Peptide capillary zone electrophoresis mass spectrometry of recombinant human erythropoietin: an evaluation of the analytical method.

An evaluation of capillary zone electrophoresis-mass spectrometry (CZE-MS) as an analytical methodology for the separation and characterization of complex glycopeptides and nonglycopeptide structures has been performed. The evaluation employed endoproteinase V8 digested recombinant human erythropoietin (rHuEPO) that was further fractionated by reverse phase chromatography. The peptides were subjected to sequence analysis and evaluated by capillary electrophoresis, with or without mass detection, for peptide purity. The peptide mass determined from the sequence was then compared to the mass obtained from CZE-MS. Glycosylation sites and carbohydrate branch patterns were easily determined, site specific microheterogeneity (either O-acetylation of N-acetylneuraminic acids or lactosamine extensions of the carbohydrate chain length) was assessed directly, glycosylation site occupancy was evaluated qualitatively, and nonglycopeptides were resolved and analyzed on-line with ease. Incomplete peptide digestion products were detected and identified by CZE-MS. Protein sequence coverage by CZE-MS was 98.2 percent complete from a single map. Off-line evaluation of peptide purity by CZE greatly aided the interpretation of multiple sequence analysis and, in validating that, the CZE-MS was detecting all peptides present. All off-line CZE and on-line CZE-MS experiments employed a capillary that was dynamically coated with Polybrene in the presence of polyethylene glycol; separations were conducted in 0.67 M formic acid.

Sample matrix effects on glycopeptide stability by high performance capillary electrophoresis.

High performance capillary electrophoresis (CE) of glycoprotein digests frequently reveals extensive microheterogeneity associated with specific protein glycosylation sites. The choice of the sample matrix can influence the electrophoretic migration time, peak shape and resolution, as well as the physical stability of the product glycopeptides. Acetic acid is a frequently employed sample matrix for both capillary electrophoresis and electrospray ionization-mass spectrometry (ESI-MS). Acetic acid appears to enhance the spontaneous hydrolysis of sialic acids from the nonreducing termini of glycopeptides in a time- and concentration-dependent manner, even at 5 degrees C, as evidenced by changes in the electrophoretic mobility and ESI-MS spectra of the resulting glycopeptides. The observed parallel electrophoretic mobility changes for specific glycoforms are consistent with the induction of peptide structure with time. Asialoglycopeptide mobilities were stable in acetic acid. Electrophoretic mobilities can be stabilized with propionic acid sample matrix with no apparent structural changes observed by ESI-MS within 31 h. Migration time reproducibility was in the range of 0.1% relative standard deviation (N = 7) with excellent peak shapes and enhanced glycopeptide resolution.

Multiple peptide fraction collection by capillary electrophoresis with reinjection analysis.

This papers addresses many of the optimization parameters necessary to convert from high resolution capillary electrophoresis (CE) analytical separation parameters to automated, micropreparative multiple fraction collection using software-controlled, interrupted applied voltage. Optimization of two parameters are crucial: 1) preparative sample loading and 2) the determination of peak collection windows. Factors affecting sample loading volume are discussed, such as capillary inner diameters, sample temperatures and sample injection times. Peak collection windows have been determined experimentally and offer an advantage to windows calculated using a linear mobility relationship, especially for long run times, high current levels, and multiple voltage ramping required for multiple fraction collection. Reinjection analysis of both non-glycopeptides and glycopeptides are examined, and clearly indicate peak mobility can be employed for identifying the collected peptides. Difficulties associated with quantitation of the collected peaks by CE are described and appear to be predominantly associated with sample matrix effects.

Identification of the residues involved in homodimer formation of recombinant human erythropoietin.

Recombinant human erythropoietin (rHuEPO) is biologically functional when in a monomeric state; upon extensive heating, rHuEPO forms a dimer. The nature of this dimeric linkage was investigated after isolation of the dimer by gel filtration. The dimer fraction was subjected to tryptic digestion, and the peptides were separated by reversed-phase HPLC. SDS-PAGE, N-terminal sequencing, capillary electrophoresis and mass spectrometry (both liquid-chromatographic electrospray and matrix-assisted laser desorption ionization) were employed to compare the tryptic peptides from heat-treated rHuEPO and untreated rHuEPO. Results demonstrated that elevated heat broke the intramolecular disulfide bond between Cys-7 and Cys-161 and an intermolecular disulfide bond then formed from these residues, producing a covalently linked rHuEPO homodimer. Dimer formation was also mathematically modeled and shown to fit a simple equilibrium.

Multiple sequential fraction collection of peptides and glycopeptides by high-performance capillary electrophoresis.

Multiple sequential fraction collection of peptides and glycopeptides by high-performance capillary electrophoresis (HPCE) under applied voltage has been demonstrated from complex tryptic peptide maps. The collection methodology was adapted from a high-resolution glycopeptide mapping procedure and, as such, requires active temperature control of the sample, electrophoresis vials, and collections vials because the electrophoresis buffer system is higher conductive. Resolution was compromised in the preparative HPCE separation due to heavy sample loading and to reduced voltage. The latter was a requirement for this buffer system in order to control Joule heating at the current levels employed; collections were routinely performed at approximately 1.5 W/m. The collection buffer was optimized by the addition of 12% methanol (v/v), thereby improving collection yields. Tryptic non-glycopeptides were group collected; secondary analysis of the HPCE collections agreed with analytical separations with respect to the number of peptides contained in a given fraction. Sequentially collected peptide fractions were analyzed by Edman sequencing and MALDI mass spectrometry to verify peptide identity and sequence. Consistent peptide sequence or mass measurements were obtained for repeat collections. The isolation of the single pure glycopeptide indicates that unique glycopeptide structures can be collected by HPCE and then analyzed by other methods.

Microheterogeneity of erythropoietin carbohydrate structure.

The microheterogeneity of the carbohydrate structures on recombinant human erythropoietin (rHuEPO) expressed in Chinese hamster ovary cells has been evaluated by electrospray ionization (ESI) mass spectrometry (MS) of glycopeptide fragments. The microheterogeneity is largely associated with the presence or absence of terminal N-acetylneuraminic acid (Neu5Ac) residues, varying amounts of O-acetylation of the Neu5Ac residues, and the presence or absence of N-acetyllactosamine extensions. The N-linked carbohydrate structures were structurally diverse; 52 different N-linked oligosaccharide structures were identified. Consistent structural assignments could be made from data obtained using different proteolytic digests, ESI solvent systems (aqueous/methanol systems with acetic or formic acid), and on-line or off-line LC/MS analysis. All glycosylation sites exhibited some level of O-acetylation of Neu5Ac residues. Interestingly, glycosylation site asparagine-83 exhibits mono-O-acetyl and di-O-acetyl Neu5Ac residues, while the other sites, asparagine-24, asparagine-38, and serine-126, exhibit mainly mono-O-acetyl Neu5Ac derivatization. This difference in O-acetylation may be site specific or due to sample handling of labile structures. However, mild base treatment of rHuEPO with NaOH on ice removed the O-acetyl groups associated with a given carbohydrate structure, without adversely affecting the underlying oligosaccharide structure, resulting in a simplified mass spectra. Nuclear magnetic resonance spectroscopy of Neu5Ac residues released by neuraminidase treatment of total rHuEPO indicated that Neu5,9Ac2 residues were present. Additional resonances were also observed that were consistent with other Neu5Ac O-acetyl linkages; these O-acetyl resonances could be removed by mild base hydrolysis of rHuEPO.

Peptide mapping and evaluation of glycopeptide microheterogeneity derived from endoproteinase digestion of erythropoietin by affinity high-performance capillary electrophoresis.

High-performance capillary electrophoresis (HPCE) has been employed to characterize the peptide map of recombinant human erythropoietin (rHuEPO) expressed from Chinese hamster ovary (CHO) cells. The methodology employs an ion pairing agent, 100 mM heptanesulfonic acid in 40 mM sodium phosphate buffer, pH 2.5, to increase peptide resolution, to decrease analyte wall interactions, and to evaluate glycopeptide microheterogeneity. The total tryptic map is segregated into two regions, nonglycosylated and glycosylated peptides. Reproducibility of the peptide map is excellent; the map results in baseline separation of 16 tryptic peptides and one doublet peak composed of two peptides (resolution 0.22). The map furthermore allows for the evaluation of the microheterogeneity associated with the three rHuEPO glycopeptides. At least 12 glycopeptide forms were separated in the initial peptide map. Peptides were identified by Edman sequencing, and the glycopeptides were further subjected to Dionex anion-exchange chromatography. To simplify the level of complexity associated with the glycopeptides, much of the characterization employed asialoglycopeptides and employed several endoproteolytic diagnosis. The relative percent distribution for each purified asialoglycopeptide was calculated to define the level of complexity and to tentatively assign a known structure to the HPCE peak. The level of structural complexity of the asialoglycopeptides appears to increase from the simplest O-linked form to the more complex N83, N38, and N24 glycosylation positions, respectively. HPCE evaluation of glycopeptide microheterogeneity appears to be simpler, faster, and just as sensitive as other more frequently employed methods for glycopeptide characterizations.

Influence of column temperature on the electrophoretic behavior of myoglobin and alpha-lactalbumin in high-performance capillary electrophoresis.

The influence of column temperature on the electrophoretic behavior of myoglobin and alpha-lactalbumin in high-performance capillary electrophoresis (HPCE) is presented. The major effect of temperature is to shorten the analysis time by decreasing the viscosity, but specific temperature effects on the protein migration behavior were also observed. Myoglobin, under high field (350 V/cm), was essentially temperature stable from 20 to 45 degrees C, but at constant current, a second form of myoglobin could be detected at both 214 and 410 nm. The initial form appeared to correspond to the Fe3+ and the second to the Fe2+ oxidation state of the heme iron. The rate of conversion from Fe3+ to the reduced Fe2+ in myoglobin, under given electrophoretic conditions, followed first-order kinetics with a rate constant at 30 degrees C of 304 s-1. A second protein, alpha-lactalbumin type III, demonstrated a conformational transition that resulted in asymmetric peaks and sigmodial mobility plots versus temperature in the transition region.

Complete amino acid sequence of fetal bovine serum acetylcholinesterase and its comparison in various regions with other cholinesterases.

The complete amino acid sequence of a mammalian acetylcholinesterase from fetal bovine serum (FBS AChE) is presented. This enzyme has a high degree of sequence identity with other cholinesterases, liver carboxyesterases, esterase-6, lysophospholipase, and thyroglobulin. The locations of 191 amino acids in 10 regions of the FBS enzyme were compared with corresponding sequences of Torpedo, human, and Drosophila AChEs and human serum butyrylcholinesterase (BChE). In one region there is a marked difference in both the number of amino acids and their sequence between mammalian AChE and other AChEs and the human serum BChE. The amino acid sequence of FBS AChE showed overall homologies of 90% with human AChE, 60% with T. california AChE, 50% with human serum BChE, and 39% with Drosophila AChE in these regions.

Physical and chemical characterization of a horse serum carboxylesterase.

The serine carboxylesterase from horse serum was characterized by amino acid composition, peptide mapping, molecular and subunit weights, and sequencing of the amino acids around the essential serine residue at the active site. A protocol was developed for using reversed-phase high-performance liquid chromatography as the final step to obtain homogeneous preparations of horse serum carboxylesterase. Amounts sufficient for determining the amino acid composition and for peptide maps were obtained from a partially purified starting material which contained approximately 55% carboxylesterase. The amino acid composition, like the subunit weight (70,800 +/- 1400), was similar to the corresponding values reported for other serine carboxylesterases. However, the amino acid sequence of the tryptic digest fragment containing the essential nucleophilic seryl residue differed significantly from the corresponding sequences of other mammalian serine carboxylesterases.

Acetylcholinesterase prophylaxis against organophosphate toxicity.

Fetal bovine serum acetylcholinesterase (FBS-AChE) protected mice from multiple LD50 doses of organophosphorus (OP) nerve agents. Mice were injected intraperitoneally (ip) with up to 3.3 mg (11,000 U) of FBS-AChE which exhibited a relatively long serum half-life and appeared well tolerated. The enzyme protected mice from the OP ethyl-S-2-diisopropylamino-ethylmethylphosphonothiolate (VX) with a stoichiometry equal to approximately 2 moles of enzyme active site per mole of VX. FBS-AChE, at a lower enzyme OP ratio, protected mice from 2 LD50s of the nerve agent methylphosphonofluoridic acid 1,2,2,-trimethylpropyl ester (soman) when used in conjunction with atropine and 2[(hydroxyimino)methyl]-1-methylpyridinium chloride. It is concluded that sequestration of highly toxic OPs by administration of AChE occurs in mice and suggests a new approach to treatment of OP intoxication.

Kinetic investigations into the interactions of aprophen with cholinesterases and a carboxylesterase.

Acetylcholinesterases, butyrylcholinesterases, and carboxylesterases appear to form kinetically a homologous enzyme series with respect to many substrates and inhibitors. The present paper evaluates the interaction of aprophen with acetylcholinesterases, butyrylcholinesterases, and carboxylesterases with respect to protecting the enzyme from organophosphate and carbamate inhibition, accelerating pralidoxime iodide (2-PAM) regeneration of the diisopropylphospho-enzyme, and comparing the inhibition and regeneration kinetics of a soluble mammalian acetylcholinesterase with that of bovine erythrocyte acetylcholinesterase. The irreversible inhibition kinetics of diisopropyl fluorophosphate (DFP) and eserine inhibition of fetal bovine serum acetylcholinesterase were typical of other acetylcholinesterases as indicated by the bimolecular inhibition rate constants, ki, of 7.7 +/- 1.3 X 10(4) M-1 min-1 and 2.9 +/- 1.7 X 10(6) M-1 min-1, respectively. Similarly, the bimolecular regeneration rate constant, kr, for 2-PAM regeneration of the diisopropylphospho-acetylcholinesterase was 14.7 M-1 min-1. The bimolecular rate constants, ki and kr, were not statistically perturbed when the reaction was monitored in the presence of aprophen with the fetal bovine serum acetylcholinesterase. Human serum butyrylcholinesterase was partially protected from DFP inhibition by aprophen with no detectable change in the bimolecular inhibition rate constant, ki. The regeneration of the diisopropylphospho-butyrylcholinesterase by 2-PAM was accelerated in the presence of aprophen by a factor of 2.7 over that of 2-PAM alone (8.4 +/- 2.2 M-1 min-1 to 23.1 +/- 2.6 M-1 min-1 respectively). Neither the inhibition (DFP) nor the regeneration (2-PAM) kinetics observed for the carboxylesterase was perturbed by the presence of aprophen.

A simplified procedure for the purification of large quantities of fetal bovine serum acetylcholinesterase.

A simple procedure has been developed for the large scale purification of fetal bovine serum acetylcholinesterase (AChE) (EC 3.1.1.7). The procedure involves two steps: batch adsorption of the AChE from 250 L of serum onto a procainamide affinity Sepharose 4B gel; and analytical procainamide affinity chromatography of the step-1 product. Over 100 mg of AChE was purified in 10 days to apparent homogeneity with this procedure.

Aprophen: a substrate and inhibitor of butyrylcholinesterase and carboxylesterases.

Aprophen, alpha-methyl-alpha-phenylbenzeneacetic acid-2-(diethylamino) ethyl ester, is a potent reversible inhibitor and a poor substrate of human serum butyrylcholinesterase (BuChE). Complex mixed competitive noncompetitive inhibition kinetics were observed; an apparent competitive inhibition constant was estimated to be 3.7 X 10(-7) M. BuChE hydrolysis of aprophen to diphenylpropionic acid and diethylaminoethanol did not appear to follow Michaelis-Menten kinetics. The BuChE turnover number for aprophen was 2.0 X 10(-3) sec-1. Rabbit liver oligomeric and monomeric carboxylesterases (CE) also hydrolyzed aprophen with a similar turnover number that varied from 1.4 X 10(-3) sec-1 to 4.3 X 10(-4) sec-1 respectively. Comparison of the catalytic rate of aprophen hydrolysis with butyrylthiocholine (BTC) and the neutral aromatic substrate, phenylthiobutyrate (phi TB), indicated that BuChE hydrolyzed BTC and phi TB 3.2 X 10(5) and 3.1 X 10(5) times more rapidly than aprophen respectively. Similarly, the CEs also hydrolyzed BTC and phi TB 17.6 and 1.9 X 10(5) times rapidly than aprophen. Acetylcholinesterases from bovine erythrocyte and electric eel were not inhibited by aprophen nor was aprophen hydrolyzed by these enzymes. The hydrolysis and inhibition reactions may best be described by a complex reaction scheme involving multiple binding sites for both the substrate and the inhibitor as well as positive cooperative ligand binding.

Acetylcholinesterase from fetal bovine serum. Purification and characterization of soluble G4 enzyme.

Acetylcholinesterase (EC 3.1.1.7) from fetal bovine serum (FBS) was purified to electrophoretic homogeneity. The procedure involved procainamide affinity chromatography with native FBS, followed by chromatography on Sepharose 6B and DEAE-Sephadex. The acetylcholinesterase was purified approximately 44,000-fold, and 13 mg was obtained corresponding to an overall yield of about 45%. The purified acetylcholinesterase was stable at 4 degrees C for at least 8 weeks but was labile to freezing; however, in 50% glycerol the enzyme was stable at -20 degrees C for at least 12 weeks. FBS acetylcholinesterase exhibited typical substrate inhibition, had a Km of 120 microM, and a turnover number of 5300 s-1 with the substrate acetylthiocholine. The enzyme was highly sensitive to the specific acetylcholinesterase inhibitor 1,5-bis(4-allyldimethylammoniumphenyl)pentan-3-one. FBS acetylcholinesterase was characterized as a G4 form of acetylcholinesterase and was distinguished from bovine erythrocyte acetylcholinesterase on the basis of lectin gel binding, [3H] Triton X-100 binding, amino acid composition, number of catalytic subunits/molecule, and hydrodynamic properties. FBS acetylcholinesterase had a Stokes radius of 76 A as judged by gel filtration, and from this a molecular weight of 340,000 daltons was calculated. The enzyme had a subunit weight of approximately 83,000 daltons by sodium dodecyl sulfate-polyacrylamide gel electrophoresis; paraoxon titration indicated a relative active site mass of 75,000 daltons. The amino acid composition of FBS acetylcholinesterase was similar to the human erythrocyte acetylcholinesterase (Rosenberry, T. L., and Scoggin, D. M. (1984) J. Biol. Chem. 259, 5643-5652). A monoclonal antibody directed against human erythrocyte acetylcholinesterase, AE-2, (Fambrough, D. M., Engel, A. G., and Rosenberry, T. L. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 1078-1082) cross-reacted with FBS acetylcholinesterase.

Fluorogenic substrates for the enkephalin-degrading neutral endopeptidase (Enkephalinase).

Rat brain neutral endopeptidase ("Enkephalinase") was shown to hydrolyze a series of fluorogenic substrates of the general structure 2-aminobenzoyl-(amino acid)n- leucylalanylglycine -4- nitrobenzylamide . The hydrolysis of these substrates was competitively inhibited by Leu5-enkephalin, demonstrating that these are indeed substrates for the rat brain neutral endopeptidase. Cleavage of the fluorogenic substrates yielded leucylalanylglycine -4- nitrobenzylamide as a common product. In addition, a series of inhibitors previously shown to inhibit thermolysin-like enzymes inhibited the hydrolysis of both Leu5-enkephalin and the synthetic substrates. The results of this study (a) demonstrate that the enkephalin-degrading endopeptidase is similar in specificity to thermolysin, (b) provide a continuous sensitive assay system for the enzyme, and (c) point out the potential use of this substrate class for probing the specificity of the enzyme.

Multiple molecular forms of rat brain enkephalinase.

Rat brain enkephalinase has been partially purified by ion exchange chromatography, chromatofocusing, and affinity chromatography on immobilized lectins. Ion exchange chromatography resolved two principle forms of enkephalinase designated A1 and A2. Both enkephalinase A1 and A2 are bound to immobilized lentil lectin while chromatography on immobilized wheat germ lectin resolved each of the principle forms into two subforms, A1, 1, A1, 2, A2, 1, and A2, 2. All four enkephalinase forms have similar, if not identical kinetic properties. The possible implications of multiple molecular forms of enkephalinases are discussed.

Inhibition of two monomeric butyrylcholinesterases from rabbit liver by chlorpromazine and other drugs.

A new form of cholinesterases has been discovered in rabbit liver; the new enzymes are monomeric butyrylcholinesterases (EC 3.1.1.8), mBuChE I and mBuChE II. These enzymes are inhibited reversibly by chlorpromazine in the pharmacologically active concentration range and they exhibit mixed competitive-noncompetitive inhibition patterns. The apparent competitive inhibition constants, Ki with chlorpromazine, are 1.8 x 10(-6) M for mBuChE I and 7.6 x 10(-6) M for mBuChE II, whereas the noncompetitive inhibition constant is 1.1 x 10(-5) M for mBuChE II as determined by spectrophotometric assay with n-butyrylthiocholine iodide substrate. Although inhibition of mBuChE I also exhibited noncompetitive behavior, a binding constant could not be determined. Human serum oligometric butyrylcholinesterase (oBuChE) was employed as a control cholinesterase and also demonstrated mixed inhibition kinetics. The competitive inhibition constant for the oBuChE was 5.5 x 10(-7) M in the low substrate region, whereas the apparent noncompetitive binding constant was 1.6 x 10(-5) M in the activated higher substrate region with n-butyrylthiocholine iodide as the substrate and chlorpromazine as the reversible inhibitor. The presence of a noncompetitive binding component indicates the presence of an operative modifier or allosteric site binding the inhibitor on both mBuChEs and the oBuChE. The inhibition constants were calculated assuming that the enzymes followed simple Michaelian kinetics.