The need to identify protein or peptide biomarkers via readily available biological samples like serum, plasma, or cerebrospinal fluid is often hindered by a few particular proteins present at relatively high concentrations. The ability to remove these proteins specifically, reproducibly, and with high selectivity is increasingly important in proteomic studies, and success in this procedure is leading to an ever-increasing list of lower abundant proteins being identified in these biological fluids. The current work addresses some of the potential problems in depleting proteins in typical biomarker studies, including nonspecific binding during depletion procedures and whether low molecular weight (LMW) species bind to the column in a so-called "sponge" effect caused by the ability of albumin or other high-abundant proteins to bind peptides or protein fragments. LC-MS/MS methods were applied to the comparative analysis of an IgG-based immunodepletion method and a Cibacron blue (CB)-dye-based method, for specificity of removing targeted proteins (binding fraction), as well as for assessing efficiency of target removal. This analysis was extended to examine the effects of repeated use of materials (cycles of binding and elution), in order to assess potential for carryover of one sample to the next. Capacity studies and efficiency of protein removal from the serum samples were followed for the IgG-based system using both immunochemical assays (ELISA) as well as LC-MS/MS methods. Additionally, the IgG-based system was further characterized for the removal of LMW polypeptides by nonspecific binding. We conclude that the IgG-based system provided effective removal of targeted proteins, with minimal carryover, high longevity, and minimal nonspecific binding. Significant differences are noted between the depletion techniques employed, and this should be considered based on the expectations set during experimental design.

We have analyzed the proteome of human cerebrospinal fluid with the help of shotgun mass spectrometry. In order to identify low-abundant proteins in these fluids, we have found it necessary to remove the abundant protein components from the mixture. Immunodepletion of the abundant proteins has allowed us to identify more than 100 proteins in cerebrospinal fluids from a patient suffering from normal pressure hydrocephalus. The identified proteins belong to a variety of different classes ranging from serum proteins to intracellular mediators that are involved in signal transduction and transcription. This work establishes a platform for future studies aimed at the comparative proteome analysis of cerebrospinal fluids from different groups of patients suffering from various psychiatric and neurological disorders.

The results of a study to determine the utility of in-source decay fragmentation of matrix-assisted laser-desorbed ions for obtaining useful sequence information on unknown peptides are presented. Six peptides were purified by high-performance liquid chromatography and submitted as single blind unknowns. The in-source decay fragment ion data were collected on a linear time-of-flight mass spectrometer equipped with delayed extraction. These fragment ion data were manually interpreted on the basis of known fragmentation pathways to determine a proposed sequence. The proposed sequences for three of the unknowns were essentially correct, with a few minor errors. A fourth unknown had significant errors associated with its proposed sequence due to misinterpretation of the fragmentation data. Two unknowns were found to have undergone significant sample degradation prior to analysis, which compromised the results for these samples. An example of the use of protein database searching of a partial peptide sequence to aid in a sequence determination is also presented.

Proteins and peptides can be sequenced from the carboxy terminus with isothiocyanate reagents to produce amino acid thiohydantoin derivatives. Previous studies in our laboratory indicated that the use of trimethylsilyl isothiocyanate (TMS-ITC) as a coupling reagent significantly improved the yields and reaction conditions and reduced the number of complicating side products [Hawke et al. (1987) Anal. Biochem. 166, 298]. The present study further explores the conditions for formation of the peptidylthiohydantoins by TMS-ITC and examines the cleavage of these peptidylthiohydantoin derivatives into a shortened peptide and thiohydantoin amino acid derivative. Schizophrenia-related peptide (Thr-Val-Leu) was used as a model peptide and was treated with acetic anhydride and TMS-ITC at 50 degrees C for 30 min, and the peptidylthiohydantoin derivatives were isolated by reverse-phase HPLC and characterized by FAB-MS. The purified derivatives were subjected to a variety of cleavage conditions, and rate constants for hydrolysis were determined. Hydrolysis with acetohydroxamate as reported originally by Stark [(1968) Biochemistry 7, 1796] was found to give excellent cleavage of the terminal thiohydantoin amino acid, but also led to the formation of stable hydroxamate esters of the shortened peptide which are poorly suited for subsequent rounds of degradation. Hydrolysis with 2% aqueous triethylamine under mild conditions (1-5 min at 50 degrees C) was found to be more suitable for carboxy-terminal sequence analysis by the thiocyanate method. The shortened peptide, which could be isolated and subjected to a second round of degradation, and the released thiohydantoin amino acid are formed in good yield (90-100%). Several other small peptides containing 15 different C-terminal amino acid side chains were also investigated in order to examine any interfering reactions that might occur when these side chains are encountered in a stepwise degradation using the thiocyanate chemistry. Quantitative yields of peptidylthiohydantoins were obtained for all the amino acids examined with the following exceptions: low yields were obtained for C-terminal Glu or Thr, and no peptidylthiohydantoins were obtained for C-terminal Pro or Asp. Asparagine was found to form cyclic imides (64%) at the penultimate position (C-2) during hydrolysis of the peptidylthiohydantoins by 2% aqueous triethylamine. Cleavage of C-terminal Asn under these conditions led to the formation of the expected shortened peptide (69%), but also to the formation of a shortened peptide (31%) with a C-terminal amide. Problems with Glu and Thr could be solved by minimizing the reaction time with acetic anhydride.(ABSTRACT TRUNCATED AT 400 WORDS)

A new reactive fluorescent adenine nucleotide analogue has been synthesized and characterized: 2-[(4-bromo-2,3-dioxobutyl)thio]-1,N6-ethenoadenosine 2',5'-bisphosphate (BDB-T epsilon ADP). This compound reacts irreversibly with NADP+-specific isocitrate dehydrogenase. Biphasic kinetics of inactivation are observed that can be described in terms of a fast initial phase of inactivation resulting in partially active enzyme of 8-10% residual activity, followed by a slower phase leading to total inactivation. NADPH protects completely against the fast phase of the reaction, indicating that modification occurs at the coenzyme binding site, whereas isocitrate with MnSO4 protects totally against the slow phase of reaction. The inactivation rate constants for both phases exhibit nonlinear dependence on BDB-T epsilon ADP concentration, consistent with the formation of a reversible complex with the enzyme prior to irreversible modification. Covalent incorporation of BDB-T epsilon ADP is limited and specific; only 0.99 mol of reagent/mol of subunit is incorporated when the enzyme is 98% inactivated in the absence of ligands. A maximum incorporation of 0.5 mol of reagent/mol of subunit is obtained in the presence of isocitrate and MnSO4, while incorporation in the presence of NADPH equals the difference between the incorporation in the absence of ligands and that measured in the presence of isocitrate and MnSO4. It appears that 0.5 mol of reagent/mol of subunit is responsible for the fast phase of inactivation and the remaining incorporation causes the slow phase. Under all conditions used in this study, isocitrate dehydrogenase has been shown to exist as a dimer by analytical ultracentrifugation and by cross-linking with dimethyl suberimidate followed by analysis on polyacrylamide gels in the presence of sodium dodecyl sulfate. It is proposed that, in the fast phase of inactivation, 2-[(4-bromo-2,3-dioxobutyl)thio]-1,N6-ethenoadenosine 2',5'-bisphosphate reacts at the coenzyme binding site of one subunit of dimeric isocitrate dehydrogenase, causing complete inactivation of the modified subunit and substantial inactivation of the other subunit. This new reagent is likely to have general applicability as an affinity label for other NADP+ binding enzymes.

NAD(P)H:quinone acceptor oxidoreductase (EC 1.6.99.2) (DT-diaphorase) is a FAD-containing reductase that catalyzes a unique 2-electron reduction of quinones. It consists of 2 identical subunits. In this study, it was found that the carboxyl-terminal portion of the 2 subunits can be cleaved by various proteases, whereas the amino-terminal portion cannot. It was also found that proteolytic digestion of the enzyme can be blocked by the prosthetic group FAD, substrates NAD(P)H and menadione, and inhibitors dicoumarol and phenindione. Interestingly, chrysin and Cibacron blue, 2 additional inhibitors, cannot protect the enzyme from proteolytic digestion. The results obtained from this study indicate that the subunit of the quinone reductase has a 2-domain structure, i.e., an amino-terminal compact domain and a carboxyl-terminal flexible domain. A structural model of the quinone reductase is generated based on results obtained from amino-terminal and carboxyl-terminal protein sequence analyses and electrospray mass spectral analyses of hydrolytic products of the enzyme generated by trypsin, chymotrypsin, and Staphylococcus aureus protease. Furthermore, based on the data, it is suggested that the binding of substrates involves an interaction between 2 structural domains.

A new reactive adenine nucleotide has been synthesized: 2-[(4-bromo-2,3-dioxobutyl)thio]-adenosine 5'-monophosphate (2-BDB-TAMP). Adenosine 5'-monophosphate 1-oxide was synthesized by reaction of AMP with m-chloroperoxybenzoic acid. Treatment with NaOH followed by reaction with carbon disulfide yielded 2-thioadenosine 5'-monophosphate (TAMP). The final product was generated by reaction of TAMP with 1,4-dibromobutanedione. The structure of 2-BDB-TAMP was determined by UV, 1H NMR, and 13C NMR spectroscopy as well as by bromide and phosphorus analysis. Rabbit muscle pyruvate kinase is inactivated by 2-BDB-TAMP at pH 7.0 and 25 degrees C. The inactivation rate exhibits a nonlinear dependence on the reagent concentration with KI = 0.57 mM. Protection against inactivation is provided by ADP and ATP, in the presence of Mn2+, as well as by phosphoenolpyruvate, in the presence of K+; in addition, partial protection is provided by AMP plus Mn2+. Incubation of pyruvate kinase with 0.075 mM 2-BDB-TAMP for 70 min in the absence of protective ligands leads to incorporation of 1.55 mol of reagent/mol of enzyme subunit when the enzyme is 53% inactive. In the presence of ADP and Mn2+, only 0.96 mol of reagent/mol of subunit is incorporated at 70 min, while the enzyme retains 100% activity. Similar results were obtained in the presence of ATP plus Mn2+. Assuming that the groups modified in the absence of ligands include those modified in the presence of the nucleotides, the 53% inactivation can be attributed to the modification of 0.59 (1.55-0.96) group per enzyme subunit.(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract