Six distinct gamma- and omega-type secalins, together with two new low molecular mass glycoproteins, have been identified as the major coeliac immunoreactive proteins from a chloroform/methanol soluble extract from rye endosperm. These components were characterized by a combination of reverse-phase high-performance liquid chromatography, immunoblotting using a coeliac serum and microsequencing analysis. This allowed the identification of a group of secalins with different molecular masses according to their N-terminal amino-acid sequence: one omega-type secalin of 40 kDa (omega 1-40); three gamma-type secalins, one of 70 kDa (gamma-70) and two of 35 kDa (gamma-35); as well as two low molecular mass glycoproteins of 15 and 18 kDa, all exhibiting coeliac serum antigenicity. Moreover, four additional rye components, including two low molecular mass proteins, which did not react with coeliac sera, have also been identified. Analysis by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) of the three main purified coeliac immunogenic secalins, gamma-70, gamma-35 and omega 1-40, indicated molecular masses of 71457, 32240 and 39117 Da, respectively. The omega 1-40 secalin displays a significant absorption in the visible region which could be related to its peculiar low capacity to bind both coeliac sera antibodies and Coomassie brilliant blue dye.

A procedure for spectroscopic characterization and partial fractionation of human protein HC populations by high-performance liquid chromatography-photodiode array ultraviolet-visible detection is reported. Human protein HC from urine or amniotic fluid fractionated by anion-exchange HPLC in a protein Pak DEAE 5PW appeared to be heterogeneous as judged by the asymmetric elution pattern, consisting of a continuous irregular broad peak with several shoulders distributed along the whole chromatogram. Selected fractions containing shoulders were rechromatographed and finally six symmetrical homogeneous peaks with different retention times were obtained from each protein HC preparation. The direct automatic absorption spectra analyses at each peak maximum, indicated that all of the homogeneous peaks seemed to be protein HC, all of them associated to the same chromophore although with different stoichiometry ratios. Isoelectric focusing showed that each peak was composed of a limited number of subpopulations of protein HC with different isoelectric points. Size microheterogeneity has been also demonstrated in both urinary and amniotic protein HC preparations by a combination of size-exclusion HPLC on a TSK 3000 SW6 column and photodiode array detection. Partial fractionation of human albumin on an analytical anion-exchange Mono-Q PC 1.6/5 column, has allowed the identification of heterogeneous chromophore-containing populations displaying significant absorption in the visible region in resemblance to that of protein HC.

A new sulfur-rich basic polypeptide, so called omega-hordothionin, has been isolated from barley endosperm by extractions with NaCl and ammonium bicarbonate followed by reverse-phase high performance liquid chromatography. Purified omega-hordothionin was found to be homogeneous by SDS/polyacrylamide gel electrophoresis, N-terminal amino-acid sequencing and electrospray-ionization mass spectrometric analysis. The complete primary structure of omega-hordothionin was determined by automatic degradation of the intact molecule and peptides obtained by proteolytic cleavage. Omega-hordothionin consists of a single polypeptide chain of 48 amino acids with a molecular mass of 5508 Da deduced from its amino acid sequence, which fully coincides with the 5508.2 Da determined by electrospray-ionization mass spectrometry. The isolated polypeptide showed a characteristic composition with a high content of basic amino acids (five arginine residues, two lysine residues and six histidine residues) and eight cysteine residues, and has strong sequence identity (66%) with the sorghum SI alpha 1 alpha-amylase inhibitor. Omega-hordothionin, like gamma-hordothionin, exhibited translation inhibitory activity on both eukaryotic cell-free systems from mammalian (rat liver and rabbit reticulocyte lysates) and prokaryotic cell-free systems (Escherichia coli). However, in contrast to gamma-hordothionin, omega-hordothionin did not inhibit plant systems such as Triticum aestivum, Cucumis sativus, Vicia sativa and Hordeum vulgare. Gamma-hordothionin also inhibited the alpha-amylase activity from human saliva, while omega-hordothionin and the other different genetic variants of thionins, alpha-hordothionin and beta-hordothionin, failed to show any inhibitory effect.

A methodology is presented for efficiently gaining structural information from electrophoresed proteins after on-gel detection by imidazole-sodium dodecyl sulfate-zinc reverse staining. As a consequence of reverse staining, (a) protein bands arise transparent against a deep white-stained background, limits of detection being in the femtomole range; (b) there is no loss of image when the gel is kept in distilled water (even during years); and (c) protein bands result immobilized, i.e., they do not diffuse upon gel storage. To recover reverse-stained proteins or fragments thereof from the gel, the immobilization of bands must first be abrogated by chelating the zinc ions from stain (protein mobilization). We had originally described mobilization at low pH by using citric acid. Here, we improve this procedure regarding the protein electrotransfer. We demonstrate that mobilization is efficiently done at neutral to alkaline pH by short-term (5 to 10 min) incubation of the gel in a buffer containing glycine or dithiothreitol prior to transfer. Moreover, mobilization was most simply performed by just adding the zinc chelating agent to the transfer buffer. Reverse staining and the new mobilization procedure made electrotransferring single protein bands from gel onto small-sized (13 x 5 mm2) PVDF membrane pieces in mini sandwich-like assemblies practical. Equipment is described for the protein electroblotting in such minisandwiches. Microsequence analysis of the electroblotted proteins showed initial yields in the range of those achieved when the transfer was done from unstained control gels. Protein bands kept in the reverse-stained gel for prolonged time periods (even for as long as 2 years) could be similarly analyzed.(ABSTRACT TRUNCATED AT 250 WORDS)

Distinct α- and γ-type gliadins, as well as a few low molecular weight components have been identified as coeliac immunoreactive proteins from a chloroform/methanol extract from wheat endosperm. Characterization of these components involved the combination of reverse-phase high-performance liquid chromatography, immunoblotting following SDS-PAGE using a coeliac serum and microsequencing analysis. This has allowed the identification of a group of gliadins with different molecular weights, according to their N-terminal amino-acid sequence: five α-type gliadins of 31, 35, 38 and two of 45 kDa, one γ2-type gliadin of 40 kDa, two γ3-type gliadins of 31, and 50 kDa, and two γ-type gliadins with an atypical gliadin N-terminal of 31, and 40 kDa, as well as a few unidentified low molecular weight components and three N-terminal blocked proteins, all exhibiting similar antigenicity.

Matrix-assisted UV laser desorption/ionization time-of-flight mass spectrometry (MALDI/TOF-MS) allows the detection of large proteins from highly complex protein mixtures such as those present in gluten prolamins. These proteins from wheat (gliadins), barley (hordeins), rye (secalins) and probably oats (avenins) are known to cause the mucosal damage in coeliac disease (gluten-sensitive enteropathy), Gluten prolamins are structurally and chemically related, and their resolution into single components is extremely difficult, Preliminary results are presented based on a novel choice of analysis and identification of prolamins from unfractionated protein complex mixtures by MALDI/TOF-MS, The high resolution and sensitivity of this technique have allowed the elucidation of protonated molecular masses of most of the gliadin, hordein, secalin and avenin components displaying a typical characteristic mass pattern: gliadins showed mass signals corresponding to gliadin components ranging from 31 to 55 kDa; barley showed mass signals for 2-3 components with 32-38 kDa; oats showed mass signals for 5-6 components with 30-35 kDa; and rye showed mass signals for two main components of 32 and 39 kDa. Analysis by this technique of gliadin-containing foods allows the immediate identification of the characteristic gliadin mass pattern, consequently permitting easy identification of gliadins in such samples. In summary, MALDI/TOF-MS seems to be a useful alternative technique for the identification of these cereal prolamins with a detection sensitivity of similar to 50-100 ng total protein loaded.

A new family of single chain (type 1) ribosome-inactivating proteins (RIPs), that we have named ebulitins, have been found in mature leaves of Sambucus ebulus L., a caprifoliaceae plant also known to contain a non-toxic two chain (type 2) RIP named ebulin I in its leaves. Ebulitins are basic proteins of M(r) 32,000, 29,000 and 29,000 for ebulitins alpha, beta and gamma, respectively. The simultaneous presence of different basic type 1 and acidic type 2 RIPs in the same plant and in the same tissue is described here for the first time and opens a new door in research into RIPs.

A new thionin from barley, omega-hordothionin, has been shown to exist in aqueous solution as a mixture of two different isoforms in a 3:2 ratio, as revealed by a complete analysis of its two-dimensional 1H-nmr spectra. The conformational heterogeneity arises from cis-trans isomerism about the Phe 12-Pro 13 peptide bond, where the major form corresponds to the cis conformation. The complete assignment of chemical shifts and nuclear Overhauser effects (NOEs) of the two isoforms allows a detailed comparative analysis of their conformational properties, even though a complete calculation of their solution structures is not possible because of a somewhat limited number of NOE constraints. Structures for the two isomers could be modeled, however, on the basis of the high structural homology between omega-hordothionin and related gamma-thionins, and under the conditions of satisfying all observed experimental data. The two isoforms adopt practically identical global folds and the structural changes imposed by cis-trans isomerization are confined to the region proximal to Pro 13. The cis-trans isomerism occurs in a conserved loop connecting the first beta-strand of the triple-stranded antiparallel beta-sheet and the alpha-helix. A comparative analysis of the sequences of this loop in the different thionins suggests that the cis-trans equilibrium about the X-Pro peptide bond depends on the size of the side chain of X (X = Gly in gamma-thionins and Phe in omega-thionin). The structural homology of this new thionin with gamma-thionins as well as with some scorpion toxins and insect defensins suggests that these proteins may share a common mode of functional activity.

Melonin, a protein from the dry seeds of Cucumis melo L. that inhibits protein synthesis in cell-free systems, is a ribonuclease (RNase). Melonin degrades poly(C) but not poly(U) at 55°C and in the presence of 7 M urea, melonin degrades both poly(C) and poly(U). Reflecting this, melonin degrades bonds containing cytidine and, to a lesser extent, uridine in non-synthetic RNA substrates. The NH2-terminal amino acid of purified melonin was blocked, but the amino acid sequences of some tryptic peptides indicate that melonin has sequence similarity to cusativin, an RNase isolated from Cucumis sativus L., and to the RNases from Petunia inflata and Nicotiana alata that determine gametophytic self-incompatibility.

We have studied the distribution of the protein synthesis inhibitory activity in the tissues of Saponaria officinalis L. (Caryophyllaceae). Seven major saporins, ribosome-inactivating proteins, were purified to apparent homogeneity from leaves, roots and seeds using a new procedure of RIPs isolation including ion-exchange and hydrophobic chromatography. They all catalysed the depurination of rat liver ribosomes, which generate the Endo's diagnostic rRNA fragment upon treatment with acid aniline, thus indicating that A4324 from the 28S rRNA has been released (Endo et al. (1987) J. Biol. Chem. 262, 5908-5912). The molecular mass of saporins by SDS-PAGE ranged between 30.2 and 31.6 kDa and by gel-filtration between 27.5 and 30.1 kDa. Amino acid composition and amino-terminal amino acid sequence indicate that all saporins may be considered isoforms. Only two saporins present in roots were glycosylated (SO-R1 and SO-R3). All saporins are very active on cell-free translation systems derived from rabbit reticulocyte lysates, rat liver, Triticum aestivum L., Cucumis sativus L. and Vicia sativa L. However, they are poor inhibitors of an Escherichia coli translation system. They inhibit protein synthesis in HeLa, BeWo and NB 100 cells, HeLa cells being the most resistant. The enzymatic activity of at least one saporin isoform was dependent on magnesium concentration in the standard rat liver cell-free system.

The molecular events which lead to the proteolytic transformation of porcine procarboxypeptidase B (PCPB) in carboxypeptidase B (CPB) have been determined. Among pancreatic and other tested proteinases, trypsin is the only one capable of generating carboxypeptidase B activity from the zymogen, in vitro. In the first step of this process, trypsin produces cleavage at the boundary between the activation region and the CPB region. Subsequently, a definite sequence of cleavages occurs at the C-terminal end of the released activation segment of 95 residues, giving rise to characteristic intermediates and to a proteolytically resistant activation fragment of 81 residues. In this process, the newly formed CPB participates in the quick-trimming of the released activation peptides. Only a single CPB species is formed in the activation process. This fact and the inability of the released activation peptides to inhibit CPB--and, therefore, their inability to slow down the kinetics of appearance of CPB activity--are two important characteristics differentiating between the activation processes of procarboxypeptidases A and B. The sequence of the 95 residues (MW = 12,835) of the activation region of porcine PCPB has also been deduced, largely from the information obtained by Edman degradation of its fragments and in part by considerations of homology with the rat precursor. The porcine PCPB activation region contains a high percentage of acidic residues, lacks cysteines, methionines, and side-chain posttranslational modifications, and presents a low but significant homology (31%) with the corresponding sequence of porcine procarboxypeptidase A.

Three different procarboxypeptidases A and two different procarboxypeptidases B have been isolated for the first time, in a pure and native state, from human pancreatic extracts. These proteins were purified in one or two quick steps by anion-exchange HPLC. All these forms have been biochemically characterized. Two of the procarboxypeptidases A, the A1 and A2 forms, are obtained in a monomeric state while the other, the A3 form, is obtained as a binary complex of a procarboxypeptidase A with a proproteinase E. This complex is stable in aqueous buffers at various ionic strengths and develops carboxypeptidase A and proteinase E activities in the presence of trypsin. The A1 and A2 forms show clear differences in electrophoretic mobility in SDS/polyacrylamide gels, isoelectric point, proteolytic activation process with trypsin and susceptibility to thermal denaturation. In contrast, these properties are similar in the A1 and A3 (binary complex) forms. On the other hand, with respect to the properties listed above, the B1 and B2 forms differ from each other mainly in isoelectric point. An overall comparison of the above properties reveals the unusual character of the A2 form, midway between the other A and B forms. N-terminal extended sequence analysis carried out on these proenzymes confirm that they constitute different isologous forms.

The complete primary structure of the activation segment of monomeric procarboxypeptidase A from porcine pancreas has been determined by automated and manual Edman-like degradation methods performed on its fragments generated by enzymatic cleavage. The polypeptide consists of 94 residues, with a molecular mass of 10,768, and presents a high proportion of acidic and hydrophobic residues and a proline-rich region in the center of the molecule. Comparison of this sequence with the already reported equivalent sequence deduced from rat procarboxypeptidase A cDNA reveals a very high degree of homology between the two propeptides (up to a 81% of identities), which is even higher in certain large zones of the molecule.

Automated Edman degradation of monomeric procarboxypeptidases A and B from porcine pancreas shows that their N-terminal regions (from residue 1 to 34–37) present a high degree of sequential homology to each other as well as to other related procarboxypeptidases. Conformational predictions based on these sequences confirm their structural homology and indicate the probable existence of two β-turns, one β-chain and a long α-helix in them. On the other hand, tryptic peptide maps on a reverse-phase column indicate great sequential similarities (if not identity) between monomeric procarboxypeptidase A and the procarboxypeptidase A subunit isolated from its binary complex with proproteinase E.

The complete amino acid sequence of the cytotoxin mitogillin has been determined by sequencing the intact chain and peptide fragments produced by cleavage at methionyl, arginyl, lysyl, and tryptophanyl residues and at one aspartic acid-proline bond. The protein consists of 149 amino acid residues with alanine at the NH2 terminus and histidine at the COOH terminus. The calculated Mr of the native mitogillin was 16 867. The native molecule presents two disulfide bridges, one between cysteine residues at positions 5 and 147 and another one between cysteine residues at positions 75 and 131. The amino acid sequence of mitogillin shows 86% homology with another cytotoxic protein called alpha-sarcin.

The complete amino acid sequence of the single polypeptide chain of cytotoxin restrictocin has been determined. Its structure was established by automated Edman degradation of the intact molecule reduced and [14C]carboxymethylated and of fragments obtained by chemical cleavage of the protein with cyanogen bromide and BNPS-skatole and by enzymatic cleavage of the polypeptide chain with trypsin. The molecule consists of 149 amino acid residues with a calculated relative molecular mass of 16836. The protein presents two disulfide bridges, one between cysteine residues at positions 5 and 147 and the other one formed by cysteine residues at positions 75 and 131. The amino acid sequence of restrictocin shows a high degree of homology (86%) with that of the cytotoxin named alpha-sarcin.