Effect of the terminal phosphate derivatization of beta- and alpha-oligodeoxynucleotides on their antisense activity in protein biosynthesis, stability and uptake by eucaryotic cells

Inhibition of polypeptide chain elongation with the mRNA-complementary (antisense) oligonucleotide has been realized through a RNase H independent mechanism. Nuclease resistant complementary non-natural alpha-17-mer oligonucleotide did not inhibit cell-free protein biosynthesis of beta-globin in the wheat germ system because it did not elicit RNase H activity. Linkage of alkylating group [4-(N-2-chloroethyl-N-methyl)-aminobenzyl]-methylamine to the 5'-terminus of the alpha-oligomer led to the formation of its covalent adduct with mRNA which could not be translated in vitro. Linkage of hydrophobic residues to the terminal phosphates of natural oligonucleotides increased their stability against nucleases and uptake by human cancer cells. A porphyrin, substituted in the meso-position by aromatic groups, gave a rise to an approximately six-fold increase of uptake and cholesterol a 30-100-fold increase. Eighty percent of bound derivatives were found in cytoplasmic cellular fractions.

Mechanisms of the inhibition of reverse transcription by antisense oligonucleotides

We have demonstrated that the synthesis of cDNA by avian myeloblastosis virus and Moloney murine leukemia virus reverse transcriptases can be prevented by oligonucleotides bound to the RNA template approximately 100 nucleotides remote from the 3' end of the primer. The RNA was truncated at the level of the antisense oligonucleotide-RNA duplex during the reverse transcription. The key role played by the reverse transcriptase-associated RNase H activity in the inhibition process was shown by the use of (i) inhibitors of RNase H (NaF or dAMP), (ii) Moloney murine leukemia virus reverse transcriptase devoid of RNase H activity, or (iii) alpha-analogues of oligomers that do not elicit RNase H-catalyzed RNA degradation. In all three cases the inhibitory effect was either reduced (NaF, dAMP) or totally abolished. However, an alpha-oligomer bound to the sequence immediately adjacent to the primer-binding site prevented reverse transcription. Therefore, initiation of polymerization can be blocked by means of an RNase H-independent mechanism, whereas arrest of a growing cDNA strand can be achieved only by an oligonucleotide mediating cleavage of the template RNA.

Modified oligonucleotides in rabbit reticulocytes: uptake, stability and antisense properties

We have investigated the behaviour of antisense oligonucleotides in rabbit reticulocytes. Both backbone-modified oligomers (methyl-phosphonate and phosphorothioate analogues), anomers of nucleotide units (alpha) and oligonucleotides linked to various ligands (intercalator, polylysine, dodecanol) were tested with respect to cellular uptake and inhibition of protein synthesis. Oligonucleotides added at an external concentration of 10 microM slowly entered the cell up to a concentration of a few hundred nanomolars. A large fraction of phosphorothioate analogues was found to be associated with the membrane. alpha-, methylphosphonate and phosphorothioate analogues remained intact during the incubation with reticulocytes although the latter were partly dephosphorylated. Antisense oligonucleotides were targeted against three different sites of the rabbit beta-globin mRNA: the 5' end of the message, the initiator AUG or the coding sequence. No specific effect on beta-globin synthesis was observed with any of the investigated compounds.

Inhibition of translation initiation by antisense oligonucleotides via an RNase-H independent mechanism

We have used alpha-oligomers as antisense oligonucleotides complementary to three different sequences of the rabbit beta-globin mRNA: a region adjacent to the cap site, a region spanning the AUG initiation codon or a sequence in the coding region. These alpha-oligonucleotides were synthesized either with a free 5' OH group or linked to an acridine derivative. The effect of these oligonucleotides on mRNA translation was investigated in cell-free extracts and in Xenopus oocytes. In rabbit reticulocyte lysate and in wheat germ extracts oligomers targeted to the cap site and the initiation codon reduced beta-globin synthesis in a dose-dependent manner, whereas the target mRNA remained intact. The anti-cap alpha-oligomer was even more efficient that its beta-counterpart in rabbit reticulocyte lysate. In contrast, only the alpha-oligomer, linked to the acridine derivative, complementary to the cap region displayed significant antisense properties in Xenopus oocytes. Therefore initiation of translation can be arrested by oligonucleotide/RNA hybrids which are not substrates for RNase-H.

Blockage of AMV reverse transcriptase by antisense oligodeoxynucleotides

Synthetic oligodeoxynucleotides, either unmodified or linked to an intercalating agent, have been used to prevent cDNA elongation by the AMV reverse transcriptase. Oligonucleotide/RNA hybrids specifically arrest primer extension. The blockage involves the degradation of the RNA part bound to the antisense oligonucleotide by the RNase-H activity associated with the retroviral polymerase.

Effect of RNA secondary structure and modified bases on the inhibition of trypanosomatid protein synthesis in cell free extracts by antisense oligodeoxynucleotides

Every messenger RNA from leishmanias and trypanosomes has at its 5' end a conserved region termed the mini-exon sequence which, however, varies from species to species. In a systematic study mRNAs from Trypanosoma brucei, Trypanosoma vivax, and Leishmania enriettii were translated in cell-free extracts in the presence of oligodeoxynucleotides complementary to part of the mini-exon sequence. The affinity of the same oligonucleotides for target and non-target mRNAs was determined by thermal elution of filter-bound complexes showing that the critical temperature of half-dissociation of the complexes was linearly related to log (l + x), where l is the length of the oligomer and x its G + C content. A few oligomers exhibited a lower Tc value than expected which was ascribed to the presence of modified RNA bases or to the existence of a hairpin structure in the L. enriettii mini-exon. In most cases the efficiency of translation inhibition by the oligonucleotides was clearly correlated to their affinity for the target RNA. The modified bases weakened the inhibition of protein synthesis by oligonucleotides complementary to these regions.

[Antisense oligonucleotides: tools of molecular genetics and therapeutic agents]

The binding of an oligodeoxynucleotide, so-called anti-sense, to the complementary sequence of a messenger RNA can prevent the synthesis of the encoded protein. This approach constitutes a very efficient and specific means to artificially regulate gene expression. Numerous chemical modifications have been introduced into synthetic oligos in order to provide them with properties that unmodified molecules do not display. For instance, oligos built up with methylphosphonate, phosphorothioate and alpha-anomer units lead to molecules that are resistant to DNases. Acridine-linked oligos exhibit an increased affinity for the target sequence due to the intercalation of the dye into the oligo/RNA duplex. Two different mechanisms account for translation inhibition by antisense oligos. Inhibition of the elongation step results only from the induced cleavage of the target RNA by RNase-H. In contrast, oligos targeted upstream of the AUG initiation codon can block the initiation step through an RNase-H independent mechanism. As a consequence, methylphosphonate- and alpha-oligos, which do not elicit RNase-H activity, targeted to the 5' region, are efficient antisense; but they are inactive if targeted to the coding sequence. Experiments performed with antisense oligos in cell-free extracts supported the notion that the mini-exon sequence, acquired by trans-splicing, was present on every message in trypanosomatids and on some of them in nematodes. Furthermore, an acridine-linked oligo complementary to the mini-exon sequence of Trypanosoma brucei induced a lethal effect on cultured procyclics. Therefore these compounds constitute promising tools in molecular genetics and could open new routes to rationally tailor therapeutic agents.

Phosphoroselenoate oligodeoxynucleotides: synthesis, physico-chemical characterization, anti-sense inhibitory properties and anti-HIV activity

Oligodeoxynucleotides with a phosphorus atom in which one of the non-bridging oxygen atoms is substituted by selenium were prepared and investigated with respect to their antisense properties. A general synthesis of phosphoroselenoate analogs of oligonucleotides is described using potassium selenocyanate as the selenium donor. The compounds, characterized by 31P NMR, were shown to decompose to phosphate with a half-life of ca. 30 days. Melting temperatures of duplexes between poly(rA) or poly(rI) with oligo(dT) and oligo(dC), respectively, indicate diminished hybridization capability of phosphoroselenoate oligomers relative to both the unmodified phosphodiester oligomers and the phosphorothioate congeners. A phosphoroselenoate 17-mer is a sequence specific inhibitor of rabbit beta-globin synthesis in wheat germ extract and in injected Xenopus oocytes. In contrast phosphoroselenoate analogs are potent non-sequence specific inhibitors in rabbit reticulocyte lysate. In vitro HIV assays were carried out on a phosphoroselenoate sequence and compared with a phosphorothioate analogue that has previously been shown to exhibit anti-HIV activity (Matsukura et al., Proc. Natl. Acad. Sci. (1987) 84, 7706-7710). The phosphoroselenoate was somewhat less active, and was much more toxic to the cells.

Comparative inhibition of rabbit globin mRNA translation by modified antisense oligodeoxynucleotides

We have studied the translation of rabbit globin mRNA in cell free systems (reticulocyte lysate and wheat germ extract) and in microinjected Xenopus oocytes in the presence of anti-sense oligodeoxynucleotides. Results obtained with the unmodified all-oxygen compounds were compared with those obtained when phosphorothioate or alpha-DNA was used. In the wheat germ system a 17-mer sequence targeted to the coding region of beta-globin mRNA was specifically inhibitory when either the unmodified phosphodiester oligonucleotide or its phosphorothioate analogue were used. In contrast no effect was observed with the alpha-oligomer. These results were ascribed to the fact that phosphorothioate oligomers elicit an RNase-H activity comparable to the all-oxygen congeners, while alpha-DNA/mRNA hybrids were a poor substrate. Microinjected Xenopus oocytes followed a similar pattern. The phosphorothioate oligomer was more efficient to prevent translation than the unmodified 17-mer. Inhibition of beta-globin synthesis was observed in the nanomolar concentration range. This result can be ascribed to the nuclease resistance of phosphorothioates as compared to natural phosphodiester linkages, alpha-oligomers were devoid of any inhibitory effect up to 30 microM. Phosphorothioate oligodeoxyribonucleotides were shown to be non-specific inhibitors of protein translation, at concentrations in the micromolar range, in both cell-free systems and oocytes. Non-specific inhibition of translation was dependent on the length of the phosphorothioate oligomer. These non-specific effects were not observed with the unmodified or the alpha-oligonucleotides.

Antimessenger oligodeoxyribonucleotides: an alternative to antisense RNA for artificial regulation of gene expression--a review

Synthetic oligodeoxyribonucleotides (oligos) are now widely used as artificial regulators for gene expression both in cell-free media and in cultured cells. We describe the biological consequence of the various chemical modifications that have been introduced into the molecules to improve their resistance against nuclease attack, their affinity for the target mRNA and their uptake by cells. We also describe the rising generation of antimessenger oligos. Covalently linked to reactive groups these molecules direct irreversible modifications of the complementary nucleic acids. We anticipate that these oligos will be targeted to double-stranded nucleic acids to interfere with gene expression at the DNA level.

A maximum of two tryptophan residues in gene-32 protein from phage T4 undergo stacking interactions with single-stranded polynucleotides

The effect of specific photochemical and radiochemical modification of tryptophyl and cysteinyl residues of the gene 32 protein (gp 32) of bacteriophage T4 on its affinity towards single-stranded polynucleotides has been investigated. Oxidation of Cys residues of gp 32 by the free-radical anion I-.2 induces a partial loss of the protein affinity, probably by affecting the metal-binding domain which includes three of the four cysteine residues of gp 32. Ultraviolet irradiation of gp 32 in the presence of trichloroethanol results in the modification of three of its five Trp residues and total loss of the protein binding. Analysis of the relative affinity of ultraviolet-irradiated gp 32 for single-stranded polynucleotides suggest that modification of a Trp of enhanced reactivity occurs first and has no effect on the protein binding. Radiochemical modification of three Trp residues of gp 32 by (SCN)-.2 results in total loss of activity. Complexation of gp 32 with denatured DNA prior to gamma-irradiation protects two Trp residues and prevents the protein inactivation. These results suggest that at most two Trp residues are involved in stacking interactions with nucleic acid bases. However, time-resolved spectroscopic methods which allow us to monitor selectively the stacked tryptophan residues have not yielded evidence of more than a single residue undergoing such interactions.

Enzymatic amplification of translation inhibition of rabbit beta-globin mRNA mediated by anti-messenger oligodeoxynucleotides covalently linked to intercalating agents

The effects of anti-messenger oligodeoxynucleotides, covalently linked to an intercalating agent, on translation of rabbit beta-globin mRNA, were investigated both in wheat germ extract and in microinjected Xenopus oocytes. A specific inhibition of beta-globin synthesis was observed in both expression systems with a modified 11-mer covalently linked to an acridine derivative. In injected oocytes a more efficient block was observed with this modified oligonucleotide than with its unsubstituted homolog. This was ascribed to stacking interactions of the intercalating agent with base pairs which provide an additional stabilization of the [mRNA/DNA] hybrid. We demonstrated that in wheat germ extract, the modified and unmodified oligonucleotides behaved similarly due to the presence of a high RNaseH activity. RNaseH was also present, although to a lesser extent, in the oocyte cytoplasm. This anti-messenger DNA-induced degradation of target mRNA resulted in amplified efficiency of hybrid-arrested translation. This additional mechanism might provide anti-sense DNAs with an advantage over anti-sense RNAs.

An acridine-linked oligodeoxynucleotide targeted to the common 5' end of trypanosome mRNAs kills cultured parasites

Anti-messenger oligodeoxynucleotides covalently linked to an intercalating agent were tested for their ability to inhibit translation of Trypanosoma brucei mRNAs in a cell-free system. The sequence of these oligodeoxynucleotides was complementary to part of the 35-nucleotide (nt) sequence which is present at the 5' end of all trypanosome mRNAs (the so-called mini-exon sequence). In a rabbit reticulocyte lysate, a nonadeoxynucleotide linked to an acridine derivative, specifically inhibited protein synthesis from T. brucei mRNAs much more efficiently than unmodified oligodeoxynucleotides of similar length. These oligodeoxynucleotides were tested on cultured trypanosomes. The acridine-linked nonadeoxynucleotide had a lethal effect on the parasites. No effect was observed with the homologous unmodified 9-mer nor with those 9-mers linked to the acridine derivative which were not complementary to the mini-exon sequence. These effects are probably a result of hybrid formation between the anti-messenger and mini-exon sequence. Trypanocidal activity of the acridine-modified nonadeoxynucleotide is most likely due to (i) increased affinity for its target, (ii) improved resistance to 3' exonucleases, and (iii) promoted membrane penetration of living parasites.

Single-strand binding proteins from phage T4 and E. coli form higher order structures with poly(dT)

Complexes of poly(dT) with gene 32 protein from phage T4 or E. coli single-strand binding protein were digested by nuclease P1 from Penicillum citrinum. Protected fragments were analyzed by gel electrophoresis. In both cases, a series of bands was obtained corresponding to multiples of a repeat unit whose size was about 80 nucleotides. Such protected fragments could not be detected under the same experimental conditions when poly(dA) was used instead of poly(dT). The formation of nucleosome-like structures is discussed in relation to the higher affinity exhibited by single-strand binding proteins towards poly(dT).

Anti-messenger oligodeoxynucleotides: specific inhibition of rabbit beta-globin synthesis in wheat germ extracts and Xenopus oocytes

Oligodeoxyribonucleotides complementary to the initiation region of rabbit beta-globin messenger RNA were used to selectively inhibit translation in a wheat germ extract and in injected Xenopus oocytes. The oligonucleotides interacted specifically with their RNA target as shown by thermal denaturation studies of hybrids on nitrocellulose filters. The longest oligonucleotide used (17-mer) efficiently blocked translation both in vitro and in vivo. In contrast the shortest one (8-mer) exhibited only a limited effect. The translation block was specific. The synthesis of endogenous proteins in oocytes and that of alpha-globin in the in vitro system were not affected by anti-beta-globin oligonucleotides. A non-complementary oligonucleotide had no inhibitory effect.

The common 5' terminal sequence on trypanosome mRNAs: a target for anti-messenger oligodeoxynucleotides

Several mature mRNAs of Trypanosoma brucei were previously shown to have a common 5' terminal sequence of 35 nucleotides (nt) encoded by a separate mini-exon. To verify whether all trypanosome mRNAs contain this mini-exon sequence at their 5' end, we have tested oligodeoxynucleotides complementary to different parts of the 35 nt leader sequence for their ability to inhibit translation of total trypanosome mRNA. All oligomers tested inhibited translation of trypanosome mRNAs in a wheat germ extract. They had no effect on translation of Brome mosaic virus mRNA and of a trypanosome mRNA for phosphoglycerate kinase modified to remove the mini-exon sequence. Three different 12mers inhibited translation 35-60%; both the 22- and 34mer inhibited translation 95-100%. Incorporation of amino acids decreased proportionally in all protein bands detected in high resolution polyacrylamide gels. Our results show that all trypanosome mRNAs that yield a product detectable in gel contain a mini-exon sequence. We infer that most, if not all, trypanosome mRNAs contain a 5' terminal mini-exon sequence acquired by discontinuous synthesis.

Specific inhibition of mRNA translation by complementary oligonucleotides covalently linked to intercalating agents

Synthetic oligodeoxynucleotides that are covalently linked at their 3' end to an acridine derivative and are complementary to the repeated sequence UUAAAUUAAAUUAAA adjacent to the ribosome binding site of the gene 32-encoded mRNA from phage T4 have been used to regulate the synthesis of gene 32-encoded protein in vitro. These modified, synthetic oligonucleotides specifically block the translation of gene 32-encoded mRNA with a higher efficiency than the homologous unsubstituted oligonucleotides. The inhibition produced by these short "anti-messengers" is due to the formation of specific mRNA . oligodeoxynucleotide hybrids that are stabilized by the intercalation of the acridine ring in the RNA . DNA duplex.

Specific recognition by the tripeptide lysyl-tryptophyl-alpha-lysine of structural damage induced in DNA by platinum derivatives

The antitumoral derivative cisPt binds to DNA, as do its inactive analogs, trans- and dienPt. Structural damage introduced into DNA after reaction with the Pt derivatives were probed by using the peptide LysTrpLys. This peptide was used for its preferential binding to single-stranded structures (Brun, F., Toulmé, J.J. and Hélène, C. (1975) Biochemistry 14, 558-563). Phosphorescence lifetime measurements show that the Pt-induced heavy atom effects are quite similar in the three peptide-DNA-Pt complexes whatever the nature of the Pt derivative used. In contrast, fluorescence quenching strongly depends on the nature of the Pt derivatives. This quenching was therefore attributed to the stacking interactions engaged by the tryptophan residue with nucleic acid bases. A comparison of fluorescence quenching data for native and modified DNAs demonstrates that modification by dienPt has no effect on stacking interactions and that high levels of modifications by trans Pt are required to observe a change in stacking efficiency. In contrast modification by cis Pt induces the formation of strong stacking sites. The results strongly suggest the existence of locally opened regions in DNA modified by cis Pt.

Oligodeoxynucleotides covalently linked to intercalating agents: a new class of gene regulatory substances

Oligodeoxynucleotides have been covalently linked to a 9-aminoacridine derivative via their 3'-phosphate group. Specific complexes are formed with the complementary sequence of the oligonucleotide. The stability is strongly increased due to intercalation of the acridine derivative. Absorption, fluorescence, nuclear magnetic resonance and circular dichroism have been used to characterize complex formation. The stability of the complexes depends on the length of the linker between the acridine derivative and the 3'-phosphate group of the oligonucleotide. Oligonucleotides covalently linked to an intercalating agent can be used to selectively control gene expression. Transcription initiation can be blocked when such an oligonucleotide binds to the transcribed strand in the open complex formed by E. coli RNA polymerase with the bla promoter. With some oligonucleotides, non-specific effects on transcription can be detected, most probably due to binding of the modified oligonucleotide to RNA polymerase. Translation of the messenger RNA from gene 32 of phage T4 can be prevented by using an oligonucleotide complementary to the sequence upstream from the Shine-Dalgarno sequence. Inhibition of translation does not occur in the absence of the intercalating agent covalently linked to the oligonucleotide nor with oligonucleotides which do not have a target sequence on the mRNA.

Recognition of damaged regions in DNA by oligopeptides and proteins

The binding of various damaged DNAs to the single-strand binding protein coded for by gene 32 from bacteriophage T4, on the one hand, and of oligopeptides containing tryptophan and lysine residues, on the other hand, is described. These molecules exhibit a higher affinity for modified DNA than for native DNA in so far as modification results in a local destabilization of the double-stranded structure of the nucleic acid. Stacking interactions between aromatic amino acids and nucleic acid bases appear to play a crucial role in the recognition of destabilized regions induced by chemical agents (carcinogens and antitumor drugs). These interactions confer to the peptide lysyl-tryptophyl-lysine an endonucleolytic activity specific for apurinic sites. From results obtained with such oligopeptides a model for the active sites of Ap-endonucleases is proposed which could account for the strategy used by the denV endonuclease from phage T4 during the first step of excision repair of pyrimidine dimers in DNA. The effect of the overall conformation of modified DNA on repair efficiency is discussed.

Influence of DNA binding on the formation and reactions of tryptophan and tyrosine radicals in peptides and proteins

The rate constant of the one-electron oxidation of the tryptophan (Trp) or tyrosine (Tyr) residues by Br- X 2 radical anions is strongly decreased when the peptides are bound to DNA. Oxidation by N X 3 is much less affected by binding. These results can be explained by electrostatic repulsion between the charged polyphosphate backbone and the Br- X 2 radicals. Once oxidized, the interacting aromatic residues react with the DNA in a first order process with a rate constant of the order 10(3) s-1. These results have been extended to the single strand binding protein: the product of gene 32 of phage T4 (gp 32). The pulse radiolysis study suggests that one Trp residue of the protein oxidized by the Br- X 2 radicals reacts with the DNA in the complex while one Tyr residue is buried upon association. It is also shown that the exposure of Trp and Tyr residues to radical attack depends on whether the T4 SSB protein is bound to native or heat-denatured DNA.

Binding of RecA protein to single-stranded nucleic acids: spectroscopic studies using fluorescent polynucleotides

Binding of the recA gene product from Escherichia coli to single-stranded polynucleotides has been investigated using poly(dA) that have been modified by chloroacetaldehyde to yield fluorescent 1,N6-ethenoadenine (epsilon A) bases. A strong enhancement of the fluorescent quantum yield of poly(d epsilon A) is induced upon RecA protein binding. A 4-fold increase is observed in the absence of ATP or ATP gamma S and a 7-fold increase in the presence of either nucleoside triphosphate. RecA protein can bind to poly(d epsilon A) in the absence of both Mg2+ ions and ATP (or ATP gamma S) but Mg2+ ions are required to observe RecA protein binding in the presence of ATP (or ATP gamma S) at pH 7.5. ATP binding to the RecA-poly(d epsilon A) complex induces a dissociation of RecA from the polynucleotide followed by re-binding of [RecA-ATP-Mg2+] ternary complex. Whereas ATP-induced dissociation of RecA-poly(d epsilon A) complexes is a fast process, the subsequent binding reaction of [RecA-ATP-Mg2+] is slow. A model is proposed whereby [RecA-ATP-Mg2+] binding to poly(d epsilon A) involves slow nucleation and elongation processes along the polynucleotide backbone. The nucleation reaction is shown to involve at least a trimer or a tetramer. Polymerization of the [RecA-ATP-Mg2+] ternary complex stops when the polynucleotide is entirely covered with 6 +/- 1 nucleotides per RecA monomer. ATP hydrolysis then induces a release of RecA-ADP complexes from the polynucleotide template.

Recognition of chemically damaged DNA by the gene 32 protein from bacteriophage T4

We have used fluorescence spectroscopy to investigate the binding of gene 32 protein from bacteriophage T4 to DNA which has been chemically modified with carcinogens or antitumor drugs. This protein exhibits a high specificity for single-stranded nucleic acids and binds more efficiently to DNA modified either with cis-diaminodichloroplatinum(II) or with aminofluorene derivatives than to native DNA. This increased affinity is related to the formation of locally unpaired regions which are strong binding sites for the single-strand binding protein. In contrast, gene 32 protein has the same affinity for native DNA, DNA containing methylated purines and DNA that has reacted with trans-diaminodichloroplatinum(II) or with chlorodiethylenetriaminoplatinum(II) chloride. These types of damage do not induce a sufficient structural change to allow gene 32 protein binding. Depurination of DNA does not create binding sites for the T4 gene 32 protein but nicked apurinic sites are strong ligands for the protein. This T4 single-strand binding protein does not exhibit a significantly increased affinity for nicked DNA as compared with native DNA. These results are discussed with respect to the recognition of DNA damage by proteins involved in DNA repair and to the possible role of single-strand binding proteins in DNA repair mechanisms.

[Recognition and cleavage of apurinic sites in DNA by the tripeptide lysyl-tryptophyl-lysine]

Tryptophan-containing peptides, such as Lys-Trp-Lys, Lys-Trp-Gly-Lys and Lys-Gly-Trp-Lys, bind with high affinity to apurinic sites in DNA. The association constant for binding of Lys-Trp-Lys to an apurinic site is two orders of magnitude higher than that for binding to a native site. This is due to a very efficient stacking of tryptophan with the bases on both sides of the vacant apurinic site. When the complexes were incubated at 37 or 45 degrees C a cleavage of the phosphodiester bond was observed. Using pBR 322 DNA containing apurinic sites, conversion of the superhelical to the relaxed circular form was observed as a result of single-strand breakage. The peptide Lys-Gly-Lys had no effect and Lys-Trp-Lys did not induce any cleavage in pBR 322 DNA which did not contain any apurinic site. Therefore, a simple tripeptide, Lys-Trp-Lys, exhibits both the specificity of recognition and the activity of an endonuclease for apurinic sites in DNA.

The binding of T4 gene 32 protein to MS2 virus RNA and transfer RNA

Fluorescence titrations, absorption spectroscopy and stopped-flow techniques were used to study the interaction of T4 coded 32-protein (P 32) with MS2 RNA and total tRNA from E. coli under different ionic conditions. It is shown that the amount of MS2 RNA and tRNA secondary structure melted by P 32 varies markedly and reversibly within a range of ionic conditions under which the binding constant of P 32 to single-stranded nucleic acids unable to form stable hairpins remains higher than 10(8) M-1. Kinetic experiments suggest that P 32 dissociates from the MS2 RNA rewinding strand with a similar rate constant as calculated for the dissociation from single-stranded regions. Possible in vivo consequences of these findings are discussed.

Fluorescence study of the association between gene 32 protein of bacteriophage T4 and poly(1-N6-ethenoadenylic acid). Evidence for energy transfer

We have studied the association of gene 32 protein of bacteriophage T4 with a fluorescent polynucleotide: poly(1-N6-ethenoadenylic acid). The presence of a bridge between the N(1) nitrogen atom and the C(6) amino group of adenine bases did not alter the affinity of the protein for the polynucleotide as compared to poly(rA) and heat-denatured DNA. This suggests that this region of the nucleic acid bases is not required in protein 32-polynucleotide complexes. The formation of gene 32 protein-poly(1-N6-ethenoadenylic acid) complex resulted in an enhancement of the polynucleotide fluorescence quantum yield which could be related to a partial unstacking of the polynucleotide bases. Energy transfer at the singlet level was demonstrated from tryptophan to 1-N6-ethenoadenine residues. The efficiency of energy transfer was calculated to be 22% which is consistent with the presence of at least one of the tryptophan residues of gene 32 protein in close vicinity of the bases.

Effect of phosphate ions on the fluorescence of tryptophan derivatives. Implications in fluorescence investigation of protein-nucleic acid complexes

In order to test the ability of phosphate groups to quench the fluorescence of tryptophan in protein-nucleic acid complexes we have studied the effect of various phosphate ions on the fluorescence of tryptophan derivatives. Unsubstituted and monoalkyl monoanions (H2PO4- and CH3OPO3H-) quench the fluorescence of all investigated indole derivatives while the dimethyl anion (CH3O)2 PO2- does not. This suggests that quenching of tryptophan fluorescence by phosphate monoanions requires the presence of an acidic OH group and could be due to a proton transfer from the phosphate ion to the indole chromophore. Trianions (PO4 3-4) which are strong proton acceptors quench the fluorescence of all tryptophan derivatives except N(1)methyl tryptophan. This result strongly supports our proposal that quenching of tryptophan fluorescence by phosphate trianions occurs through deprotonation of the NH indole group. Bianions (HPO '4(7), and CH3O PO3 2-3) quench the fluorescence of several indole derivatives including N-acetyl tryptophanamide but have no effect on tryptophan or N(1)-methyl tryptophan. From our results we conclude that phosphate groups of nucleic acids are not able to quench the fluorescence of tryptophyl residues in protein-nucleic acid complexes except if an accessible residue is located near a phosphorylated polynucleotide chain end.

Specific recognition of single-stranded nucleic acids. Interaction of tryptophan-containing peptides with native, denatured, and ultraviolet-irradiated DNA

The binding of the tripeptide Lys-Trp-Lys to native, denatured, and ultraviolet-irradiated DNAs has been investigated by fluorescence spectroscopy. Two types of complexes are formed which both involve electrostatic interactions. Only one of them involves a stacking of the tryptophyl ring with nucleic acid bases. Quantitative analysis of fluorescence data shows that this stacking interaction is strongly favored in denatured as compared to native DNA. In ultraviolet-irradiated DNA, the peptide Lys-Trp-Lys binds selectively to unpaired regions around thymine dimers. Due to the stacking interaction of the aromatic amino acid with nucleic acid bases, this simple tripeptide is therefore able to discriminate between single-stranded and double-stranded regions in a nucleic acid.

[Spectroscopic study of the structural changes of scorpion hemocyanin, induced by pH variations and addition of various salts]

Structural modifications of the scorpion haemocyanin induced by pH variations and salt addition are studied by U.V. absorption, fluorescence, circular dichroism and light scattering. Haemocyanin fluorescence is due to both aromatic amino-acids tyrosine and tryptophan. Deoxygenation or denaturation lead to a fourfold enhancement of its intensity. At acidic pH the active site is modified and the protein is dissociated, but at alkaline pH the haemocyanin aggregates. The addition of different salts (sodium citrate, potassium bromide and iodide...) involves protein dissociation, the amplitude of which depends on the anion. But pH variations and salt addition don't change the haemocyanin secondary structure as shown by circular dichroism. The C.D. spectrum of scorpion haemocyanin exhibits the characteristic bands of Arthropod haemocyanine.

Interactions of aromatic residues of proteins with nucleic acids. Fluorescence studies of the binding of oligopeptides containing tryptophan and tyrosine residues to polynucleotides

The binding of oligopeptides of general structure Lys-X-Lys (where X is an aromatic residue) to several polynucleotides has been studied by fluorescence spectroscopy. Two types of complexes are formed, both involving electrostatic interactions between lysyl residues and phosphate groups as shown by the ionic strength and pH dependence of binding. The fluorescence quantum yield of the first complex is identical with that of the free peptide. The other complex involves a stacking of the nucleic acid bases with the aromatic amino acid whose fluorescence is quenched. Fluorescence data have been quantitatively analyzed according to a model involving these two types of complexes. Association constants and the size of binding sites have been determined. Stacking interactions are favored in single-stranded polynucleotides as compared to double-stranded ones. A short oligopeptide such as Lys-X-Lys is thus able to distinguish between single-stranded and double-stranded nucleic acids. Fluorescence results are compared to those obtained by proton magnetic resonance and circular dichroism.

Specific recognition of single-stranded regions in ultraviolet-irradiated and heat-denatured DNA by tryptophan-containing peptides

Fluorescence studies of the binding of peptides containing lysyl and tryptophyl residues to nucleic acids show that two types of complexes are formed. One of them involves a direct interaction of the tryptophyl ring with nucleic acid bases, which leads to fluorescence quenching. Comparison with proton magnetic resonance and circular dichroism data indicates that this fluorescence quenching is due to a stacking of the indole ring with bases. Quantitative analysis of fluorescence data leads to the conclusion that stacking is favored in single-stranded regions of DNAs, which are produced either by heating or by UV-irradiation of the native DNA sample. The binding of the peptide Lys-Trp-Lys is about ten times tighter in these single-stranded regions as compared with double-stranded ones. A short tripeptide such as Lys-Trp-Lys is, therefore, able to discriminate between single-stranded and double-stranded regions. Moreover, bound peptide molecules photosensitize the splitting of thymine dimers in UV-irradiated DNA, thus providing a model for DNA photoreactivation.