Identification of novel potent hydroxamic acid inhibitors of peptidyl deformylase and the importance of the hydroxamic acid functionality on inhibition.

Peptidyl deformylase (PDF) is a metallo protease that catalyzes the removal of a formyl group from the N-termini of prokaryotic prepared polypeptides, an essential step in bacterial protein synthesis. Screening of our compound collection using Staphylococcus aureus PDF afforded a very potent inhibitor with an IC(50) in the low nanomolar range. Unfortunately, the compound that contains a hydroxamic acid did not exhibit antibacterial activity (MIC). In order to address the lack of activity in the MIC assay and to determine what portion of the molecule was responsible for binding to PDF, we prepared several analogues. This paper describes our findings that the hydroxamic acid functionality found in 1 is mainly responsible for the high affinity to PDF. In addition, we identified an alternative class of PDF inhibitors, the N-hydroxy urea 18, which has both PDF and antibacterial activity.

Identification of a C-terminal cdc25 sequence required for promotion of germinal vesicle breakdown.

Glutathione S-transferase (GST)-cdc25B(31-566) induced germinal vesicle breakdown (GVBD) when microinjected into Xenopus oocytes. Purified, N-terminally truncated forms of cdc25B did not induce GVBD, even though many had phosphatase activity and activated cdc2 in vitro. N-terminally truncated forms of cdc25B inhibited induction of GVBD by longer forms of the enzyme suggesting a direct interaction in vivo. cdc25B(356-556), but not cdc25B(364-529), inhibited GVBD induction by GST-cdc25B(31-566) suggesting that a region of cdc25B near to the C-terminus was responsible for the inhibition. To determine the region of peptide sequence that was inhibitory, cdc25B(356-556) was subjected to proteolysis with endoproteinase lys-C. Following a demonstration that the resulting peptide mixture inhibited GST-cdc25B-dependent GVBD, a series of peptides spanning amino acids at the C-terminus were synthesized. The peptide TRSWAGERSR inhibited GVBD induced by GST-cdc25B. An alanine scan of the peptide revealed residues critical for GVBD inhibition, and site-directed mutagenesis of the corresponding residues in GST-cdc25B(31-566) eliminated its ability to induce GVBD. These results demonstrate that a cdc25B C-terminal domain, involved in dominant-negative inhibition of GVBD-competent cdc25B, is required for induction of GVBD following microinjection into oocytes.

Crystal structure of the catalytic subunit of Cdc25B required for G2/M phase transition of the cell cycle.

Cdc25B is a dual specificity phosphatase involved in the control of cyclin-dependent kinases and the progression of cells through the cell cycle. A series of minimal domain Cdc25B constructs maintaining catalytic activity have been expressed. The structure of a minimum domain construct binding sulfate was determined at 1.9 A resolution and a temperature of 100 K. Other forms of the same co?nstruct were determined at lower resolution and room temperature. The overall folding and structure of the domain is similar to that found for Cdc25A. An important difference between the two is that the Cdc25B domain binds oxyanions in the catalytic site while that of Cdc25A appears unable to bind oxyanions. There are also important conformational differences in the C-terminal region. In Cdc25B, both sulfate and tungstate anions are shown to bind in the catalytic site containing the signature motif (HCxxxxxR) in a conformation similar to that of other protein tyrosine phosphatases and dual specificity phosphatases, with the exception of the Cdc25A. The Cdc25B constructs, with various truncations of the C-terminal residues, are shown to have potent catalytic activity. When cut back to the site at which the Cdc25A structure begins to deviate from the Cdc25B structure, the activity is considerably less. There is a pocket extending from the catalytic site to an anion-binding site containing a chloride about 14 A away. The catalytic cysteine residue, Cys473, can be oxidized to form a disulfide linkage to Cys426. A readily modifiable cysteine residue, Cys484, resides in another pocket that binds a sulfate but not in the signature motif conformation. This region of the structure is highly conserved between the Cdc25 molecules and could serve some unknown function.

Fluorescence and site-directed mutagenesis studies of interleukin 1 beta.

The authors mutated two key residues in the sequence of the cytokine interleukin 1 beta, namely the double mutant Phe46 to Trp46 and Trp120 to Phe120 and the single point mutation Lys103 to Leu103 and measured the resulting receptor binding and biological activities. The biological and receptor binding activities of the Trp46 mutein was reduced by a factor of 12 and 25, respectively, and surprisingly, those of the Leu103 mutein, 2600 and 600-fold relative to the wild-type protein. The authors had previously showed that Lys103 was unusually reactive to a variety of derivatizing agents. Furthermore, the Trp to Phe mutation allowed us to monitor the local environment of that residue by studying its intrinsic fluorescence properties, as well as any change in the fluorescence properties of Trp120 of the Leu103 mutein. The results of these studies show that mutation of Lys103 to Leu103 produces subtle long-range changes in the micro-environment of Trp120, indicative of a key role for this residue in the folding of the entire protein.

The cyclosporin A-binding immunophilin CyP-40 and the FK506-binding immunophilin hsp56 bind to a common site on hsp90 and exist in independent cytosolic heterocomplexes with the untransformed glucocorticoid receptor.

We have recently shown that hsp56, the FK506-binding immunophilin component of both the heat shock protein (hsp90.hsp70.hsp56) heterocomplex and the untransformed glucocorticoid receptor heterocomplex, is bound directly to hsp90 (Czar, M. J., Owens-Grillo, J. K., Dittmar, K. D., Hutchison, K. A., Zacharek, A. M., Leach, K. L., Deibel, M. R., and Pratt, W. B. (1994) J. Biol. Chem. 269, 11155-11161). In this work, we show that both untransformed glucocorticoid receptor and hsp90 heterocomplexes contain CyP-40, a 40-kDa immunophilin of the cyclosporin A-binding class. CyP-40 is present in both native glucocorticoid receptor heterocomplexes and receptor heterocomplexes reconstituted with rabbit reticulocyte lysate, and the presence of CyP-40 in the receptor heterocomplex is stabilized by molybdate. Immunoadsorption of hsp90 from cell lysate yields coimmunoadsorption of both hsp56 and CyP-40, showing that both immunophilins are in native heterocomplex with hsp90. However, immunoadsorption of hsp56 does not yield coimmunoadsorption of CyP-40; thus, the two immunophilins do not exist in the same heterocomplex with hsp90. Both purified CyP-40 and hsp56 bind directly to purified hsp90, and excess CyP-40 blocks the binding of hsp56, consistent with the presence of a common immunophilin binding site on hsp90. Our data also suggest that there are at least two types of untransformed glucocorticoid receptor-hsp90 heterocomplexes, one that contains hsp56 and another that contains CyP-40. The role played by the immunophilins in steroid receptor action is unknown, but it is clear that the peptidylprolyl isomerase activity of immunophilins is not required for glucocorticoid receptor-hsp90 heterocomplex assembly and proper folding of the hormone binding domain by the hsp90-associated protein folding system of reticulocyte lysate.

The hsp56 immunophilin component of untransformed steroid receptor complexes is localized both to microtubules in the cytoplasm and to the same nonrandom regions within the nucleus as the steroid receptor.

In their unliganded state, mouse glucocorticoid receptors (GR) that are overexpressed in the WCL2 line of Chinese hamster ovary cells are distributed in a nonrandom manner throughout all planes of the nucleus. These untransformed nuclear receptors exist in a heterocomplex containing three heat shock proteins, hsp90, hsp70, and hsp56, the latter being an immunophilin of the FK506 binding type whose cellular function is unknown. Because a knowledge of the cellular distribution of hsp56 could provide important clues to its function in steroid-receptor heterocomplexes, we have examined hsp56 localization in intact cells by indirect immunofluorescence using the UPJ56 antibody. The majority of hsp56 is located in the nucleus, with substantial amounts also visualized in the cytoplasm of intact cells. The cytoplasmic hsp56 was examined in rat pulmonary endothelial cells where the protein was found to colocalize with microtubules. The nuclear hsp56 was examined in the WCL2 cells, where the protein was found by confocal imaging to colocalize throughout all planes of the nucleus in the same mottled pattern as the overexpressed GR. Like the GR, the nuclear hsp56 is recovered largely in the cytosolic fraction after hypotonic rupture of WCL2 cells. An observation potentially related to the microtubule-associated fraction of hsp56 is that immunoadsorption of hsp56 from WCL2 cytosol is accompanied by coadsorption of the microtubule-associated protein-1C complex. These observations are discussed with respect to the possible biological functions of hsp56 in the folding and/or cytoplasmic-nuclear trafficking of the receptor.

A soluble binding assay for measuring 3H-FK506 binding to the hsp56 immunophilin.

Heat shock protein 56 (hsp56) was previously identified as an immunophilin based on its ability to specifically bind to FK506-Affi-Gel 10. In this report, we have quantitated human Jurkat T cell hsp56 binding to 3H-FK506, as well as to the immunosuppressant rapamycin. Binding was measured utilizing immunoadsorbed hsp56, and, in addition, we demonstrate that 3H-FK506 binds to hsp56 in solution. Hsp56 bound to an antibody-Sepharose column binds 3H-FK506 with an affinity of 19.4 +/- 4.6 nM, as compared to 23.2 +/- 6.8 nM for soluble hsp56. In competition experiments, the apparent affinity constant for rapamycin was 11.6 +/- 2.8 nM, using immobilized hsp56, and 17.3 +/- 7.7 nM, using the soluble hsp56 preparation. These results demonstrate that hsp56 binds FK506 and rapamycin with similar affinities, and suggest that hsp56 may play a role in mediating the cellular function of both of these drugs.

The native v-Raf.hsp90.p50 heterocomplex contains a novel immunophilin of the FK506 binding class.

We have recently reported that v-Raf forms a native heterocomplex with rat heat shock protein (hsp) 90 and a 50-kDa phosphoprotein (p50) in stably transfected 3Y1 fibroblasts (Stancato, L. F., Chow, Y-H., Hutchison, K. A., Perdew, G. H., Jove, R., and Pratt, W. B. (1993) J. Biol. Chem. 268, 21711-21716). Several members of the nuclear receptor family exist in heterocomplexes containing hsp90 and various members of the immunophilin protein family, including hsp56, an immunophilin of the FK506 binding class (Pratt, W. B. (1993) J. Biol. Chem. 268, 21455-21458). In this work, we have asked if Raf is also associated with an immunophilin. We have immunoadsorbed v-Raf from stably transfected rat 3Y1 fibroblasts and show that the immunoadsorbed v-Raf.hsp90.p50 heterocomplex binds the immunosuppressant drug [3H]FK506. The binding is of high affinity (KD 82 nM) and specific in the sense that it is competed by nonradioactive FK506 and rapamycin but not by cyclosporin A. The [3H]FK506 binding activity is eliminated when the heterocomplex proteins are dissociated from v-Raf. Using the 22W mutant of c-Raf in which the NH2-terminal half has been deleted, we show that the catalytic domain of the kinase is sufficient for the immunophilin association. We have shown that hsp90 and p50 do not bind FK506, and the v-Raf heterocomplex does not contain any of the established FK506-binding proteins. Thus, we propose that Raf is associated with a novel immunophilin.

Solution structure of human interleukin-1 receptor antagonist protein.

Interleukin-1 receptor antagonist protein (IRAP) is a naturally occurring inhibitor of the interleukin-1 receptor. In contrast to IL-1 beta, IRAP binds to the IL-1 receptor but does not elicit a physiological response. We have determined the solution structure of IRAP using NMR spectroscopy. While the overall topology of the two 153-residue proteins is quite similar, functionally critical differences exist concerning the residues of the linear amino acid sequence that constitute structurally homologous regions in the two proteins. Structurally homologous residues important for IL-1 receptor binding are conserved between IRAP and IL-1 beta. By contrast, structurally homologous residues critical for receptor activation are not conserved between the two proteins.

Initial crystallographic analysis of a recombinant human interleukin-1 receptor antagonist protein.

We report the crystallization of samples of a recombinant preparation of human interleukin-1 receptor antagonist protein (IRAP) and solution of the crystal structure by isomorphous replacement methods. Crystals were obtained by the hanging-drop vapor-diffusion method at 277 K from solutions of PEG 4000 containing sodium chloride, dithiothreitol and PIPES [sodium piperazione-N,N'-bis(2-ethanesulfonate)] buffer at pH 7.0. Crystals appear within about a week and grow as truncated tetragonal bipyramids to 0.3-0.6 mm on an edge. X-ray diffraction data from these crystals specify space group P4(3)2(1)2 and unit-cell dimensions of a = b = 72.35(26), c = 114.7(8) A and Z = 16 (two molecules per asymmetric unit). Fresh crystals diffract to about 2.3 A resolution. The search for heavy-atom derivatives has produced two, potassium gold cyanide and trimethyl lead chloride, as same-site, single-site derivatives. Inspection of an electron-density map at 4 A resolution calculated with these derivatives confirms that the IRAP molecule is a member of the interleukin-1 structural family.

An active FK506-binding domain of 17,000 daltons is isolated following limited proteolysis of chicken thymus hsp56.

We have previously identified hsp56, a protein component of steroid receptor complexes, as an FK506 binding protein [Yem et al. (1992) J. Biol. Chem. 267, 2868-2871]. We now report that hsp56 is also found to be a major immunophilin in chicken thymus, by virtue of binding to FK506-Affi-Gel-10 as well as positive cross-reactivity with a polyclonal antiserum directed against human hsp56. Limited digests of purified chicken hsp56 with endoproteinase Lys C result in the production of a unique polypeptide having a mass of about 17 kDa (p17), as judged by Western blotting. Peptide mapping provided additional proof that p17 is a fragment which comprises the entire FK506 binding domain I of chicken hsp56, terminating with an Arg-Lys which might represent a processing site. Binding of radiolabeled dihydro FK506 to p17 is saturable with a calculated KD of 42 nM. Since size exclusion chromatography of drug-p17 complexes indicates that the active species is a homodimer with a mass of 30-40 kDa, the stoichiometry calculated for the drug-protein complex is approximately 1:1. Furthermore, unlike FKBP-12, chicken p17 bound to FK506 does not bind to calcineurin-calmodulin complexes. This work demonstrates the excision of a domain from an hsp56 protein that is active in binding FK506 and functionally distinct from FKBP-12, a protein of similar size and structure.

Chemical modification of interleukin-1 beta: biochemical characterization of a carbodiimide-catalyzed intramolecular cross-linked protein.

We have modified recombinant human Interleukin-1 beta using 1-ethyl-3(3-dimethylaminopropyl)-carbodiimide at pH 6.5, resulting in the formation of an internally cross-linked protein. The major product (30% yield) of the reaction (17 kD; pI = 6.2) was purified and fully characterized by peptide mapping using Endoproteinase Lys C. When digests were conducted under nondenaturing conditions, we found that the modified protein is different from the native protein. The native protein yielded 14 peptides after digestion, whereas only two large peptides and a tetrapeptide, Asn-Tyr-Pro-Lys, were released from the cross-linked protein (i.e., cleavage occurs only at residues Lys88 and Lys92). Using gel filtration, the two peptides were found to co-elute as a single species (15 kD), which represent a noncovalent complex of the amino terminal and C-terminal portions of the molecule. Further analysis of the modified protein by peptide mapping under denaturing conditions and by FAB MS analysis showed that Glu111 and Lys138 were internally cross-linked. The cross-linked protein had bioactivity (T-cell proliferation), fluorescence, and circular dichroism spectra similar to native IL-1 beta. In contrast, while having similar secondary structure, the digested cross-linked protein had less than 1% of T-cell proliferative activity of the undigested protein. These data show that the structural integrity surrounding and perhaps including the Asn-Tyr-Pro-Lys region may be crucial for the biological activity of rIL-1 beta and may be important for the binding of IL-1 to its receptor.

Proton, carbon, and nitrogen chemical shifts accurately delineate differences and similarities in secondary structure between the homologous proteins IRAP and IL-1 beta.

1H alpha, 13C alpha, and 15N alpha secondary shifts, defined as the difference between the observed value and the random coil value, have been calculated for interleukin-1 receptor antagonist protein and interleukin-1 beta. Averaging of the secondary chemical shifts with those of adjacent residues was used to smooth out local effects and to obtain a correlation dependent on secondary structure. Differences and similarities in the placement of secondary structure elements in the primary sequences of these structurally homologous proteins are manifested in the smoothed secondary chemical shifts of all three types of nuclei. The close correlation observed between the secondary chemical shifts and the previously defined locations of secondary structure, as defined by traditional methods, exemplifies the advantage of chemical shifts to delineate regions of secondary structure.

Tissue distribution and cellular localization of hsp56, an FK506-binding protein. Characterization using a highly specific polyclonal antibody.

Heat shock protein 56 (hsp56) has been shown to be involved in two cellular pathways, as an immunophilin for FK506 and as a component of steroid receptor complexes. To help define its role in these cellular pathways, we have developed UPJ56, a polyclonal antibody raised against hsp56 purified from Jurkat cells. In Western blot experiments, hsp56 was highly expressed in rat thymus, liver, and spleen, with low levels in lung and muscle. In immunofluorescence experiments using untreated LLC-PK1 cells, fibrillar staining was seen in the cytoplasm, suggesting a cytoskeletal localization of hsp56. The nuclei were brightly stained, except for the nucleoli. Confocal microscopy demonstrated that the staining was present in all planes of the nucleus. These results suggest that hsp56 is expressed in tissues enriched in steroid receptors and is highly expressed in tissues involved in T cell function. Furthermore, the localization of hsp56 with the cytoskeleton and throughout the nucleus is consistent with its association with steroid receptor complexes.

Secondary structure and topology of interleukin-1 receptor antagonist protein determined by heteronuclear three-dimensional NMR spectroscopy.

Interleukin-1 (IL-1) proteins, such as IL-1 beta, play a key role in immune and inflammatory responses. Interaction of these cytokines with the IL-1 receptor induces a variety of biological changes in neurologic, metabolic, hematologic, and endocrinologic systems. Interleukin-1 receptor antagonist protein (IRAP) is a naturally occurring inhibitor of the interleukin-1 receptor. The 153-residue protein binds to the receptor with an affinity similar to that of IL-1 beta but does not elicit any physiological responses. As a first step toward understanding IRAP's mode of action, we have used multidimensional, heteronuclear NMR spectroscopy to determine the antagonist's solution secondary structure and global fold. Using a combination of 3D 1H-15N NOESY-HMQC and TOCSY-HMQC and 3D 1H-15N-13C HNCA and HN(CO)CA experiments on uniformly 15N- or doubly 13C/15N-enriched IRAP, we have made resonance assignments for more than 90% of the main-chain atoms. Analysis of short- and long-range NOE's indicates that IRAP is predominantly beta-sheet, with the same overall topology as IL-1 beta but with different regions of the primary sequence comprising the beta-strands. Two short helical segments also were identified. The 14% sequence identity between IL-1 beta and IRAP increases to 25% when differences in the locations of secondary structure elements in the primary sequences are taken into account. Still, numerous differences in side chains, which ultimately play a major role in receptor interaction, exist.(ABSTRACT TRUNCATED AT 250 WORDS)

FKBP-12 is not an inhibitor of protein kinase C.

It was recently noted that the amino acid sequence of FK506 binding protein (FKBP-12) is nearly identical to that of an endogenous inhibitor of protein kinase C, PKCI-2. To follow up on this observation, we have tested the hypothesis that FKBP-12 is an inhibitor of PKC. The kinase activity of rat brain protein kinase C (PKC) was not inhibited by the presence of up to 700 micrograms recombinant human FKBP-12 per ml, in either the presence or absence of FK506. FKBP-12 also did not affect PMA-induced phosphorylation of an endogenous PKC substrate, an 80 kDa protein, in permeabilized cells. To test whether FKBP-12 could account for endogenous PKC inhibitory activity in cells, Jurkat cell lysate was chromatographed on an anion exchange column. A peak of PKC inhibitory activity was eluted at approximately 200 mM NaCl. As shown by both Western blots and FK506 binding activity, FKBP-12 was eluted only in the flow-through and wash fractions. These results demonstrate that FKBP-12 is clearly distinct from endogenous PKC inhibitory activity.

Site-specific chemical modification of interleukin-1 beta by acrylodan at cysteine 8 and lysine 103.

Acrylodan, which normally modifies cysteine residues, was employed to derivatize recombinant interleukin-1 beta (rIL-1 beta) under native conditions, using a reagent:protein ratio of 3:1. Two major covalent protein/acrylodan adducts were generated and subsequently purified by DEAE TSK 5PW ion exchange chromatography. Peptide mapping and mass spectrometry were used to locate the probe on the modified proteins. Both modified proteins carried one molecule of acrylodan each, one at Cys-8 and the other at Lys-103. Neither Cys-71 nor any of the other 13 lysine residues of rIL-1 beta was modified. Cysteine 71 is inaccessible to acrylodan, but the unusual specificity for Lys-103 could be caused by the location of that residue at the bottom of a hydrophobic pocket which might specifically bind the reagent. No double-labeled protein was detected, indicating that the introduction of the label at either site interferes with the labeling at the other. Both acrylodan-modified proteins exhibited bioactivity in the thymocyte proliferation assay at a level equivalent to that of the unmodified control protein (1.7 x 10(7) units/mg), which shows that the modification of either the Cys-8 or Lys-103 position with acrylodan does not interfere with the cellular bioactivities of the respective proteins. Furthermore, receptor binding assays yielded a Kd = 32.0 +/- 4.8 pM for the Lys-103-labeled protein, Kd = 69.5 +/- 12.7 pM for the unmodified protein, and Kd = 75.0 +/- 11.6 pM for the Cys-8-labeled protein. Thus, Cys-8 or Lys-103 modification of rIL-1 beta by acrylodan also does not interfere with the ability of the molecule to bind to its receptor. The slightly higher affinity of the Lys-103-labeled protein for the receptor suggests that the positive charge on this residue in the native molecule may interfere with IL-1 receptor binding. The two fluorescent labeled IL-1 proteins described herein should provide interesting probes for the study of IL-1/IL-1 receptor interactions.

Spectral characterization of environment-sensitive adducts of interleukin-1 beta.

We have determined the fluorescence properties of two covalently attached acrylodan derivatives of recombinant human interleukin-1 beta, namely the Cys-8 and Lys-103 adducts. The emission and excitation maxima indicated the presence of two operationally distinct conformers for each probe. The iodide quenching and the kinetics of fluorescence changes associated with guanidinium chloride-induced denaturation show that each covalent adduct exists both in hydrated and dehydrated environments. Furthermore, fluorescence changes associated with the binding of the adducts to a recombinant soluble murine receptor indicated that only the conformers with the label in the hydrophobic environment are competent toward the soluble murine interleukin receptor and that the hydrated and dehydrated conformers are in a dynamic equilibrium on the time scale of the binding experiments.

The Hsp56 component of steroid receptor complexes binds to immobilized FK506 and shows homology to FKBP-12 and FKBP-13.

Extracts from human Jurkat cells or from calf thymus contain a 60-kDa protein that is bound to immobilized FK506. As expected, the NH2-terminal sequences of the 60-kDa protein from these two species were found to be nearly the same. We were surprised to discover, however, that the sequence of the human protein was identical to that of Hsp56, a heat shock protein of unknown function that has been shown to be a component of several steroid receptor complexes. Further analysis of the calf thymus protein revealed a peptide with homology to a region near the COOH terminus of both FKBP-12 and FKBP-13. It would appear, therefore, that this 60-kDa protein, or as we refer to it provisionally, "Hsp56," could have the capacity to bind FK506 directly. These observations lead us to speculate that "Hsp56" may mediate immunosuppression and inhibition of T-cell proliferation by FK506 and may do so via a cytosolic signal transduction pathway separate, but not necessarily exclusive, from that of FKBP-12 and FKBP-13.