High yield expression and purification of human endothelin-1

A DNA construct encoding human big endothelin (Big ET) preceded by the factor Xa protease recognition site (Ile-Glu-Gly-Arg), fused in frame to the maltose binding protein sequence, has been introduced in DH5-alpha cells. The fusion product (MBP-Big ET) was expressed at a concentration close to 100 micrograms/ml of culture broth and constituted approximately 50% of the total protein content. Crude cell extracts containing the fusion product have been directly treated with trypsin under mild denaturing conditions in order to release big endothelin (1-37) from the adduct. Cleavage yield of the MBP-Big ET adduct was close to 70%. Big ET(1-37) was separated from unrelated peptides derived from the tryptic digest of the bacterial extract by affinity chromatography. The affinity column was prepared by immobilizing a protease resistant peptide ligand able to recognize Big ET with sufficient affinity, selectivity, and specificity. From the affinity step (recovery, 90%), recombinant Big ET(1-37) was obtained with a purity close to 80%. The affinity-purified recombinant product was then digested with alpha-chymotrypsin in order to release endothelin (1-21), which was then purified by RP-HPLC. With this two-step purification protocol, 3 micrograms of endothelin was recovered from 1 ml of bacterial broth, with a purity close to 95%.

Identification of an epitope of tumor necrosis factor (TNF)-receptor type 1 (p55) recognized by a TNF-alpha-antagonist monoclonal antibody

The relationships between epitope topography and agonistic/antagonistic effects of anti-TNF receptor type 1 (TNF-R1) antibodies on TNF-alpha cytotoxic activity have been studied. To this purpose various monoclonal antibodies (mAbs) against the soluble form of TNF-R1 (sTNF-R1) have been generated and characterized. Epitope topography studies identified at least four distinct epitopes located outside (4E10) or within (or close to) the TNF-alpha binding site of urinary sTNF-R1 (7H3, 4C1, 9B11). mAbs 7H3 and 4C1 were able to neutralize the inhibition of human TNF-alpha cytotoxicity on L-M cells by sTNF-R1, while 4E10 was unable. Moreover, 7H3 and 4C1 were able to antagonize the TNF-alpha cytotoxicity on human U937 cells, while they were uneffective on mouse L-M cells, suggesting that these antibodies recognize, in a species-specific mode, also the membrane form of the human receptor. No agonistic effects were observed when these antibodies were used in the absence of TNF-alpha. Epitope topography studies carried out using overlapping decapeptides covering most of the sTNF-R1 sequence showed that residues 143-148 of the fourth cysteine-rich domain of the receptor (FFLREN) contain antigenic determinants recognized by the antagonist antibody 7H3. These results suggest that at least part of residues 143-148 of sTNF-R1 are surface exposed on the soluble as well as on the membrane forms of TNF-R1 and are accessible to TNF-alpha antagonists.

PCR amplification of antibody variable regions using primers that anneal to constant regions

Images

Binding of type I IL-1 beta receptor fragment 151-162 to interleukin-1 beta

The relevance of hydropathically complementary sequences in ligand receptor interactions has been evaluated in the interleukin-1 beta/receptor type I case. Computer assisted comparison of the hydropathic profiles of IL-1 beta and its receptor (type I) identified residues 88-99 in IL-1 beta and 151-162 in the receptor as the sequences pair characterized by the highest level of hydropathic complementarity. These fragments, once produced by chemical synthesis and derivatized with biotin, displayed specific recognition properties for each other, as detected by solid phase binding assays. Binding between the two fragments occurred independently from the assay format, was saturable and specifically inhibited by unlabeled peptides. Receptor fragment (151-162) derivatized with biotin recognized also full length recombinant IL-1 beta, and binding was inhibited to 50% in the presence of 3 microM IL-1 beta (88-99) peptide. Interaction specificity was further confirmed by the non competitive effect on the interaction of a sequence scrambled IL-1 beta (88-99) peptide. In a similar way, full length biotinylated IL-1 beta recognized immobilized IL-1 beta receptor fragment (151-162), and this interaction was diminished in the presence of unlabeled receptor fragment or IL-1 beta Results indicate that IL-1 beta receptor fragment (151-162) binds IL-1 beta recognizing the IL-1 beta (88-99) sequence, thus suggesting a possible role of these fragments in the protein/receptor recognition surface.

Kinetic analysis of TNF-alpha oligomer-monomer transition by surface plasmon resonance and immunochemical methods

In this work we have studied the kinetic parameters of oligomeric tumour necrosis factor alpha (TNF-alpha) dissociation using biospecific interaction analysis (BIA), based on surface plasmon resonance (SPR) of TNF-alpha immobilized on a sensor chip, and by an ELISA technique able to detect TNF-alpha oligomers in solution. Validation studies, carried out with sensor chips bearing TNF-alpha oligomers or bovine albumin monomers, verified that: (a) TNF-alpha can be immobilized in the oligomeric form onto sensor chips; (b) the covalent linkage between TNF-alpha and sensor chips is stable under the experimental conditions; (c) TNF-alpha monomers are present on the sensor chips after dissociation; (d) immobilization and dissociation rate constant (kdiss) measurements are reproducible. The kdiss of recombinant TNF-alpha, measured under non denaturing conditions at pH 7.4 by BIA and ELISA were in good agreement, being 0.92 x 10(-5)/s and 1.1 x 10(-5)/s respectively (corresponding to a half life of about 20.9 h and 17.5 h, respectively). The dissociation rate was found to be significantly affected by the presence of detergents and by the pH of the solution, suggesting that TNF-alpha, at low concentrations, exists in solution with different molecular forms depending on the time of storage and buffer composition. Real-time BIA is rapid and does not require particular antibodies or reagents. Thus, the stability of the quaternary structure of natural or recombinant TNF-alpha from human or animal species as well as that of other oligomeric cytokines can probably be studied using this method.

Evidences that syngeneic alpha-type anti-idiotypic antibodies may non-competitively inhibit idiotype/oligomeric antigen interactions by affecting idiotype avidity

The effects of syngeneic anti-Id antibodies on the multivalent interaction between human TNF-alpha, a homotrimeric Ag, and an anti-TNF mAb (mAb(1)78) have been studied. Eight anti-mAb(1)78 Ig secreting hybridoma, able to inhibit TNF binding in a competitive or non-competitive mode, have been generated. Two representative clones (mAb(2)1G3 and mAb(2)9F1) were selected for studying the inhibition mechanism of TNF-mAb(1)78 interaction. Idiotype-paratope topography studies indicated that mAb(2)1G3 (IgG2a) and mAb(2)9F1 (IgG1) bind two sterically distinct idiotopes on mAb(1)78 (IgG1) V regions. In particular, mAb(2)1G3 was found to bind an idiotope located within (or spatially close to) the Ag combining site suggesting that competitive inhibition of TNF binding to mAb(1)78 by mAb(2)1G3 occurs through paratope blockade. On the other hand, mAb(2)9F1 recognizes an idiotope located outside the paratope, being able to bind mAb(1)78 even in the presence of saturating amounts of TNF. mAb(1)78-TNF molar ratio in complexes, at stoichiometric equivalence, was unchanged in the presence of a large excess of mAb(2)9F1, suggesting that the functional bivalency of mAb(1)78 was not impaired by this anti-Id antibody. However, bivalent mAb(2)9F1 was able to partially inhibit the binding of bivalent mAb(1)78 to oligomeric TNF in liquid-phase as well as in solid-phase assays, whereas no inhibition was observed with monovalent mAb(2)9F1-F(ab) or mAb(1)78-F(ab). This suggests that inhibition is based on a decrease of the avidity of bivalent mAb(1)78 and not on allosteric effects on antigen binding sites. The effect of mAb(2)9F1 on mAb(1)78 arm flexibility and paratope orientation is discussed. In conclusion, the results indicate that anti-Id antibodies may inhibit Ag-antibody multivalent interactions by paratope blockade or by affecting the antibody avidity.

Peptide-based assay for the identification of endothelin-converting enzyme inhibitors

A simple, solid-phase assay usable for the detection of endothelin-converting enzyme activity and inhibitors has been developed. It uses a multimeric peptide immobilized on microtiter plates that is able to specifically recognize the Big Endothelin fragment 16-32 derivatized with biotin. This fragment is cleaved between residues 21-22 by alpha-chymotrypsin with almost the same proteolysis rate as Big Endothelin, and after enzyme treatment it does not bind to the multimeric peptide adsorbed on the microtiter plates. The amount of uncleaved peptide bound to the plate is detected by subsequent treatment with streptavidin conjugated to peroxidase, followed by a chromogenic reaction. Model inhibitor profiles generated for alpha-chymotrypsin, a protease known to convert Big Endothelin in endothelin, demonstrated the utility of this assay as a rapid high-throughput aid in the study of Big Endothelin enzymatic processing and possibly in the identification of putative enzyme inhibitors.

AB023, novel polyene antibiotics. II. Isolation and structure determination

AB023, a complex of new polyene antibiotics, was isolated from a soil Streptomyces strain. The two main components, pentaene antibiotics AB023a and AB023b, were separated and purified by preparative chromatographic methods and their structures were determined by extensive NMR and mass spectrometric studies.

Detection of human chromosomes in somatic cell hybrids by PCR analysis

The detection of human chromosomes in somatic cell hybrids is usually made by chromosomal analysis, Southern blot analysis with human probes, and starch-gel electrophoresis of isoenzymes. We describe here a new, quick, and very efficient method to detect human chromosomes in somatic cell hybrids between human and rodent (rat and mouse) cells. The method is based on the polymerase chain reaction to promote amplification of human DNA, using primers derived from localized genes or DNA fragments from each human chromosome.

Binding of human tumor necrosis factor alpha to multimeric complementary peptides

A peptide with binding properties for tumor necrosis factor (TNF alpha) sequence 144-157 has been designed, using a computer-assisted method able to create peptide sequences hydropathically complementary to a given sequence. The complementary peptide was synthesized in a multimeric form starting from an octadentate polylysine core, to facilitate its immobilization and to provide interaction multivalency. Once immobilized on a solid support to prepare an affinity column, it recognized the target TNF144-157 peptide selectively from crude peptide mixtures containing TNF fragments encompassing the entire TNF alpha sequence. Similar selectivity and specificity were shown for full-length recombinant TNF alpha, allowing its purification from crude Escherichia coli extracts. The octameric complementary peptide preserved its recognition properties for TNF alpha and biotinylated TNF alpha even after coating on microtiter plates. Competitive binding occurred with unlabeled TNF alpha in the range between 0.01 and 10 micrograms/ml, in the presence of detergent such as 0.05% Tween 20 and in the presence of 1% normal goat serum. The effect of complementary peptide multimerization was evidenced by its enhanced binding affinity for TNF alpha, which exists in solution as a trimer, while the target TNF[144-157] peptide was recognized with much lower strength. The dissociation constant for interaction with TNF alpha was close to 10 nM, allowing its easy detection by solid phase assays in concentrations as low as 10 pmol/ml.

Oligomeric tumour necrosis factor alpha slowly converts into inactive forms at bioactive levels

The stability of oligomeric human tumour necrasis factor alpha (TNF) at bioactive levels has been studied by two immunoenzymatic assays: one able to specifically detect oligomeric and not monomeric TNF (O-e.l.i.s.a.) and the other able to detect both forms (OM-e.l.i.s.a.). The selectivity of O-e.l.i.s.a. and OM-e.l.i.s.a. for oligomeric and monomeric TNF was demonstrated with isolated forms prepared by partial dissociation of recombinant TNF with 10% (v/v) dimethyl sulphoxide and gel-filtration h.p.l.c. Evidence for instability of oligomeric TNF were obtained in physiological buffers, as well as in serum and cell-culture supernatants, as a function of TNF concentration. In particular, only a half of the TNF antigen was recovered in the oligomeric form after 72 h incubation (37 degrees C) at 0.12 nM, whereas no apparent dissociation was detected at 4 nM. The structural changes observed at picomolar concentrations were rapidly reversed by raising the concentration of TNF to about 2 nM by ultrafiltration, suggesting that subunit dissociation and reassociation reactions occur in the picomolar and nanomolar range respectively. The cytolytic activity of L-M cells correlates with oligomeric-TNF levels after incubation at picomolar concentrations. Moreover, isolated oligomeric TNF was cytotoxic towards L-M cells, whereas monomeric TNF was virtually inactive. In conclusion, the results suggest that bioactive oligomeric TNF is unstable at picomolar levels and slowly converts into inactive monomers, supporting the hypothesis that quaternary-structure changes in TNF may contribute to the fine regulation of TNF cytotoxicity.

Design of hydropathically complementary peptides for Big Endothelin affinity purification

Molecular recognition between Big Endothelin (Big ET) and a computer generated peptide hydropathically complementary to Big ET[16-29] sequence has been studied by analytical high performance liquid affinity chromatography (HPLAC), circular dichroism (CD) and nuclear magnetic resonance (NMR) experiments. Specific binding was observed between solid support immobilized complementary peptide and Big ET[1-38], [1-32], and [16-32], but not with Big ET fragments [1-21], [16-21], [22-32], and [22-38], obtained by chymotrypsin proteolytic degradation. Selectivity in the recognition process was clearly demonstrated by the ability of complementary peptide affinity column to purify the Big ET molecule from complex peptide mixtures, even when present in very low concentrations. Similar selectivity was evidenced with the Big ET fragment [16-32], [NH2-HLDIIWVNTPEHIVPYG-COOH] containing the entire hydropathically complementary sequence. Binding was followed by marked spectroscopic changes, as monitored by circular dichroism and one- and two-dimensional nuclear magnetic resonance experiments. The NMR spectra of the complementary peptides 1:1 mixture showed variations in the chemical shifts of proton resonances in several residues, both in the main chain (amide protons) and in the side chains (aliphatic and aromatic protons). These data support the hypothesis of a multilocalized type of interaction between complementary peptides, where many residues along the peptide chains participate in co-operative stabilizing contacts in the forming complex.

Affinity enhancement of complementary peptide recognition

A peptide hydropathically complementary to Big Endothelin [Big ET] residues 16-29 has been synthesized in a multimeric form starting from an octadentate polylysine core, essentially in a way similar to the procedure used for the production of multiple antigenic peptides [MAP's]. Interaction between the multimeric complementary peptide [8 delta ET] and the Big ET fragment 16-32 containing the target complementary region, also synthesized in a multimeric form [8ET], was evaluated by analytical high performance affinity chromatography and solid phase binding assays. While the binding interaction between the monomerics peptide pair was in the micromolar range, the recognition between the corresponding multimeric form was characterized by enhanced binding affinity of at least two orders of magnitude. In solution, complex formation between multimeric complementary peptide and target Big ET sequence in the monomeric and multimeric form was accompanied by precipitation at concentrations higher than 0.5 mg/mL and 0.1 mg/mL, respectively. Polyclonal antibodies raised against the multimeric target sequence recognized multimeric and monomeric ET target sequences with binding affinities similar to binding affinities exhibited by the multimeric complementary peptide. Multimerization of hydropathically complementary peptides could provide an improved opportunity to measure and thus probe quantitative binding properties of complementary peptides.

Antigenic regions of tumor necrosis factor alpha and their topographic relationships with structural/functional domains

The immunogenic regions of human Tumor Necrosis Factor alpha (huTNF) have been mapped by studying the interaction between various mouse anti-huTNF sera and synthetic huTNF fragments, spanning the entire sequence of huTNF. Three main immunogenic regions were identified within residues 1-23, 95-116 and 137-157 of huTNF and two other less immunogenic regions within residues 117-136 and 37-55. The same huTNF regions were found to contain antigenic sites by binding studies with cognate anti-peptide sera. Competitive binding experiments with shorter synthetic subfragments provided evidence for the location of strong antigenic sites within residues 1-10, 17-23, 104-112 and 137-143. In particular the immunodominant site was found to be located within residues 104-112. huTNF regions corresponding to residues 24-36, 56-75, 76-94, and 147-157 resulted to be not or poorly antigenic. However, treatment of huTNF with Triton X-100 under conditions that partially dissociate the oligomeric quaternary structure resulted in the exposition of sites recognized by sera against peptides huTNF [56-75] and huTNF [76-94], suggesting that antigenic sites not accessible in the oligomeric huTNF are exposed in the dissociated form. The principal antigenic sites in the oligomeric molecule are localized in the flexible N-terminal part and in hydrophilic segments located in the "middle/top" region of the molecule, opposite to the C-terminus. Protein segments of the "bottom" region, close to the C-terminus, were poorly immunoreactive. Neutralization assays of TNF cytolytic activity on L-M cells showed that binding of antibodies to epitopes located in the "middle/top" regions of huTNF does not affect its cytolytic activity, supporting the hypothesis of a receptor binding site location at the "bottom" of TNF trimer.

Design and recognition properties of a hydropathically complementary peptide to human interleukin 1 beta

A computer-designed hydropathically complementary peptide to human interleukin 1 beta (IL1 beta) precursor sequence 204-215 recognized the 204-215 peptide as well the entire IL1 beta protein with binding affinities in the micromolar range. Interaction between the complementary pair was characterized by analytical high-performance liquid affinity chromatography on columns derivatized with the computer-generated peptide. Recognition selectivity was clearly shown by the ability of the computer-generated complementary peptide columns to purify the IL1 beta-(204-215)-peptide from complex synthetic mixtures with high yields, independently of the type of solid support used. Recognition specificity was demonstrated by the inability of the IL1 beta-(204-215)-peptide and IL1 beta molecules to interact with blank columns or columns derivatized with other non-related peptides. Furthermore, scrambling the sequence of the computer-generated peptide or the IL1 beta-(204-215)-peptide in such a way as to alter their hydropathic profiles had the effect of abolishing binding. The complementary pair failed to interact in the presence of competing peptide, thus providing further evidence of specificity. Computer-generated complementary peptide affinity columns also proved useful for purification of recombinant human IL1 beta protein directly from crude Escherichia coli lysates.

Assignment of the urokinase-type plasminogen activator receptor gene (PLAUR) to chromosome 19q13.1-q13.2

The urokinase-type plasminogen activator receptor (uPAR) is a key molecule in the regulation of cell-surface plasminogen activation and, as such, plays an important role in many normal as well as pathological processes. We applied a cDNA probe from the corresponding gene (PLAUR) in a location analysis using a panel of human/rodent cell hybrids and in a multipoint linkage analysis of 40 CEPH families. These two independent studies both found PLAUR to be located on chromosome 19. The cell hybrid study suggested that PLAUR is located at chromosome 19q13-qter, and the multipoint analysis indicated that PLAUR is located at chromosome 19q13.1-q13.2 and surrounded by DNA markers in the following way (with distances given in recombination fractions): D19S27-.11-CYP2A-.06-PLAUR-.03-D19S8-.04-APOC 2-.24-PRKCG. Further, a ligand-binding study performed on cell hybrids verified the species specificity of the uPAR and confirmed the chromosome assignment.

Rapid and simple DNA detection by hybridization on absorbent filters

Images

A rapid method for monitoring DNA labelling reactions with haptens

A rapid and simple method for evaluating the efficiency of DNA labelling reactions with haptens is described. The method, called the Flow-Through Hapten-DNA Assay (FT-HDA), relies on binding of anti-hapten antibodies/alkaline phosphatase conjugates to hapten-DNA, immobilized on disposable capillary absorbent filters, and visual detection of blue-grey coloured spots appearing on the filter after chromogenic reaction with enzyme substrates. FT-HDA of hapten-DNA is markedly faster and simpler than conventional diffusion assays on membranes.

The differential glycosylation of human pro-urokinase from various recombinant mammalian cell lines does not affect activity and binding to PAI-1

Human chromosomal DNA encoding single-chain urokinase-type Plasminogen Activator (scu-PA, or pro-urokinase) was inserted in an expression plasmid and transfected in human A431, mouse LB6 and CHO cells. LB6 cells were also transfected with a Bovine Papilloma Virus derivative containing the scu-PA gene. Human scu-PA was purified from cell supernatants of recombinant clones and characterized for structure and function. All recombinant scu-PAs are undistinguishable from human urine-derived scu-PA for peptide backbone, but possess a higher sugar content, as revealed by SDS-PAGE analysis after digestion with glycopeptidase F. This difference is partly due to an increased sialic acid content, as shown by analysis of neuraminidase-treated scu-PAs. No difference was found, however, among recombinant and natural scu-PAs in the kinetics of conversion into two-chain active forms (tcu-PAs) by human plasmin, and in the KM and kcat values of tcu-PA activity on the chromogenic substrate S-2444 and on human plasminogen. Also, recombinant and non-recombinant tcu-PAs displayed similar dose-response curves for binding to the endothelial inhibitor PAI-1. In conclusion, the glycosylation pattern of u-PA does not affect its interaction with the plasma proteins directly involved in its fibrinolytic function.

Epitope mapping of the anti-urokinase monoclonal antibody 5B4 by isolated domains of urokinase

The amino terminal fragment (ATF) of urokinase-type plasminogen activator (uPA) is a degradation product comprising the entire growth factor-like and kringle domains. It has been previously shown that ATF is able to bind to the u-PA receptor through the growth factor-like domain and that the anti u-PA monoclonal antibody 5B4 (Mab 5B4) binds to ATF preventing u-PA receptor binding. To localize more precisely the epitope recognized by Mab 5B4, ATF was subfragmented by controlled enzymatic proteolysis with V8 protease. Three subfragments of 4,000 Mr (F-4k), 11,000 Mr (F-11k) and 12,000 Mr (F-12k) were purified from the reaction mixture and characterized. SDS-PAGE under reducing and non-reducing conditions, N-terminal aminoacid sequence analysis and C-terminal aminoacid analysis of each fragment indicate that F-4k and F-11k correspond to intact growth factor-like domain and kringle domain (residues 4-43 and 44-135 respectively) while F-12k corresponds to the kringle domain cleaved in the first loop at the glu52-gly53 bond. By Western blot and competitive binding experiments we show that Mab 5B4 recognizes an epitope located on the kringle domain of u-PA and that the binding is strongly reduced when the kringle contains an additional cleavage in its first loop. Since the receptor binding site of u-PA has been previously shown to be located on the growth factor-like domain, Mab 5B4 inhibits the binding of uPA to its cellular receptor likely by steric hindrance. Besides the proven utility in epitope localization of anti u-PA monoclonal antibodies, these u-PA fragments may represent powerful tools for studies of structure-function relationship of u-PA.

Charge heterogeneity of recombinant pro-urokinase and urinary urokinase, as revealed by isoelectric focusing in immobilized pH gradients

When analysing homogeneous preparations of recombinant pro-urokinase and urinary urokinase by isoelectric focusing (IEF) in immobilized pH gradients, an extreme charge heterogeneity was detected (at least ten major and ten minor bands in the pH range 7-10). This extensive polydispersity was not caused by different degrees of glycosylation, or by IEF artefacts, such as binding to carrier ampholytes or carbamylation by urea. A great part of this heterogeneity could be traced back to the existence of a multitude of protein molecules containing Cys residues at different oxidation levels (-SH, -S-S-, even cysteic acid). Owing to the very large number of Cys residues in pro-urokinase (24 out of a total of 411 amino acids) and to the relatively high pI of its native forms (pI 9.5-9.8; the native form is believed to contain all Cys residues as -S-S- bridges), the presence of SH or cysteic acid residues would increase the negative surface charge, as even SH groups would be extensively ionized. In pro-urokinase, part of the heterogeneity was also due to spontaneous degradation to urokinase and possibly also to cleavage into lower-molecular-mass fragments. When all these causes of heterogeneity were removed, the pI spectrum was reduced to only four, about equally intense bands. The cause of this residual heterogeneity is unknown.