Automation of micro-preparation and enzymatic cleavage of gel electrophoretically separated proteins.

To achieve high throughput, protein microcharacterization sample preparation must be automated. We describe a cartesian robot capable of processing 32 protein samples in parallel. The system is based on specially designed flow-through reactors for contamination-free reagent delivery and removal. Washing of excised gel pieces, reduction and alkylation, proteolytic cleavage and peptide extraction are performed in these reactors. Compatibility of the system with HPLC peptide separation and Edman degradation as well as with laser desorption mass spectrometry of the unseparated mixture is demonstrated. This is the first report describing automated preparation and processing of multiple protein samples.

Techniques for sample preparation including methods for concentrating peptide samples.

In the current era of proteomics two main analytical techniques are employed for protein identification. By far the fastest and most sensitive procedure for protein identification employs biological mass spectrometry, while de novo sequence analysis by classical Edman degradation is currently diminishing. In order to achieve the highest sensitivity for both techniques, great demands need to be put on sample preparation. In this paper we review three different aspects of protein sample preparation. Firstly, we discuss the use of polyacrylamide or agarose gel systems in which, during electrophoresis, proteins present in multiple primary gel pieces are eluted and simultaneously concentrated in a small secondary gel volume, whereby the overall sensitivity of Edman sequencing can be greatly increased. In a second chapter we review automation strategies occurring in the protein field which allow the automatic handling of multiple protein spots at the same time. In this context, we describe the use of auto-sampling techniques for further mass spectrometric studies and protein digestion robots allowing the simultaneous preparation of tens of gel-separated proteins. Finally we discuss various strategies for the preparation of biological peptide samples such as protein digests for both matrix-assisted laser desorption ionisation and electrospray ionisation mass spectrometry.

Amino acid analysis and protein database compositional search as a rapid and inexpensive method to identify proteins.

The identification of protein samples in minute quantities of protein samples, e.g., from two-dimensional polyacrylamide gel electrophoresis analysis, is an everyday problem in biology laboratories. Here we show that computer-assisted amino acid analysis can fulfill this task. Amino acid analysis data can be used to compare the amino acid composition of an unknown protein with protein compositions in a database (compositional search). Routine amino acid analysis data can, despite a certain margin of error, be used to identify a protein. Compared to protein sequencing, amino analysis is much cheaper, faster, and allows higher sample throughput. Thus, the method may replace protein sequencing as a first attempt in identification, provided a homolog can be found in the database.

C-terminal phosphorylation of the serum-response factor.

The serum-response factor (SRF) is essential for the induction and repression of the protooncogene c-fos. Phosphorylation of SRF has been implicated to be involved in these processes and five phosphorylation sites have already been mapped within the N-terminal region. Here we show that in vivo additional phosphorylation of SRF does occur. This modification is located primarily within amino acids 206-289, which probably contain more than one phosphorylation site. Microsequencing allowed the identification of one phosphorylation site at Ser253, which is a potential target of casein kinase II. Mutational analysis revealed that, in contrast to N-terminal phosphorylation, Ser253 phosphorylation does not affect DNA-binding properties. In addition, phosphorylation at Ser253 does not seem to change transactivation activity of SRF but rather influences its contribution to transcriptional repression. Thus, C-terminal phosphorylation of SRF may modulate c-fos basal repression.

Stability and proteolytic domains of Nef protein from human immunodeficiency virus (HIV) type 1.

Proteolytic experiments in conjunction with 1H-NMR spectroscopy show that the Nef (negative factor) protein from human immunodeficiency virus type 1 probably consists of two main domains, the N-terminal anchor domain at amino acid positions 2-65 and the C-terminal core domain at positions 66-206. The N-terminal domain is likely to be located at the surface of the protein, while the C-terminal domain has a compactly folded core and is stable in the absence of the anchor domain. It is conceivable that the core domain represents a functional domain of the Nef protein, activated after the removal of the membrane anchor by the human-immunodeficiency-virus protease or cellular proteases. Nef is stable at pH 5-12 and denatures at 317-322 K. The Nef protein remains in its native conformation in dimethyl-sulfoxide/water mixtures up to 35% (by vol.), and in acetonitrile/water up to 14% (by vol.). Nef refolds spontaneously after denaturation with urea or guanidinium hydrochloride. The 1H-NMR parameters and pKa values of five of the nine histidine residues and one of the seven tyrosine residues were determined and were found in four cases to be typical for residues which are not located in the interior of the protein.

Identification of multiple SRF N-terminal phosphorylation sites affecting DNA binding properties.

Human serum response factor (SRF) bearing a histidine tag was expressed using vaccinia virus. The recombinant protein was purified and shown to be phosphorylated mainly in its N-terminal part. The corresponding phosphorylation sites were mapped by microsequencing and also appear to be phosphorylated in endogenous serum response factor. Four phosphorylation sites are located on serines within amino acids 77-85, while another phosphorylation site has been identified at Ser103. Mutations that considerably reduced or abolished phosphorylation at amino acids 77-85 caused a decrease in binding to the c-fos serum response element accompanied by markedly reduced association and dissociation rates. In contrast, replacing Ser103 by alanine decreased DNA binding activity without drastically affecting the on/off rates. The combination of abolishing phosphorylation at amino acids 77-85 and 103 displayed greatly reduced on/off rates of DNA binding, but the reduction of DNA binding activity was partially alleviated. None of these mutations affect either the ability to interact with p62TCF or stimulation of transcription in vitro. These findings imply possible roles for SRF phosphorylation in the regulation of c-fos transcription.

Automated protein preparation techniques using a digest robot.

Since the introduction of fast analysis methods for peptide mixtures such as MALDI-MS, peptide micropreparation and digest methods have become an important bottleneck in the protein characterization process. We therefore developed and describe here a digest robot capable of processing 30 protein samples in parallel [Houthaeve et al. (1995), FEBS Lett. 376, 91-94]. Briefly, after gel pieces or blots are cut out, they are loaded in flowthrough reactors and these are loaded in a thermocontrolled reactor block. The proteins are then washed, reduced, and alkylated, proteolytically or chemically cleaved, and resulting peptides extracted. The system allows the parallel use of different reagents and enzymes during the same run, and is compatible with RP-HPLC peptide separation and Edman degradation, MALDI-MS, and NanoES-MS/MS. The digest robot is now also commercially available from ABIMED. In an ongoing project aimed at elucidating proteinaceous structures involved in the functional and structural maintenance of the Golgi apparatus, we illustrate the strength of the digest robot for the fast analysis of several proteins. We conclude that the performance of the digest robot is comparable to currently used manual digestion methods. The approach outlined makes sample preparation procedures faster, simpler, and less labor-intensive.

Femtomole sequencing of proteins from polyacrylamide gels by nano-electrospray mass spectrometry.

Molecular analysis of complex biological structures and processes increasingly requires sensitive methods for protein sequencing. Electrospray mass spectrometry has been applied to the high-sensitivity sequencing of short peptides, but technical difficulties have prevented similar success with gel-isolated proteins. Here we report a simple and robust technique for the sequencing of proteins isolated by polyacrylamide gel electrophoresis, using nano-electrospray tandem mass spectrometry. As little as 5 ng protein starting material on Coomassie- or silver-stained gels can be sequenced. Multiple-sequence stretches of up to 16 amino acids are obtained, which identify the protein unambiguously if already present in databases or provide information to clone the corresponding gene. We have applied this method to the sequencing and cloning of a protein which inhibits the proliferation of capillary endothelial cells in vitro and thus may have potential antiangiogenic effects on solid tumours.

A peptide concentration and purification method for protein characterization in the subpicomole range using matrix assisted laser desorption/ionization-postsource decay (MALDI-PSD) sequencing.

We here describe the use of added reversed-phase chromatographic beads to concentrate peptides from highly diluted solutions. In the procedure developed, peptide-bead suspensions are dried under vacuum to complete dryness; peptides are subsequently eluted in a small volume of matrix-assisted laser desorption/ionization (MALDI)-matrix containing organic/aqueous solvent and transferred to a MALDI-target for mass analysis. We show that by using this bead-peptide concentration procedure, low femtomole amounts of peptides are efficiently concentrated, up to 1000 times, to volumes smaller than 0.7 microL. We have used this concentration procedure in combination with MALDI-post-source decay analysis to identify subpicomole amounts of proteins present in polyacrylamide gels. Furthermore, we show that the bead-peptide concentration method can be elegantly used to clean up samples contaminated with high concentrations of substances normally deleterious to MALDI-mass spectrometry (MS) experiments. We have found additionally that the bead-peptide concentration procedure can be successfully used to store low femtomole amounts of peptide for prolonged periods of time without severe losses of peptide material. This bead-peptide concentration procedure therefore seems to be a simple and convenient step in the MALDI-MS sample preparation process.

Interactions of rab5 with cytosolic proteins.

Rab proteins, one of the subfamilies of ras-like small GTP-binding proteins, are attached to cellular compartments or transport vesicles and may determine the specificity of fusion between these compartments and vesicles. It has been proposed that they alternate between a membrane-bound and a cytosolic state during their functional cycle. We have used a photo-crosslinking approach to identify their cytosolic interaction partners. In vitro synthesized rab5 was cross-linked in the presence of ATP mainly to three cytosolic proteins of 52, 65, and 85 kDa. Sucrose density gradient centrifugation of the cross-linked products suggested that they were part of a 10-14 S complex. Furthermore, rab5 was cross-linked to these and additional cytosolic proteins of 42, 48, and 160 kDa in the absence of ATP. Unexpectedly, upon ATP depletion of the cytosol cross-linked and noncross-linked rab5 was found in a sedimentable high molecular weight structure. Other members of the rab subfamily, but not N-ras, also sedimented under these conditions. Electrophoretic and electron microscopic analysis of the pelleted material revealed that it contained actin filament bundles and intermediate filaments. Our data suggest that cytosolic rab proteins interact with several proteins in a 10-14 S complex, and that the rab proteins may interact directly or indirectly via this complex with the cytoskeleton.

Identification of the major membrane and core proteins of vaccinia virus by two-dimensional electrophoresis.

Vaccinia virus assembly has been well studied at the ultrastructural level, but little is known about the molecular events that occur during that process. Towards this goal, we have identified the major membrane and core proteins of the intracellular mature virus (IMV). Pure IMV preparations were subjected to Nonidet P-40 (NP-40) and dithiothreitol (DTT) treatment to separate the core proteins from the membrane proteins. These proteins were subsequently separated by two-dimensional (2D) gel electrophoresis, and the major polypeptide spots, as detected by silver staining and 35S labeling, were identified by either matrix-assisted laser desorption/ionization mass spectrometry, N-terminal amino acid sequencing, or immunoprecipitation with defined antibodies. Sixteen major spots that partitioned into the NP-40-DTT-soluble fraction were identified; 11 of these were previously described virally encoded proteins and 5 were cellular proteins, mostly of mitochondrial origin. The core fraction revealed four major spots of previously described core proteins, two of which were also detected in the membrane fraction. Subsequently, the NP-40-DTT-soluble and -insoluble fractions from purified virus preparations, separated by 2D gels, were compared with postnuclear supernatants of infected cells that had been metabolically labeled at late times (6 to 8 h) postinfection. This relatively short labeling period as well as the apparent shutoff of host protein synthesis allowed the selective detection in such postnuclear supernatants of virus-encoded proteins. These postnuclear supernatants were subsequently treated with Triton X-114 or with sodium carbonate to distinguish the membrane proteins from the soluble proteins. We have identified the major late membrane and nonmembrane proteins of the IMV as they occur in the virus as well as in infected cells. This 2D gel map should provide an important reference for future molecular studies of vaccinia virus morphogenesis.

Rabbit skeletal muscle alpha alpha-tropomyosin expressed in baculovirus-infected insect cells possesses the authentic N-terminus structure and functions.

When expressed in E. coli, skeletal muscle alpha-tropomyosin has an unacetylated N-terminus. Unacetylated alpha-tropomyosin lacks important functions; this is non-polymerizable and has a low affinity to actin. In the present work, in order to obtain fully functional recombinant alpha-tropomyosin, rabbit skeletal muscle alpha-tropomyosin (alpha-tropomyosin BV) has been expressed in baculovirus-infected insect cells. alpha-TropomyosinBV was not distinguishable from the authentic tropomyosin, not only in functional properties but also in blocked N-terminus. To know the N-terminus structure of alpha-tropomyosinBV, the N-terminal segment six amino acids long, MDAIKK, has been specifically and efficiently removed from alpha-tropomyosinBV by use of an immobilized proteolytic enzyme system based on E. coli cell bodies which carry the ompT gene product, a proteolytic enzyme localized on the outer cell wall of E. coli. The structure of recombinant alpha-tropomyosinBV was shown to be identical to the authentic protein by electrospray mass spectrometry and protein sequencing analysis. Additionally, electrospray mass spectrometry indicated a single phosphorylation not only in alpha- but also beta-tropomyosin chains in the rabbit skeletal muscle. The differentiated susceptibilities of potential ompT cleavage sites are indicative of a non-coiled-coil conformation of the N-terminus of alpha-tropomyosin.

Peptides adsorbed on reverse-phase chromatographic beads as targets for femtomole sequencing by post-source decay matrix assisted laser desorption ionization-reflectron time of flight mass spectrometry (MALDI-RETOF-MS).

We here describe a procedure for concentrating peptides from solutions by adsorbing them onto reverse-phase beads that were added to these solutions. The beads are then transferred to the target disc of the matrix assisted laser desorption ionization-reflectron time of flight (MALDI-RETOF) mass spectrometer. Because of their hydrophobic nature, these beads cluster in a very small area on the target disc assuring an important concentration step. After drying, peptides are desorbed from the beads by adding a small volume of 50% acetonitrile in 0.1% trifluroacetic acid in water containing the matrix components. Hereby we focus the original amount of peptide material on the target disc on a very small surface, producing highly concentrated peptide-matrix mixtures. This permits high yield identification and sequence tagging by post-source-decay analysis on peptides derived from proteins only available in the femtomole range from one-dimensional (1-D) or two-dimensional (2-D) gels. The procedure is illustrated by the identification of 38 proteins from human thrombocyte membrane skeletons.