a second independent verification of the identity of a peptide. The search engine gives out a list of best matching proteins.
In general not all peptide masses in the spectra can be matched to the theoretical digestion of a protein. However, the larger the sequence to be covered by the fragments, the more statistically probable the result will be. Deviations between the theoretical peptide mass fingerprint and the experimental one might be due to the following reasons:
Sometimes several less stringent criteria, such as more than one allowed missed cleavage, several kinds of possible modifications, larger mass deviation tolerances have to be granted in order to match the experimental peptide mass spectra to their theoretical ones. If this is not possible, it is advisable to sequence an abundant peptide in the spectrum by MALDI-QTOF tandem mass spectrometry.
The following databases are used for confirmation of the protein matches: PeptideMass, BLAST2SEQUENCE, Swiss-Prot-TrEMBL, and NCBI (see “The list of internet sites”).
|DiML DTD Version 3.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin