Characterization of the ATP-dependent binding of wheat germ protein synthesis initiation factors eIF-(iso)4F and eIF-4A to mRNA.

The ATP-dependent binding of wheat germ protein synthesis initiation factors eIF-(iso)4F and eIF-4A to an oligoribonucleotide has been investigated by direct fluorescence titration techniques. In addition, the effect of ATP on the interaction between another cap-binding initiation factor, eIF-4F, and eIF-4A was studied using the same methods. Comparison of the equilibrium association constants (K(eq)) indicate that 1) hydrolyzable ATP affects the affinity of eIF-(iso)4F for eIF-4A, regardless of whether or not mRNA was previously bound to the eIF-(iso)4F; in contrast, ATP had no effect on the eIF-(iso)4F/oligoribonucleotide interaction; 2) in the presence of ATP, the binding of the binary eIF-(iso)4F.eIF-4A complex to the oligoribonucleotide is of similar affinity as the binding of the oligoribonucleotide to the eIF-(iso)4F alone; the stoichiometry of this ternary eIF-(iso)4F.eIF-4A.mRNA complex was found to be 1:1:1; and 3) a similar ATP effect is observed for the eIF-4F/eIF-4A interaction as for the eIF-(iso)4F.eIF-4A complex.

Differences in unwinding of supercoiled DNA induced by the two enantiomers of anti-benzo[a]pyrene diol epoxide.

The unwinding of supercoiled phi X174 RFI DNA induced by the tumorigenic (+) and non-tumorigenic (-) enantiomers of trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE) has been investigated by agarose slab-gel and ethidium titration tube gel electrophoresis. The differences in adduct conformations were verified by flow linear dichroism techniques. Both enantiomers cause a reversible unwinding by the formation of noncovalent intercalative complexes. The effects of covalently bound BPDE residues on the electrophoretic mobilities of the RF I DNA form in agarose gels were investigated in detail in the range of binding ratios rb approximately 0.0-0.06 (covalently bound BPDE residues/nucleotide). In this range of rb values, there is a striking difference in the mobilities of (+)-BPDE- and (-)-BPDE-adducted phi X174 DNA in agarose slab-gels, the covalently bound (+)-BPDE residues causing a significantly greater retardation than (-)-BPDE residues. Increasing the level of covalent adducts beyond rb approximately 0.06 in the case of the (+)-BPDE enantiomer, leads to further unwinding and a minimum in the mobilities (corresponding to comigration of the nicked form and the covalently closed relaxed modified form) at rb 0.10 +/- 0.01; at still higher rb values, rewinding of the modified DNA in the opposite sense is observed. From the minimum in the mobility, a mean unwinding angle (per BPDE residue) of theta = 12 +/- 1.5 degrees is determined, which is in good agreement the value of theta = 11 +/- 1.8 degrees obtained by the tube gel titration method. Using this latter method, values of theta = 6.8 +/- 1.7 degrees for (-)-BPDE-phi X174 adducts are observed. It is concluded that agarose slab gel techniques are not suitable for determining unwinding angles for (-)-BPDE-modified phi X174 DNA because the alterations in the tertiary structures for rb < 0.06 are too small to cause sufficiently large changes in the electrophoretic mobilities. The major trans (+)-BPDE-N2-guanosine covalent adduct is situated at external binding sites and the mechanisms of unwinding are therefore different from those relevant to noncovalent intercalative BPDE-DNA complexes or to classical intercalating drug molecules; a flexible hinge joint and a widening of the minor groove at the site of the lesion may account for the observed unwinding effects. The more heterogeneous (-)-BPDE-nucleoside adducts (involving cis and trans N2-guanosine, and adenosine adducts) are less effective in causing unwinding of supercoiled DNA for reasons which remain to be elucidated.

Characterization of the interaction of wheat germ protein synthesis initiation factor eIF-3 with mRNA oligonucleotide and cap analogues.

Direct fluorescence titration experiments of wheat germ protein synthesis initiation factor eIF-3 with mRNA cap and oligoribonucleotide analogues were performed in order to determine the equilibrium association constants (Keq) for the eIF-3.mRNA interaction as a function of pH and temperature. These data suggest that (i) the eIF-3.mRNA interaction is not cap-specific (i.e., m7G-specific), (ii) ATP hydrolysis is not involved in the interaction, and (iii) the interaction is primarily ionic in nature. Competition experiments between a rabbit alpha-globin mRNA oligoribonucleotide analogue and either mRNA cap analogues or nucleoside triphosphates (NTPs) are also reported; these experiments indicate that NTPs act as both activators and competitive inhibitors of the mRNA.eIF-3 association. The results are consistent with a partially uncompetitive binding mechanism, whereby at low NTP concentrations (less than or equal to 10 microM) the bound NTP enhances subsequent mRNA binding to eIF-3, perhaps by inducing a conformational change, and at higher NTP concentrations, the NTP acts as a competitive inhibitor for the mRNA binding site on eIF-3.

A spectroscopic study of the binding of m7GTP and m7GpppG to human protein synthesis initiation factor 4E.

The binding of analogues of the 7-methylguanosine-containing cap, m7GTP and m7GpppG, to eIF-4E from human erythrocytes as a function of pH, temperature, and ionic strength is described. From the pH-dependent binding of m7GTP and m7GpppG to eIF-4E, a new model describing the nature of the cap.eIF-4E interaction is proposed. The thermodynamic values and ionic strength dependence of binding are consistent with a binding site which is primarily hydrophobic. Fluorescence and circular dichroism data indicate that tryptophan residues may be involved in base-stacking interactions with the cap in a somewhat buried environment. The model presented here confirms the earlier proposal [Rhoads et al. (1983) Biochemistry 22, 6084-6088] that the enolate tautomer of the cap is preferred for interaction and further proposes that the interaction is with a protonated amino acid residue, such as histidine, while stacking with an aromatic amino acid, such as tryptophan.

Reactions of stereoisomeric non-bay-region benz[a]anthracene diol epoxides with DNA and conformations of non-covalent complexes and covalent adducts.

The reactions of the non-bay-region diol epoxides racemic trans-8,9-dihydroxy-anti-10,11-epoxy-8,9,10,11-tetrahydrobenz[a]an thracene (anti-BA-10,11-DE) and racemic trans-8,9-dihydroxy-syn-10,11-epoxy-8,9,10,11-tetrahydrobenz[a]ant hracene (syn-BA-10,11-DE) with native double-stranded DNA in aqueous solutions (5 mM sodium cacodylate buffer, pH 7.0, 23 degrees C) was investigated utilizing various spectroscopic techniques. The results of linear dichroism experiments suggest that both diastereomers form non-covalent, intercalative complexes with DNA prior to undergoing chemical reactions; the association constant for the anti stereoisomers is about twice as large (850 +/- 100 M-1) as that for the syn-diastereomers, thus qualitatively paralleling the behavior established previously for the bay-region diol epoxides of benzo[a]pyrene and benz[a]anthracene. The reaction rates of both anti- and syn-BA-10,11-DE are significantly accelerated in the presence of DNA, and the fraction of diol epoxide molecules which bind covalently to DNA is 13 +/- 2% and 3 +/- 1% respectively; these levels of covalent binding are lower by factors of about two respectively, than in the case of the bay-region diol epoxides of benz[a]anthracene. The phenanthrenyl residues in the covalent anti-BA-10,11-DE-DNA adducts are tilted with their long axes closer to the average orientations of the normals to the DNA bases; in contrast, the adducts derived from the binding of the syn diastereomers, appear to be characterized by intercalative-type conformations; however, the overall degrees of orientations are weak in the cases of these non-bay-region diol epoxide-DNA adducts. Nevertheless, these adduct conformations resemble those derived from the highly tumorigenic anti and the less active syn diasteromers of benzo[a]pyrene and benz[a]anthracene, thus providing one additional example to the previously observed correlations between adduct structure and biological activity.

Reactions of stereoisomeric and structurally related bay region diol epoxide derivatives of benz[a]anthracene with DNA. Conformations of noncovalent complexes and covalent carcinogen-DNA adducts.

The modes of reaction of the tumorigenic bay region diol epoxide anti-BADE [+/-)-trans-3,4-diol-anti-1,2-epoxy-1,2,3,4-tetrahydrobenz[a]anthr acene) and the less potent tumor initiating diastereomer syn-BADE [+/-)-trans-3,4-diol-syn-1,2-epoxy-1,2,3,4-tetrahydrobenz[a]anthra cene) with native, double-stranded DNA were compared. The bay-region diol epoxide derived from 3-methylcholanthrene (3-MCDE, racemic trans-9,10-diol-anti-7,8-epoxy-7,8,9,10-tetrahydromethylcholanthrene+ ++) was included in this study in order to assess the effects of the methyl and methylene substituents on the reactivity with DNA. Utilizing linear dichroism and other spectroscopic methods, it is shown that all three diol epoxides forn non-covalent complexes with DNA. The diastereomers anti-BADE and syn-BADE form intercalative physical complexes, but the association constant K of the syn-diastereomer is about 6-7 times smaller than for anti-BADE; this effect is ascribed to the bulky quasi-diaxial conformation of the diol epoxide ring in the syn diastereomer. The value of K (4000 M-1) is similar for anti-BADE and 3-MCDE, although the latter is not intercalated in the classical sense since the short axis of the molecule is tilted closer to the axis of the DNA double helix. The conformations of the covalent DNA adducts are interpreted in terms of a quasi-intercalative conformation (site I), and a conformation in which the long axes of the polycyclic molecules are tilted closer to the axis of the helix (site II). Both tumorigens, anti-BADE and 3-MCDE, undergo a marked re-orientation from a non-covalent site I to a covalent site II conformation upon binding chemically with the DNA bases, although a small fraction of the covalent anti-BADE adducts remains quasi-intercalated; in contrast, the alkyl substituents in 3-MCDE not only prevent the formation of intercalative physical complexes, but also the formation of site I covalent adducts. In the case of the less tumorigenic syn-BADE, both the non-covalent complexes and the covalent adducts are of the site I-type. The bay-region diol epoxide of benz[a]anthracene and of 3-methylcholanthrene display a similar pattern of reactivities and covalent adduct conformations as the bay region diol epoxide derivatives of benz[a]pyrene, suggesting that adduct conformation might be an important factor in determining the levels of mutagenic and tumorigenic activities of this class of compounds.

Subunit dissociation of Busycon canaliculatum hemocyanin.

The hemocyanin of the channeled whelk, Busycon canaliculatum, is a multisubunit protein with a molecular weight close to 9 X 10(6). The increase in pH above neutrality and the addition of 0-5 M urea and 0-2 M GdnHCl is found to dissociate the whole molecules to half-molecules and smaller dimeric and monomeric fragments of one-tenth and one-twentieth mass of the parent hemocyanin. The molecular weight transitions investigated at constant protein concentration of 5 X 10(-2) g X l-1 show no clearly discernible plateau regions, where essentially only half-molecules and one-tenth molecules are present. The ultracentrifugation patterns in much of the dissociation region produced by urea at pH 6.9 suggests the presence of three distinct components consisting of whole molecules, half-molecules and largely one-tenth molecular weight fragments. At pH 8.2 and higher, where whole molecules are largely absent, the effects of urea on the dissociation of half-molecules to tenths and tenth-molecules to twentieth molecule was investigated by means of light scattering. Analysis of the urea data based on a decamer to dimer and dimer to monomer scheme of dissociation used in our earlier studies gave apparent estimates of about 90 amino acid groups at the contact areas of the dimers in the half-molecules and 110 groups at the monomer contacts forming the dimers. The latter relatively large estimate of groups suggests that the dissociation of the tenth molecules or dimers must occur by longitudinal splitting of the contact areas along both the folded domains and the connecting chain segments of the twentieth molecules. Circular dichroism, absorbance and viscosity data suggest that the secondary structure and conformation of the folded domains of the hemocyanin subunits are largely retained at both high pH and in 3-8 M urea solutions. The molecular weights at pH 9.0-10.6 and in 3-8 M urea are found to be (4.2-4.7) X 10(5), close to one-twentieth of the mass of the parent hemocyanin. Denaturation and unfolding of the subunit domains is observed between 3 and 6 M GdnHCl solutions, as evidenced by the abolition of the characteristic copper absorbance in the neighborhood of 346 nm and the relatively pronounced changes in circular dichroism at 222 nm and intrinsic viscosity. The further decrease in molecular weights to about (2.6-3.2) X 10(5), below one-twentieth of the mass of hemocyanin suggests the presence of hidden breaks or scissions in the polypeptide chains suffered during isolation, which become exposed as a result of complete unfolding in GdnHCl solutions.(ABSTRACT TRUNCATED AT 400 WORDS)