Confirmation by mass spectrometry of a trisulfide variant in methionyl human growth hormone biosynthesized in Escherichia coli.

A sulfur-containing compound found in acid hydrolysates of proteins was identified 30 years ago as a trisulfide: bis-(2-amino-2-carboxyethyl) trisulfide (cysteine2S3). At that time, studies concerning the chemistry of sulfur-transferring enzyme systems suggested that cysteine2S3 also existed in biological systems. Two decades later, a cystine trisulfide structure was postulated in the regulator protein molecule for the activation of delta-aminolevulinate synthetase. Recently, a trisulfide bond was reported to occur in the minor loop disulfide at Cys182-Cys189 in human growth hormone. We have detected a trisulfide structure in methionyl human growth hormone in the major loop disulfide Cys53-Cys165. The development of mass spectral analyses of high molecular weight molecules, such as proteins, led to the eventual identification of the modification. A tandem mass spectral analysis on a Sciex electrospray instrument localized an addition of 32 Da to the Cys53-Cys165 fragment. Elemental composition was determined by accurate mass measurement obtained by peak matching to a synthetic peptide and established that an extra sulfur atom was involved.

Preparative isolation of recombinant human insulin-like growth factor 1 by reversed-phase high-performance liquid chromatography.

The isolation of recombinant human insulin-like growth factor 1 (rhIGF-1) is complicated by the presence of several rhIGF-1 variants which co-purify using conventional chromatographic media. These species consist primarily of a methionine-sulfoxide variant of the properly folded molecule and a misfolded form and its respective methionine-sulfoxide variant. An analytical reversed-phase high-performance liquid chromatography procedure using a 5-micron C18 column, an acetonitrile-trifluoroacetic acid (TFA) isocratic elution, and elevated temperature gives baseline resolution of the four species. Using this analytical method as a development tool, a process-scale chromatography step was established. The 5-micron analytical packing material was replaced with a larger-size particle to reduce back-pressure and cost. Since the TFA counter-ion binds tightly to proteins and is difficult to subsequently dissociate, a combination of acetic acid and NaCl was substituted. Isocratic separations are not good process options due to problems with reproducibility and control. A shallow gradient elution using premixed mobile phase buffers at the same linear velocity was found to give an equivalent separation at low load levels and minimized solvent degassing. However, at higher loading there was a loss of resolution. A matrix of various buffers was evaluated for their effects on separation. Elevated pH resulted in a significant shift in both the elution order and relative retention times of the principal rh-IGF-1 variants, resulting in a substantial increase in effective capacity. An increase in the ionic strength further improved resolution. Several different media were evaluated with regard to particle size, shape and pore diameter using the improved mobile phase. The new conditions were scaled up 1305-fold and resulted in superimposable chromatograms, 96% recovery and > 99% purity. Thus, by optimizing the pH, ionic strength and temperature, a high-capacity preparative separation of rhIGF-1 from its related fermentation variants was obtained.

Identification and structural analysis of the tetrasaccharide NeuAc alpha(2-->6)Gal beta(1-->4)GlcNAc beta(1-->3)Fuc alpha 1-->O-linked to serine 61 of human factor IX.

O-Linked fucose has been found attached to Thr/Ser residues within the sequence Cys-X-X-Gly-Gly-Thr/Ser-Cys in the N-terminal EGF domains of several coagulation/fibrinolytic proteins. Carbohydrate composition and mass spectrometric analyses of tryptic and thermolytic peptides containing the corresponding site (Ser-61) in the first EGF domain of human factor IX indicated the presence of a tetrasaccharide containing one residue each of sialic acid, galactose, N-acetylglucosamine, and fucose. The Ser-61 tetrasaccharide was not susceptible to alpha-fucosidase digestion. Fragments generated during mass spectrometric analysis indicated that fucose was the attachment sugar residue. The involvement of fucose in the carbohydrate-peptide linkage was confirmed by two-dimensional 1H NMR spectroscopic analysis of the glycopeptide containing factor IX residues 57-65. The complete structure of the tetrasaccharide was obtained by methylation analysis and two-dimensional 1H TOCSY and ROESY experiments as NeuAc alpha(2-->6)Gal beta(1-->4)GlcNAc beta(1-->3)Fuc alpha 1-->O-Ser61.

Biochemical characterization of recombinant human nerve growth factor.

Recombinant human nerve growth factor (rhNGF) was expressed and secreted by Chinese hamster ovary cells and purified to homogeneity using ion-exchange and reversed-phase (RP) chromatography. The isolated product was shown to be consistent with a 120-amino-acid residue polypeptide chain by amino acid composition, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), RP-HPLC, and mass spectrometry and with an N-terminal sequence consistent with that expected from the cDNA for human nerve growth factor. By size-exclusion chromatography, rhNGF behaves like a noncovalent dimer. Limited enzymatic digests of the 120-residue monomer produced additional species of 118 (trypsin, removal of the C-terminal Arg119-Ala120 sequence) and 117 (trypsin plus carboxypeptidase B, removal of the C-terminal Arg118-Arg119-Ala120 sequence) residues. Each of these species was isolated by high-performance ion-exchange chromatography and characterized by amino acid and N-terminal sequence analyses, SDS-PAGE, RP-HPLC, and mass spectrometry. All three species were present in the digests as both homodimeric and heterodimeric combinations and found to be equipotent in both the chick dorsal root ganglion cell survival and rat pheochromocytoma neurite extension assays.

Chemical heterogeneity as a result of hydroxylamine cleavage of a fusion protein of human insulin-like growth factor I.

Recombinant DNA techniques were used to biosynthesize human insulin-like growth factor I (hIGF-I) as a fusion protein wherein the fusion polypeptide is an IgG-binding moiety derived from staphylococcal protein A. This fusion protein is produced in Escherichia coli and secreted into the fermentation broth. In order to release mature recombinant-derived hIGF-I (rhIGF-I), the fusion protein is treated with hydroxylamine, which cleaves a susceptible Asn-Gly bond that has been engineered into the fusion protein gene. Reversed-phase h.p.l.c. was used to estimate the purity of the rhIGF-I preparations, especially for the quantification of the methionine sulphoxide-containing variant. It was determined that hydroxylamine cleavage of the fusion protein produced, as a side reaction, hydroxamates of the asparagine and glutamine residues in rhIGF-I. Although isoelectric focusing was effective in detecting, and reversed-phase h.p.l.c. for producing enriched fractions of the hydroxamate variants, ion-exchange chromatography was a more definitive procedure, as it allowed quantification and facile removal of these variants. The identity of the variants as hydroxamates was established by Staphylococcus aureus V8 proteinase digestion, followed by m.s., as the modification was transparent to amino acid and N-terminal sequence analyses. The biological activity of rhIGF-I was established by its ability to incorporate [3H]thymidine into the DNA of BALB/c373 cells and by a radioreceptor assay utilizing human placental membranes. Both assays demonstrate that the native, recombinant and methionine sulphoxide and hydroxamate IGF-I variants are essentially equipotent.

O-linked fucose is present in the first epidermal growth factor domain of factor XII but not protein C.

Epidermal growth factor (EGF) domains are found in many proteins, particularly those of the coagulation/fibrinolytic system. We and others have demonstrated that tissue plasminogen activator (t-PA) and prourokinase are modified by the attachment of fucose to equivalent threonine residues within their EGF domains. Factor XII and protein C each contain two EGF domains; in both proteins, the EGF domain nearest the N terminus has a threonine residue in a position homologous to that which is fucosylated in t-PA. In protein C, this site is 3 residues from the position of another post-translational modification, beta-hydroxylation of Asp-71. We isolated peptides containing these sites to determine, primarily by mass spectrometric analysis, the presence of O-linked fucose and/or beta-hydroxyaspartate. We found that factor XII is fully fucosylated at Thr-90. Protein C is unmodified at the equivalent site (Thr-68) and is completely beta-hydroxylated at Asp-71. It has been recently reported that the first EGF domain of human factor VII has O-linked fucose at the equivalent position (Ser-60) (Bjoern, S., Foster, D. C., Thim, L., Wiberg, F. C., Christensen, M., Komiyama, Y., Pedersen, A. H., and Kisiel, W. (1991) J. Biol. Chem. 266, 11051-11057), while it is unmodified at Asp-63 despite having the consensus sequence for beta-hydroxylation at the latter site. These observations raise the possibility that O-linked fucosylation and beta-hydroxylation of EGF domains are mutually exclusive post-translational modifications.

Transpeptidation during the analytical proteolysis of proteins.

Since peptide mapping with proteolytic enzymes such as trypsin and Staphylococcus aureus V8 protease is a powerful tool for the characterization of proteins, investigators should be cognizant of possible artifacts due to the technique itself. This article describes the identification of minor peaks found in the maps of recombinant human relaxin and insulin-like growth factor I as transpeptidation products. Both proteins have some homology to insulin with relaxin being composed of two chains designated A and B, while insulin-like growth factor I is composed of a single polypeptide chain. Digestion of relaxin with trypsin at pH 7.2 yields two peptides, T2,3(A10-18) and T7(B10-13), linked together by a disulfide bond. An unexpected component at a 10% level was identified to be the T2-T7 peptide pair where T3(ArgA18) has formed a peptide bond with the amino-terminal LeuB10 of the T7 peptide. It was also observed that the digestion of insulin-like growth factor I with V8 protease normally yields two peptides V4(13-20) and V9(59-70) linked by a disulfide bridge. A minor peak at a 1 to 2% level was identified to be a single polypeptide resulting from the formation of a peptide bond between the amino-terminal Met59 of V9 and the carboxyl-terminal Asp20 of V4, with the disulfide bond intact. These transpeptidation products were isolated by reversed-phase HPLC and identified using amino-terminal sequence and mass spectrometric analyses.

Structural characterization by mass spectrometry of native and recombinant human relaxin.

Mass spectrometry has played a key role in characterizing the primary structure of native and recombinant relaxin, a peptide hormone that induces ripening of the cervix prior to childbirth. The peptide is composed of two chains, A and B, and is formed from a single-chain prohormone, as is insulin. Aside from conserved cysteines, though, it has little sequence homology with insulin. Due to the small amounts of native peptide initially available (less than 10 pmol), traditional techniques could not provide information on the blocked A-chain sequence, on the carboxyterminal sequences, nor on other possible post-translational modifications. Mass measurements by fast atom bombardment (FAB) were made on reduced human relaxin isolated from corpora lutea. The detection limit by FAB for reduced relaxin was 500 fmol. The B-chain was four amino acids shorter than expected from comparison of the previously known cDNA sequence with homologous rat and porcine sequences. The A-chain, as predicted, was 24 amino acids in length and had a pyroglutamic acid residue on the amino-terminus. The purified samples were homogeneous with no other post-translational modifications. The recombinant relaxin molecule was also extensively characterized by mass spectrometry. In addition to the intact molecule, all tryptic peptides were characterized by FAB. A capillary high-performance liquid chromatography continuous-flow FAB system, developed for high-sensitivity peptide mapping, aided in these analyses. Finally, the three disulfide bonds were shown by tandem mass spectrometry to match those of insulin.

Preparation of biologically active 32P-labeled human relaxin. Displaceable binding to rat uterus, cervix, and brain.

Relaxin is a member of the insulin family of polypeptide hormones and is known to exert its biological effects on various parts of the mammalian reproductive system. Biologically active human relaxin has been chemically synthesized based on the nucleotide sequence obtained from an ovarian cDNA clone. In the present study synthetic human relaxin was radiolabled by phosphorylation with cAMP-dependent protein kinase and [gamma-32P]ATP to a specific activity of 5000 Ci/mmol. The phosphorylated relaxin was purified on cation exchange high performance liquid chromatography and was shown to co-migrate with relaxin on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Mass spectrometry revealed a single phosphorylated site on the B chain of relaxin. The 32P-relaxin was able to bind to a goat anti-relaxin antibody, and this binding could be displaced by unlabeled relaxin in a concentration-dependent manner. A comparison of the concentration responses of cellular cAMP production stimulated by relaxin and phosphorylated relaxin in a primary human uterine cell line showed that phosphorylation did not affect the in vitro biological efficacy of relaxin. This made it suitable for in situ autoradiographic localization of relaxin binding sites in rat uterine, cervical, and brain tissue sections. Displacement of the binding of 100 pM 32P-relaxin by 100, 10, and 3 nM unlabeled relaxin, but not by 100 nM insulin, insulin-like growth factor-I, and an insulin-like growth factor-I analog, demonstrated the high affinity and specificity of such binding. We conclude that 32P-labeled human relaxin is biologically and immunologically active and that this novel probe binds reversibly and with high affinity to classical (e.g. uterus) and unpredicted (e.g. brain) tissues.