Src homology-2 domain binding assays by scintillation proximity and surface plasmon resonance.

Various SH2 competitive binding assays, based on different techniques, have been described in the literature to identify and characterize SH2 ligands. The consideration that most reported methods show experimental limitations associated with assay parameters has prompted us to base our Src-SH2 inhibitor discovery program on the use of two different assays. In this study, two conceptually different biochemical methods designed to discover Src-SH2 inhibitors, respectively scintillation proximity assay (SPA) and surface plasmon resonance (SPR), have been evaluated and compared. For its high sensitivity and adaptability to automation SPA was chosen for high capacity screening (primary screen), whereas SPR was used for hits confirmation (secondary screening). However with the drastic improvement of inhibitor affinities, the limit of sensitivity was rapidly reached for the SPR assay based on the canonical pYEEI ligand. The substitution of the natural, monophosphorylated peptide ligand with a triphosphorylated peptide has allowed us to remarkably increase its sensitivity, so that molecules with nanomolar affinities could be easily differentiated in terms of IC(50) ranking. Such a new, improved SPR assay can be of great interest for the study of high affinity ligands of different SH2-based drug targets.

Discovery of thioazepinone ligands for Src SH2: from non-specific to specific binding.

The structure-based design and synthesis of new thioazepinones as ligands for Src SH2 protein is presented. From benzothioazepinones, ligands with somewhat unspecific binding properties, simpler thioazepinones were designed, the best ones demonstrated nanomolar affinity for Src SH2. A few of these new ligands were crystallized with the protein and demonstrated a specific binding mode with the protein.

Endogenous glutathione-binding proteins of insect cell lines: characterization and removal from glutathione S-transferase (GST) fusion proteins.

After affinity purification on immobilized glutathione, insect-cell-derived glutathione S-transferase (GST) fusion proteins contain variable amounts of protein contaminants of about 23-24 kDa. We have isolated these glutathione-binding proteins from the widely used Sf9 and Hi5 insect cell lines and characterized them by LC-MS and N-terminal sequencing. Based on the observation that these proteins have higher affinity for glutathione than GST fusions, we have found that by using differential elution conditions the amount of such contaminants in GST fusion preparations can be strongly reduced directly during the affinity purification step. The main interest of these results is that they are not restricted to a specific construct, but rather they seem to apply to various insect-cell-derived GST fusions.

Human IL-1 receptor antagonist from Escherichia coli: large-scale microbial growth and protein purification.

Interleukin-1 receptor antagonist (IL-1ra) is a recently discovered cytokine which specifically inhibits IL-1 pro-inflammatory activities in various experimental conditions. In this work, the growth conditions of a recombinant E. coli strain which in laboratory studies expressed human IL-1ra mostly in insoluble form, have been optimized at the level of 6-1 bioreactors and then scaled up to a 50-1 process. As a result, a high amount (0.43 g l-1 of microbial culture) of soluble, active IL-1ra has been directly obtained in the large-scale cell lysate with no need for protein solubilization. Also, an efficient purification procedure has been developed for the soluble protein, based on cation exchange expanded bed adsorption directly followed by anion exchange chromatography. This process, which does not include any intermediate dialysis step or gradient elutions, can be easily scaled up to larger production volumes and is therefore well-suited for manufacturing. As a result of the overall optimization study, more than 12 g of pure IL-1ra have been obtained from a single 50-1 fermentation run, without any denaturation/renaturation process. The final product, whose identity and purity have been checked also by MALDI-TOF and ESI-MS, shows full biological activity both in cellular assays and in in vivo experiments with Cynomolgus monkeys.

Inhibitors of type-I interleukin-1 receptor from microbial metabolites.

We describe here the results of a screening program conducted to discover inhibitors of the type-I interleukin-1 receptor (IL-1RI) from samples of microbial origin. An innovative approach, based on automated, nonradioactive receptor binding assays has been employed. Specially prepared cell-free systems have allowed the use of high concentrations of microbial metabolites in the reaction mixtures with a low percentage of false positives. More than 30,000 microbial samples from different species of soil isolates have been tested and two interesting activities have been purified and characterized. One of these, isolated from Streptomyces sp. GE48009, was identified as niphimycin, an antifungal agent also known as scopafungin. Preliminary evidence suggests that this molecule and azalomycin F, a structural analogue, inhibit IL-IRI by virtue of their long-chain guanidinium moiety. The other activity, isolated from Aspergillus sp. GE49752, was identified as flavipin, a substituted o-phthalaldehyde.

A new cytokine-receptor binding mode revealed by the crystal structure of the IL-1 receptor with an antagonist.

Inflammation, regardless of whether it is provoked by infection or by tissue damage, starts with the activation of macrophages which initiate a cascade of inflammatory responses by producing the cytokines interleukin-1 (IL-1) and tumour necrosis factor-alpha (ref. 1). Three naturally occurring ligands for the IL-1 receptor (IL1R) exist: the agonists IL-1alpha and IL-1beta and the IL-1-receptor antagonist IL1RA (ref. 2). IL-1 is the only cytokine for which a naturally occurring antagonist is known. Here we describe the crystal structure at 2.7 A resolution of the soluble extracellular part of type-I IL1R complexed with IL1RA. The receptor consists of three immunoglobulin-like domains. Domains 1 and 2 are tightly linked, but domain three is completely separate and connected by a flexible linker. Residues of all three domains contact the antagonist and include the five critical IL1RA residues which were identified by site-directed mutagenesis. A region that is important for biological function in IL-1beta, the 'receptor trigger site' is not in direct contact with the receptor in the IL1RA complex. Modelling studies suggest that this IL-1beta trigger site might induce a movement of domain 3.

Particle concentration fluorescence as an alternative to radioactivity for cytokine receptor binding assays.

To detect inhibitors of the type-I interleukin-1 receptor in a safe and cost-effective nonisotopic environment, the use of particle concentration fluorescence (PCF) as a detection method for a high volume receptor binding assay (RBA) has been explored. A cell-free PCF-RBA has been developed using a soluble form of the type-I IL-1 receptor (sIL-1R), a nonneutralizing anti-sIL-1R monoclonal antibody and two fluorescein-labelled tracers: IL-1 alpha and IL-1ra fused to the Escherichia coli maltose binding protein (MBP-IL-1ra). Both ligands showed saturable specific binding (KA values in the 10(9) M-1 range), although signal intensity was about 3-fold higher with MBP-IL-1ra. The data presented here suggest that the combined use of recombinant DNA and PCF technologies can replace the traditional cell-based, radioactive RBA when screening for inhibitors of cytokine receptors.

An elongation factor Tu (EF-Tu) resistant to the EF-Tu inhibitor GE2270 in the producing organism Planobispora rosea.

Using a cell-free protein-synthesis system, we have established that the elongation factor (EF) Tu (EF-Tu) of the actinomycete Planobispora rosea, the producer of the thiazolyl peptide GE2270, a specific EF-Tu inhibitor, is highly resistant to its own antibiotic, while it is completely inhibited by kirromycin, which is another inhibitor of this factor. P. rosea was found to possess a single tuf gene, located between fus and rpsJ, encoding other components of the protein-synthesis machinery. The P. rosea tuf gene was expressed as a translational fusion to malE in Escherichia coli, and the resulting EF-Tu with an N-terminal Gly-Met extension was able to promote poly(U)-directed poly(Phe) synthesis in cell-free systems. This activity was not affected by GE2270, and the recombinant protein was incapable of binding the antibiotic, indicating that the P. rosea EF-Tu is intrinsically resistant to this inhibitor. Inspection of the translated tuf sequence revealed a number of amino acid substitutions in highly conserved positions. These residues, which are likely to be involved in conferring GE2270 resistance, map in EF-Tu domain II, as do the only two known mutations conferring resistance to this class of thiazolyl peptides in Bacillus subtilis.

A cell-free, nonisotopic, high-throughput assay for inhibitors of type-I interleukin-1 receptor.

A cell-free, nonisotopic assay has been developed to discover molecules that compete with the natural ligands for binding to the active site of the Type-I interleukin-1 receptor. The key reagents are the interleukin-1 receptor antagonist, a recombinant soluble form of the receptor (sIL-1R), and a specific anti-sIL-1R nonneutralizing monoclonal antibody (MAb79). With these molecules a sensitive assay has been developed using a reversed format: the ligand is immobilized and the receptor is in solution. The ligand-bound receptor is detected using MAb79 and an enzyme-linked secondary antibody. Since no cells or cell membranes are used, the assay is very robust, with no interference from membrane-perturbing agents and high resistance to the organic solvents normally used to resuspend compounds of chemical libraries. The microplate format and colorimetric detection have allowed the complete automation of the immobilized-ligand IL-1 receptor binding assay, which has been used for high-throughput screening of synthetic compounds and natural products.

Crystals of soluble interleukin-1 receptor complexed with its natural antagonist reveal a 1:1 receptor-ligand complex.

Interleukin-1 is a cytokine involved in the acute phase response against infection and injury. We obtained crystals of a complex of soluble, recombinant human interleukin-1 receptor and recombinant human interleukin-1 receptor antagonist, a naturally occurring antagonist. The crystals are suitable for X-ray analysis and diffract to 2.7 A resolution. Solvent content calculations indicate that the crystals contain one receptor and one antagonist molecule per asymmetric unit. Other receptor to antagonist ratios are highly unlikely. These results suggest that the interleukin-1 antagonist binds a single receptor molecule and does not cause receptor aggregation.

Refined crystal structure of the interleukin-1 receptor antagonist. Presence of a disulfide link and a cis-proline.

Interleukin-1 (IL-1) molecules are cytokines involved in the acute-phase response against infection and injury. Three naturally occurring IL-1 molecules are known, two agonists: IL-1 alpha and IL-1 beta, and one antagonist, the IL-1 receptor antagonist (IL-1ra). Although IL-1 action protects the organism by enhancing the response to pathogens, its overproduction can lead to pathology and has been implicated in disease states that include septic shock, rheumatoid arthritis, graft versus host disease and certain leukemias. The crystal structure of IL-1ra has been solved at 0.21-nm resolution by molecular replacement using the IL-1 beta structure as a search model. The crystals contain two independent IL-1ra molecules which are very similar. IL-1ra has the same fold as IL-1 alpha and IL-1 beta. The fold consists of twelve beta-strands which form a six-stranded beta-barrel, closed on one side by three beta-hairpin loops. Cys69 and Cys116 are linked via a disulfide bond and Pro53 has been built in the cis-conformation. Comparison of the IL-1ra structure with the IL-1 alpha and IL-1 beta structures present in the Protein Data Bank shows that a putative receptor interaction region, involving the N-terminus up to the beginning of strand beta 1 and the loops D and G, is very different in the three IL-1 molecules. Other putative interaction regions, as identified with mutagenesis studies, are structurally conserved and rigid, allowing precise and specific interactions with the IL-1 receptor.

Epitope mapping of a monoclonal antibody which binds HIV-1 Gag and not the Gag-derived proteins.

Monoclonal antibody (MAb) 1G12 binds the uncleaved HIV-1 Gag polypeptide (p55), but fails to recognize the final products of the proteolytic processing [Sarubbi, E. et al. (1991) FEBS Lett. 279, 265-269]. In this report we show that binding of MAb 1G12 to a 110-residue Gag fragment containing the p17-p24 cleavage site prevents proteolysis of this site by the HIV-1 protease. Competition studies with synthetic peptides have been performed to map the binding site of MAb 1G12 on Gag. The antibody recognizes a sequential epitope that spans the HIV-1 protease cleavage site; determinants located on both p17 and p24 are required for antibody binding. MAb 1G12 is also shown to lack any cross-reactivity with other HIV-1 protease cleavage sites.

Peptide aldehydes as inhibitors of HIV protease.

We have recently shown that alpha-MAPI, a peptidic aldehyde of microbial origin, inhibits the HIV protease with a potency comparable to pepstatin, having, differently from pepstatin, no activity on other aspartic proteases. In this study different peptide derivatives containing a C-terminal aldehyde have been tested to assess the potential of this function for the inhibition of HIV protease. The results of our analysis correspond with the recently published subsite preferences of the viral enzyme, indicating that aldehydes bind to the active site of the HIV protease. Our data suggest that peptide aldehydes can act in their hydrated forms as transition state analogues with the most potent inhibitor having an IC50 of 0.9 microM.

A high throughput assay for inhibitors of HIV-1 protease. Screening of microbial metabolites.

A novel method for discovery of HIV-1 protease inhibitors in complex biological samples has been developed. The assay is based on two specific reagents: a recombinant protein constituted by a portion of the HIV-1 Gag polyprotein comprising the p17-p24 cleavage site, fused to E. coli beta-galactosidase, and a monoclonal antibody which binds the fusion protein in the Gag region. Binding occurs only if the fusion protein has not been cleaved by the HIV-1 protease. The assay has been adapted for the screening of large numbers of samples in standard 96-well microtiter plates. Using this method about 12000 microbial fermentation broths have been tested and several HIV-1 protease inhibitory activities have been detected. One of these has been studied in detail.

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.

Characterization of the spoT gene of Escherichia coli.

The Escherichia coli spoT gene encodes a guanosine-3',5'-bispyrophosphate (ppGpp) 3'-pyrophosphohydrolase known to be responsible for cellular (ppGpp) degradation. The DNA sequence of the spoT region is presented. The spoT gene is deduced to be 702 codons long, with a probable UUG initiation codon, and a deduced mass of 79,342 daltons. Two spoT mutations (spoT202 and spoT203) have been localized to an open reading frame by complementation of function as well as by genetic marker rescue. The ability to overexpress the spoT gene is limited, but enough ppGppase activity can be made to reverse ppGpp accumulation during the stringent response to amino acid starvation. The spoT gene is located within a larger spo operon and is flanked by two smaller genes. The first gene in the operon encodes omega, a protein that copurifies with RNA polymerase (Gentry, D. R., and Burgess, R. R. (1986) Gene (Amst.) 48, 33-40). The spoT gene is the second gene in the operon; it is followed by a third open reading frame deduced to encode a protein with a mass of 25,343 daltons. Insertion of a kanamycin resistance gene in the omega gene reduces spoT gene expression as judged by lowered ppGppase activity, relA-dependent reduction of growth rate, and abolition of spoT mutant complementation activity. These effects are reversed by expression of the spoT gene, but not the omega gene, in trans. Transcription of the spo operon occurs in a clockwise direction on the E. coli chromosome and is probably directed by at least two promoters.

Protein sequences encoded by the relA and the spoT genes of Escherichia coli are interrelated.

relA and spoT are designations for two unlinked Escherichia coli genes whose products function in the synthesis and degradation of guanosine 3',5'-bispyrophosphate during the stringent regulatory response to amino acid deprivation. The RelA protein catalyzes an ATP:GTP 3'-pyrophosphoryl group transfer reaction, and the SpoT protein has a guanosine 3',5'-bispyrophosphate 3'-pyrophosphohydrolyase activity. Both genes have been sequenced recently; the relA gene produces an 84-kDa protein, and the spoT gene is deduced to encode a 79-kDa protein. We report here that the protein sequences of the relA and spoT genes are extensively interrelated.