Analysis of the ligand binding properties of recombinant bovine liver-type fatty acid binding protein.

The coding part of the cDNA for bovine liver-type fatty acid binding protein (L-FABP) has been amplified by RT-PCR, cloned and used for the construction of an Escherichia coli (E. coli) expression system. The recombinant protein made up to 25% of the soluble E. coli proteins and could be isolated by a simple two step protocol combining ion exchange chromatography and gel filtration. Dissociation constants for binding of oleic acid, arachidonic acid, oleoyl-CoA, lysophosphatidic acid and the peroxisomal proliferator bezafibrate to L-FABP have been determined by titration calorimetry. All ligands were bound in a 2:1 stoichiometry, the dissociation constants for the first ligand bound were all in the micro molar range. Oleic acid was bound with the highest affinity and a Kd of 0.26 microM. Furthermore, binding of cholesterol to L-FABP was investigated with the Lipidex assay, a liposome binding assay and a fluorescence displacement assay. In none of the assays binding of cholesterol to L-FABP was observed.

Three-dimensional structure of bovine heart fatty-acid-binding protein with bound palmitic acid, determined by multidimensional NMR spectroscopy.

The three-dimensional structure of the holo form of recombinant cellular bovine heart fatty-acid-binding protein (H-FABPc), a polypeptide of 133 amino acid residues with a molecular mass of 15 kDa, has been determined by multidimensional homonuclear and heteronuclear NMR spectroscopy applied to uniformly 15N-labeled and unlabeled protein. A nearly complete set of 1H and 15N chemical shift assignments was obtained. A total of 2329 intramolecular distance constraints and 42 side-chain chi 1 dihedral-angle constraints were derived from cross-relaxation and J coupling information. 3D nuclear Overhauser enhancement and exchange spectroscopy combined with heteronuclear multiple-quantum coherence (NOESY-HMQC) experiments, performed on a sample of uniformly 13C-labeled palmitic acid bound to unlabeled cellular heart fatty-acid-binding protein revealed 10 intermolecular contacts that determine the orientation of the bound fatty acid. An ensemble of protein conformations was calculated with the distance-geometry algorithm for NMR applications (DIANA) using the redundant dihedral-angle constraint (REDAC) strategy. After docking the fatty acid into the protein, the protein-ligand arrangement was subject to distance-restrained energy minimization. The overall conformation of the protein is a beta-barrel consisting of 10 antiparallel beta-strands which form two nearly orthogonal beta-sheets of five strands each. Two short helices form a helix-turn-helix motif in the N-terminal region of the polypeptide chain. The palmitic acid is bound within the protein in a U-shaped conformation close to the two helices. The obtained solution structure of the protein is consistent with a number of fatty-acid-binding-protein crystal structures.

Members of the fatty acid binding protein family are differentiation factors for the mammary gland.

Mammary gland development is controlled by systemic hormones and by growth factors that might complement or mediate hormonal action. Peptides that locally signal growth cessation and stimulate differentiation of the developing epithelium have not been described. Here, we report that recombinant and wild-type forms of mammary-derived growth inhibitor (MDGI) and heart-fatty acid binding protein (FABP), which belong to the FABP family, specifically inhibit growth of normal mouse mammary epithelial cells (MEC), while growth of stromal cells is not suppressed. In mammary gland organ culture, inhibition of ductal growth is associated with the appearance of bulbous alveolar end buds and formation of fully developed lobuloalveolar structures. In parallel, MDGI stimulates its own expression and promotes milk protein synthesis. Selective inhibition of endogenous MDGI expression in MEC by antisense phosphorothioate oligonucleotides suppresses appearance of alveolar end buds and lowers the beta-casein level in organ cultures. Furthermore, MDGI suppresses the mitogenic effects of epidermal growth factor, and epidermal growth factor antagonizes the activities of MDGI. Finally, the regulatory properties of MDGI can be fully mimicked by an 11-amino acid sequence, represented in the COOH terminus of MDGI and a subfamily of structurally related FABPs. This peptide does not bind fatty acids. To our knowledge, this is the first report about a growth inhibitor promoting mammary gland differentiation.

Matrix assisted laser desorption/ionization mass spectrometry of enzymatically synthesized RNA up to 150 kDa.

Enzymatically synthesized RNA samples (in vitro transcripts) were analysed by matrix assisted laser desorption/ionization mass spectrometry (MALDI-MS). Spectra of RNA up to 150 kDA (461 nucleotides) are shown. Polymerase generated sample heterogeneity and its contribution to mass resolution are discussed. A time course exonuclease digest of a 55 nt in vitro transcript was analyzed to investigate the performance of MALDI-MS on complex mixtures. Based on these data, the analysis by MALDI-MS of DNA sequencing reactions, produced by the action of an RNA polymerase, is discussed.

Comparison of IR- and UV-matrix-assisted laser desorption/ionization mass spectrometry of oligodeoxynucleotides.

UV-matrix assisted laser desorption/ionization mass spectrometry (UV-MALDI-MS) with 3-hydroxypicolinic acid as matrix and IR-MALDI-MS with succinic acid as matrix have proved their feasibility for highly accurate and sensitive mass determination of nucleic acids (DNA and RNA). In this work, a detailed comparison of these two MALDI-methods and between positive- and negative ion mass spectra for the analysis of oligodeoxynucleotides is undertaken. Mass spectra of DNA sequences with up to 40 nucleotides are shown. Both linear and reflectron time-of-flight mass analyzers were used within this study and are compared for their potential in the MALDI analysis of oligodeoxynucleotides. The role of molecule-ion fragmentation is also discussed.

N-terminal variants of fatty acid-binding protein from bovine heart overexpressed in Escherichia coli.

An expression vector for bovine heart fatty acid-binding protein (H-FABP) was constructed by introducing the coding part of the cDNA into the pET-3d vector. Transformed Escherichia coli strain BL21 (DE3)pLysS produced functional recombinant H-FABP up to 40% of the soluble proteins. The expression of fatty acid-binding protein was under the control of the T7-phi 10 promoter and the corresponding T7-RNA-polymerase in turn was induced by isopropyl beta-D-thiogalactopyranoside. By combination of cation exchange chromatography and gel filtration pure recombinant protein was obtained exhibiting isoelectric heterogeneity. Recombinant H-FABP was resolved into at least six variants with isoelectric points between 5.1 and 5.6. After separation by preparative isoelectric focusing the four major variants were digested with trypsin and the resulting peptides were characterized by high performance liquid chromatography (HPLC), matrix assisted laser desorption/ionization (MALDI) mass spectrometry, amino acid sequencing and chemical modification. The structural differences were traced back to the N-termini beginning with either methionine, as expected from the cDNA, or methionine sulfoxide, valine and N-formyl methionine. The latter three arise from oxidation, cleavage of N-terminal methionine and incomplete deformylation, respectively.

Isoforms of fatty-acid-binding protein in bovine heart are coded by distinct mRNA.

In the course of our studies on structure/function relationships of fatty-acid-binding proteins, we reported earlier that the two isoforms of the 15-kDa cardiac fatty-acid-binding protein (cFABP) from bovine heart only differ in one position; Asn98 in pI 5.1-cFABP; Asp98 in pI 4.9-cFABP [Unterberg, C., Börchers, T., Højrup, P., Knudsen, J. and Spener, F. (1990) J. Biol. Chem. 265, 16,255-16,261]. In the present study, we elucidate the origin for this heterogeneity. Isoelectric focusing analysis of immunoprecipitated in vitro translation products from total mRNA and positive-hybrid-selected cFABP/mRNA revealed two L-[35S]methionine-labeled proteins corresponding to pI 5.1-cFABP and pI 4.9-cFABP. In a control experiment, recombinant mRNA derived from cDNA encoding pI 5.1-cFABP was translated and produced only pI 5.1-cFABP as shown by isoelectric focusing of the translation products. We could observe co-translational acetylation but not post-translational deamidation of the cFABP isoforms. Taken together, our results demonstrate that the isoforms of cardiac fatty-acid-binding protein found in bovine heart are coded by distinct mRNA species.

Solution structure of bovine heart fatty acid-binding protein (H-FABPc).

Fatty acid-binding protein (FABP) from bovine heart, a 15 kDa cytoplasmic protein has been investigated by multi-dimensional homonuclear and heteronuclear NMR-spectroscopy. Perdeuterated palmitic acid has been used as fatty acid ligand. The tertiary structure has been determined from distance geometry calculations with the variable target functions algorithm (DIANA) utilizing 1027 interproton distance constraints, which were obtained from 1H-homonuclear NOESY spectra. Overlapping NOE crosspeaks were assigned by heteronuclear multidimensional NMR-experiments with a 15N-labelled sample. The tertiary structure resembles a beta-barrel (beta-clam) consisting of ten anti-parallel beta-strands and a short helix-turn-helix motif. The beta-strands are arranged in two nearly orthogonal beta-sheets composed of 5 strands each. The solution structure is compared with the x-ray crystal structure of bovine heart and rat intestinal FABPs.

Expression of fatty acid-binding protein from bovine heart in Escherichia coli.

The coding part of the cDNA of cardiac fatty acid-binding protein (cFABP) from bovine heart was cloned into the vector pKK233-2. After induction with isopropyl-beta-D-thiogalactopyranoside cFABP was found in a soluble form in the cytosol of plasmid transformed E. coli amounting up to 5.7% of the soluble protein. cFABP was detected after SDS-polyacrylamide gelelectrophoresis and/or isoelectric focusing and Western blot by immuno-staining and was determined quantitatively by a solid phase enzyme-linked immuno sorbent assay. The cFABP produced by bacteria binds oleic acid with high affinity as shown by comigration of protein and ligand in both gelfiltration and isoelectric focusing. cFABP was purified from bacterial lysates to near homogeneity and resolved into four isoproteins.

Expression of a functionally active cardiac fatty acid-binding protein in the yeast, Saccharomyces cerevisiae.

The unicellular eukaryotic microorganism, Saccharomyces cerevisiae, transformed with a plasmid containing a cDNA fragment encoding bovine heart fatty acid-binding protein (H-FABP) under the control of the inducible yeast GAL10 promoter, expressed FABP during growth on galactose. The maximum level of immunoreactive FABP, identical in size to native protein as judged from SDS-polyacrylamide gel electrophoresis, was reached after approximately 16 hours of induction. Analysis of particulate and soluble subcellular fractions showed that FABP was exclusively associated with the cytosol. FABP expressed in yeast cells was functional as was demonstrated by its capacity to bind 14C-oleic acid in an in vitro assay. Growth of the transformants on galactose as the carbon source was significantly retarded at 37 degrees C. Whereas the fatty acid pattern of total lipids was not altered in transformed cells, desaturation of exogenously added 14C-palmitic acid was significantly reduced both at 30 and 37 degrees C. The lowest percentage of radioactively labeled unsaturated fatty acids was found in the phospholipid fraction.

Cloning of a full-length complementary DNA for fatty-acid-binding protein from bovine heart.

A full-length cDNA for bovine heart fatty-acid-binding protein (H-FABP) was cloned from a lambda gt11 cDNA library established from bovine heart muscle. The cDNA sequence shows an open reading frame coding for a protein with 133 amino acids. Colinearity with the amino acid sequences of four tryptic peptides was asserted. H-FABP isolated from bovine heart begins with an N-acetylated valine residue, however, as derived from analysis of the tryptic, amino-terminal-blocked peptide and the molecular mass of the peptide obtained via secondary-ion mass spectrometry. The molecular mass of the total protein is 14673 Da. Bovine H-FABP is 89% homologous to rat H-FABP and 97% homologous to the bovine mammary-derived growth-inhibition factor described recently by Böhmer et al. [J. Biol. Chem. 262, 15137-15143 (1987)]. Significant homologies were also found with bovine myelin protein P2 and murine adipocyte protein p422. Secondary-structure predictions were proposed for these proteins, based on computer analysis, which reveal striking similarities.

On the structure of poly d(A-S4T).

The helical twist of poly d(A-s4T) was determined from the periodicity of the cleavage patterns of the double stranded polydeoxynucleotide adsorbed on calcium phosphate and found to be 14 bp per turn. Both cleavage patterns and 31P NMR spectra indicate a mononucleotide structure rather than an alternating B DNA like poly d(A-T). The failure of nucleosome formation excludes a B type structure. The discrepancy of the mononucleotide structure found in 31P NMR spectra and the dinucleotide structure given by X ray fiber diffraction is explained by an alternating tilt of the planes of the base pairs (base roll) as a consequence of a strong propeller twist. The importance of interstrand stacking interactions of adjacent 4-thiothymidines for the helical stability is discussed.

Base pairing structure in the poly d(G-T) double helix: wobble base pairs.

High resolution nuclear magnetic resonance (NMR) and ethidium bromide binding studies are used to demonstrate that poly d(G-T) forms an ordered double helical structure at low temperatures (below 24 degrees C in 0.3 M NaCl) in which G and T are hydrogen bonded together in a wobble base pair hydrogen bonding scheme as proposed earlier by Lezius and Domin. Alternative hydrogen bonding schemes involving the tautomeric form of either T or G, such as have been proposed to account for mutation rates in DNA synthesis, are eliminated.

An ATPase depending on the presence of single-stranded DNA from mouse myeloma.

An ATPase was purified from mouse myeloma MOPC 70E the activity of which depends on the presence of single-stranded DNA and divalent cations such as Mg2+, Mn2+, Ca2+, Ni2+ or Fe2+. The enzyme splits both ribonucleoside and deoxyribonucleoside triphosphates but preferentially ATP and dATP yielding nucleoside diphosphates and inorganic phosphate. The enzyme has an absolute requirement for single-stranded DNA. Alternating double-stranded polydeoxynucleotides are only slight effective, and native double-stranded DNA, single-stranded and double-stranded RNAs as well as DNA - RNA hybrids are ineffective in stimulating the ATPase. The enzyme has further characterized by sedimentation in a sucrose density gradient (s20, w = 5.5 S) and by isoelectric focussing in an ampholine pH gradient (pI = 6.5).

Rapid isolation of highly active RNA polymerase from Escherichia coli and its subunits by matrix-bound heparin.

1. RNA polymerase from Escherichia coli is selectively and strongly retained by a heparin-substituted agarose and can be eluted therefrom by a neutral buffer containing 0.6 M salt. The method is applicable to relatively crude preparations of the enzyme on a preparative scale giving highly purified RNA polymerase in excellent yield. The enzyme obtained by this procedure shows the highest specific activity so far reported and is pure and enriched in factor sigma as indicated by dodecylsulfate gel electrophoresis. 2. Based on the differential affinity of the subunits of the enzyme for the heparin-carrying gel matrix, a method for separation of alpha, beta' + beta and sigma subunits by application of urea and salt-containing buffers is described. Upon recombination and dialysis with urea-free buffer 40-50% of the enzyme activity is restored.

High-molecular-weight DNA polymerases from mouse myeloma. Purification and properties of three enzymes.

Cytoplasmic (high-molecular-weight) DNA polymerase was partially purified from mouse myeloma. Upon chromatography on DEAE-Sephadex, following fractionation on phosphocellulose, the enzyme was resolved into three species named CI, CII, and CIII. The species CI and CII have equal sedimentation coefficients (10.5 S) in sucrose gradients without salt. In the presence of 125 mM ammonium sulfate the sedimentation coefficients are reduced to 8.6 S. The species CIII shows sedimentation coefficients of 5.7 S and 5.2 S without salt and in the presence of 125 mM ammonium sulfate, respectively. This species is assumed to be an artifact arising from either CI or to a minor extent from CII. The optima for pH, KCl and Mg2+ concentration, and the extent of inhibition by N-ethylmaleimide are the same. However, the enzymes differ in their responses to Mn2+ (substituting for Mg-2+), and to addition of ethanol, dimethylsulfoxide, and various phospholipids in the assay mixture. The enzymes prefer poly[d(A-T - d(A-T)] or partially degraded (activated) DNA as template rather than double-stranded or single-stranded DNA. The activity on activated DNA relative to that on poly[d(A-T) - D(A-T)] was found to be 93, 66, and 29% for DNA polymerases CI, CII, and CIII, respectively.