gamma-L-Glutaminyl-4-hydroxybenzene, an inducer of cryptobiosis in Agaricus bisporus and a source of specific metabolic inhibitors for melanogenic cells.

A stable phenol, gamma-L-glutaminyl-4-hydroxybenzene (GHB), is oxidized by tyrosinase in the gill tissues of the mushroom Agaricus bisporus to a quinone and a second oxidation product which together suppress mitochondrial energy production and the synthesis of proteins and nucleic acids in the zygote, thus establishing dormancy in the spores. Brief incubation of cultured murine L1210 leukemia and B-16 melanoma cells with muM concentrations of the purified quinone notably prolonged survival times or blocked tumor growth in histocompatible mice inoculated i.p. with high concentrations of the exposed cells. The instability of the quinone precluded in vivo administration. The short incubation of cultured B-16 melanoma cells with mM concentrations of GHB markedly prolonged survival times or abolished tumor growth in histocompatible C57BL/6J mice inoculated i.p. with 5 X 10(6) exposed cells. This response did not occur with L1210 leukemia cells, which lack the enzyme tyrosinase. The survival times of mice bearing B-16 melanoma, but not of those with L1210 leukemia, were slightly prolonged by a single injection and were significantly extended by daily i.p. injections of GHB. Normal C57BL/6J mice, given GHB i.p. as single or multiple 400-mg/kg doses, manifested no systemic toxicity but showed depigmentation of the hair after 2 to 3 weeks. These studies provide evidence that GHB exerts cytotoxicity specifically for cells that by their content of tyrosinase convert the phenol to the quinone. This targeted response minimizes systemic toxocity and underscores the potential therapeutic application of this agent to melanocarcinoma.

Synthesis of iodine-125 labeled (+/-)-15-(4-azidobenzyl)carazolol: a potent beta-adrenergic photoaffinity probe.

(+/-)-15-(4-Azidobenzyl)carazolol (2), a potent beta-adrenergic photoaffinity ligand developed in our laboratories, has been radioiodinated to theoretical specific activity (2175 Ci/mmol) and shown to label covalently beta-adrenergic receptor peptides in avian and amphibian erythrocyte membrane preparations. The radioiodinated analogues of the desired compound (2) were optimally prepared by two synthetic steps from (+/-)-15-(4-aminobenzyl)carazolol (8). The latter was iodinated with carrier-free Na125I and chloramine T to yield two major isotopomers (the monoiodinated derivatives 9 and 10), which were separated by thin-layer chromatography and converted via diazonium salt formation to their respective 4-azides, 12 and 6. These azides can be used interchangeably in ligand binding or photoaffinity labeling experiments. Compound 8 was obtained by catalytic reduction of the nitro derivative (7), which was arrived at by direct reaction of 1,1-dimethyl-2-(4-nitrophenyl)ethylamine (3) with 4-(2,3-epoxypropoxy)carbazole (5). Of the desired isomers, (+/-)-15-(4-azido-3-iodobenzyl)carazolol (6) could be synthesized from 1,1-dimethyl-2-(4-azido-3-iodophenyl)ethylamine (4) by direct reaction with 5. This and the preceding sequence of reactions were carried out by using nonradioactive materials, and separation and purification of products were accomplished by high-performance liquid chromatography. The compounds described have been shown to be potent beta-adrenergic antagonists by virtue of their ability to inhibit beta-adrenergic stimulation of adenylate cyclase or to compete for the binding of another beta-adrenergic ligand, [125I]cyanopindolol, to the beta-adrenergic receptors of frog erythrocytes. The photoactive azide derivatives of these compounds (6 and 12) have been shown to covalently incorporate into the beta-adrenergic receptor binding subunit of frog and turkey erythrocyte membrane preparations. Incorporation of the ligands into these polypeptides can be blocked specifically by both beta-adrenergic agonists and antagonists.

Actinoidins A and A2: structure determination using 2D NMR methods.

The structural analysis of the intact glycopeptide antibiotics, actinoidins A (1a) and A2 (1b), by two-dimensional 1H NMR is described. The location of the single chlorine at the A3 position and the sites of attachment of the four carbohydrate substituents in actinoidin A are elucidated based on correlation spectroscopy (COSY), double quantum coherence experiments (DQCE), homonuclear Hartmann-Hahn experiments (HOHAHA) and nuclear Overhauser spectroscopy (NOESY). Similar 2D correlation and NOE NMR experiments are then performed on the novel analog, actinoidin A2, to determine its structure. The structural difference between actinoidins A and A2 is shown to reside in the presence of L-rhamnose in actinoidin A2 in place of L-acosamine in actinoidin A. All questions concerning the stereo-chemistry of the chiral centers in both the heptapeptide core and the carbohydrate moieties in each of these antibiotics could be successfully addressed with the exception of Gl', the alpha-carbon on the N-terminal amino acid which is known to have the R-configuration from previous studies.

Structural characterization of glycopeptide antibiotics related to vancomycin by fast atom bombardment mass spectrometry.

A series of glycopeptide antibiotics related to the vancomycin-ristocetin family have been successfully analyzed by fast atom bombardment mass spectrometry (FABMS). The FAB mass spectra of glycopeptides weighing up to 2,100 daltons exhibit intense molecular ions and fragment ions from which information concerning carbohydrate composition and sequence are readily obtained. Careful adjustment of the FABMS experimental conditions has enabled the accurate masses of the glycopeptides to be determined by high resolution FABMS with an accuracy of better than six ppm. Comparison of the observed molecular ion cluster pattern with calculated isotope distributions reveals the precise number of chlorine atoms in these molecules, which, together with the accurate mass data, can be used to restrict the number of possible elemental compositions to a meaningfully small value. These techniques have been used to characterize several glycopeptides of known structure including ristocetin, actinoidin, avoparcin, vancomycin and A35512B, as well as aridicins A, B and C which are three new, novel members of the vancomycin class.

Biosynthetic studies of aridicin antibiotics. I. Labeling patterns and overall pathways.

Biosynthetic feeding experiments with 14C and 13C-labeled precursors in Kibdelosporangium aridum have established the biosynthetic origins of the heptapeptide aglycone of the aridicin antibiotics, and the tentative sequence of the later stage biosynthetic transformations. The aglycone moiety has been found to be derived from tyrosine, sodium acetate and L-methionine. It is suggested that the preformed aglycone is first mannosylated and then followed by the attachment of the glycolipid. The sugar oxidation to the glucuronic acid level was found to take place as a terminal step of the biosynthesis.

Aridicins, novel glycopeptide antibiotics. III. Preparation, characterization, and biological activities of aglycone derivatives.

The aglycone and two pseudoaglycones of aridicin A were prepared by selective hydrolysis and characterized, chemically and biologically. These new analogs demonstrate improved activities in vitro over the parent antibiotics against methicillin sensitive and resistant staphylococci. The major determinant of activity is the mannose substituent, the presence of which results in less potent compounds. The analogs have potent activity against enterococci.

Kibdelins (AAD-609), novel glycopeptide antibiotics. II. Isolation, purification and structure.

A new glycopeptide antibiotic complex was isolated from the fermentation culture of Kibdelosporangium aridum subsp. largum (SK&F AAD-609) by affinity chromatography on a D-alanyl-D-alanine agarose column. This major components of the complex were resolved by preparative reversed-phase HPLC. Mild acid hydrolysis showed that the new antibiotics have the same mannosyl aglycon (2) as the aridicins. FAB mass spectrometry, isoelectric focusing, potentiometric titration and carbohydrate and fatty acid analyses were used to determine the structures of the five major components of the complex. These studies showed that the kibdelins differ from the aridicins only in the oxidation level at the C-6 position of the amino sugar. Kibdelin A (5), B (6), C1 (7), C2 (8) and D (9) are a series of N-acylglucosamine analogs containing saturated straight and branched chain C10-C12 fatty acids whereas, in kibdelin D the fatty acid component is (Z)-4-decenoic acid.

Unusual conformation of a 3'-thioformacetal linkage in a DNA duplex.

The DNA.DNA duplex d(CGCGTTSCH2OTTGCGC).d(GCGCAAAACGCG) (designated duplex III) containing a 3'-thioformacetal (3'-TFMA) linkage in the center of the sequence was characterized in detail by two- and three-dimensional homonuclear NMR spectroscopy. The NMR results were analyzed and compared with those of two duplexes of the same sequence: One is an unmodified reference sequence and the other contains a formacetal (OCH2O) linkage at the central T--T step (designated duplex I and duplex II, respectively). In general, the NMR spectra of duplex III closely resemble those of the analogous duplexes I and II, suggesting an overall B-type structure adopted by the 3'-TFMA-modified duplex III. Nonetheless, the detection of several distinct spectral features originating from the protons at the T6(3'-SCH2O)T7 modification site is indicative of a local conformation that is clearly different from the corresponding region in duplexes I and II. The 3'-thioformacetal linker, in contrast to the formacetal (FMA) linkage, cannot be accommodated in a conformation usually found in natural nucleic acid duplexes. As a consequence, the 3'-TFMA-modified T6 sugar adopts an O4'-endo form (an intermediate structure between the usual C2'-endo and C3'-endo forms). This change is accompanied by a change in the epsilon (C4'-C3'-S3'-CH2) dihedral angle and by subsequent adjustments of other torsion angles along the backbone. Notably, this conformational readjustment at the T6-T7 backbone linkage is localized; its collective result has negligible effect on base-base stacking of the T6 and T7 residues. A close examination of the COSY data in all three duplexes reveals a subtle variation in sugar geometry, with more S-type character adopted by the modified duplexes II and III. The results of this study illustrate that, although the difference between FMA and 3'-TFMA linkages is merely in the substitution of the T6(O3') in the former by a sulfur atom in the latter, the stereoelectronic difference in a single atom can induce significant local structural distortion in an otherwise well-structured oligonucleotide duplex.

Sequence-dependent conformational heterogeneity of a hybrid DNA.RNA dodecamer duplex.

Two- and three-dimensional homonuclear NMR studies of a hybrid duplex RI, formed by annealing r(GCGCAAAACGCG) and d(CGCGTTTTGCGC) strands are described. NMR parameters, such as intra- and interresidue proton-proton NOEs and sugar proton coupling constants were analyzed with reference to those of the corresponding DNA.DNA duplex. Furthermore, spectral analyses were conducted on the basis of model structures of nucleic acid duplexes. Distinctive spectral patterns of the hybrid duplex reveal unique heterogeneous conformations which co-exist throughout the sequence and are significantly different from those of model structures of either canonical A- or B-forms. Features of an intermediate conformation were observed in the DNA and RNA strands in duplex RI, the former being more B-like and the latter more A-like. Three-dimensional NOESY-NOESY spectra were analyzed and their use was demonstrated for resolving superimposed resonances and cross peaks, especially those originating from the RNA strand. The application of a useful strategy that combines the use of 2D NMR data and the known structural information for efficient 3D spectral analyses is demonstrated.

The role of 2,4,5-trihydroxyphenylalanine in melanin biosynthesis.

Both 3,4-dihydroxyphenylalanine and 2,4,5-trihydroxyphenylalanine were oxidized with periodate and mushroom tyrosinase to determine whether the latter compound is an intermediate in melanin biosynthesis. Matrix analysis of the spectra obtained with a rapid scan spectrophotometer and comparison of the spectra of quinone intermediates with model quinones disclosed that, although 2,4,5-trihydroxyphenylalanine can be oxidized to 2-carboxy-2,3-dihydroindole-5,6-quinone (dopachrome), this oxidation proceeds through a stable intermediate, 5-(2-carboxy-2-aminoethyl)-2-hydroxy-1,4-benzoquinone, which does not appear in the oxidation of 3,4-dihydroxyphenylalanine to dopachrome. Thus, these studies are in agreement with the original postulate, that 4-(2-carboxy-2-aminoethyl)-1,2-benzoquinone and leukodopachrome are the intermediates in the major pathway for dopachrome synthesis.

Solution conformation of human big endothelin-1.

The solution conformation of human big endothelin-1, a 38-residue peptide which serves as the putative precursor to the potent vasoconstrictor endothelin-1 has been examined by 1H NMR. NOEs were utilized as distance restraints in the distance geometry program DSPACE to generate initial structures. Further refinement of these structures was accomplished through molecular mechanics/molecular dynamics in an iterative process involving the incorporation of stereospecific assignments of prochiral centers and the use of back-calculation of NOESY spectra. A family of structures consisting of a type II beta-turn for residues 5-8 and an alpha-helix extending from residues 9-16 constitute a well-defined region, as reflected by the atomic root-mean-square (RMS) difference of 1.56 A about the mean coordinate positions of the backbone atoms (N, C, C alpha and O). This core region (residues 1-15) is very similar to the core residues of endothelin-1 (Donlan, M. et al. (1991) J. Cell. Biochemistry, S15G, 85). While the evidence from NOESY and coupling constant data suggests that the C-terminal region, residues 17-34, is not a mixture of randomly distributed chain conformations, it is also not consistent with a single chain conformation. Under the conditions studied, residues 17-38 in human big endothelin-1 in water at pH 3.0 between 20-30 degrees C appear to be represented by a series of conformers in dynamic equilibrium.

1H NMR assignment and secondary structural elements of human transforming growth factor alpha.

The 1H NMR spectrum of human transforming growth factor alpha (hTGF-alpha) has been completely assigned, and secondary structural elements have been identified as a preliminary step in determining the structure of this protein by distance geometry methods. Many of these structural elements closely correspond to those previously found in a truncated human EGF [Cooke et al. (1987) Nature (London) 327, 339-341] and murine EGF [Montelione et al. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 5226-5230]. These include the presence of an antiparallel beta-sheet between residues G19 and C34 with a type I beta-turn at V25-D28, a type II beta-turn at H35-Y38, and another short beta-sheet between residues Y38-V39 and H45-A46.

NMR restraint analysis of transforming growth factor alpha: a key component for NMR structure refinement.

Structures of the protein, transforming growth factor alpha (TGF-alpha), have been derived from NMR data using distance geometry and subsequent energy refinement. Analysis of the sequential NOE distance bounds using a template algorithm provides a check for consistency in the calculation of bounds, stereospecific assignment of prochiral centers, and secondary structure assignment. Application of the template algorithm to the long range NOEs found within the NMR data sets collected at pH 6.3 and pH 3.4 is used to assess the confidence levels for the accuracy of the structures obtained from modeling. The method also provides critical insight in differentiating regions of the structure that are well defined from those that are not. Use of the restraint analysis protocol is shown to be a powerful adjunct to currently used methods for the assignment of protein structures from NMR data.

Phenylalanine in the second membrane-spanning domain of alpha 1A-adrenergic receptor determines subtype selectivity of dihydropyridine antagonists.

The alpha 1-adrenergic receptors (alpha 1-AR) belong to the G-protein coupled seven-transmembrane biogenic amine receptor family. Three subtypes have been successfully cloned in the alpha 1-adrenergic receptor family, and they share 50% identical amino acid sequences and 70% similarity. We have constructed seven chimeric receptors of the alpha 1A-AR. Each of the chimeras contains alpha 1D-subtype amino acid sequences within the membrane-spanning domains. Comparisons of ligand affinities with these chimeras has provided information on the importance of certain amino acid residues in determining receptor subtype specificity in the alpha 1A- and alpha 1D-ARs. With ligands in the dihydropyridine series, the niguldipine analog 1 was found to have respective pKi's of 9.32 +/- 0.17 for alpha 1A-AR; 6.84 +/- 0.24 for alpha 1D-AR; and 6.76 +/- 0.28 for alpha 1A/D(TM2), respectively. This trend was also exhibited by two other niguldipine analogs, 2 and 3, which had similar pKi's toward alpha 1D-AR and alpha 1A/D(TM2). This subtype selectivity was also maintained in the piperdine derivative, 4, and alpha 1A-AR selective ligand, which showed the same parallel trends in binding affinities with alpha 1A-AR and the six chimeras as the niguldipine analogs. Since in considering the second membrane-spanning domain, the alpha 1A- and alpha 1D-ARs only differ at positions 76, 77, 85, and 86, we were able to show through mutational studies that phenylalanine 86 is solely responsible for the selectivity found in the chimeric receptor alpha 1A/D(TM2) exhibited against the ligands 1-4 used in this study. A model based on the rhodopsin structure places the amino acid at position 86 in the final turn toward the extracellular region. This is four helical turns above aspartic acid-79, a conserved amino acid in the second membrane-spanning domain. This is the first report that suggests a significant involvement of the second membrane-spanning domain in antagonist binding in the biogenic amines class of the superfamily of seven-transmembrane receptors.