Thermostability of double-stranded deoxyribonucleic acids: effects of covalent additions of a psoralen

Preparation and Purification of Oligodeoxynucleotide Duplexes Containing a Site-Specific, Reduced, Chemically Stable Covalent Interstrand Cross-Link Between a Guanine Residue and an Abasic Site

Methods for the preparation of DNA duplexes containing interstrand covalent cross-links may facilitate research in the fields of biochemistry, molecular biology, nanotechnology, and materials science. Here we report methods for the synthesis and isolation of DNA duplexes containing a site-specific, chemically stable, reduced covalent interstrand cross-link between a guanine residue and an abasic site. The method uses experimental techniques and equipment that are common in most biochemical laboratories and inexpensive, commercially available oligonucleotides and reagents.

Protein Recognition in Drug-Induced DNA Alkylation: When the Moonlight Protein GAPDH Meets S23906-1/DNA Minor Groove Adducts

DNA alkylating drugs have been used in clinics for more than seventy years. The diversity of their mechanism of action (major/minor groove; mono-/bis-alkylation; intra-/inter-strand crosslinks; DNA stabilization/destabilization, etc.) has undoubtedly major consequences on the cellular response to treatment. The aim of this review is to highlight the variety of established protein recognition of DNA adducts to then particularly focus on glyceraldehyde-3-phosphate dehydrogenase (GAPDH) function in DNA adduct interaction with illustration using original experiments performed with S23906-1/DNA adduct. The introduction of this review is a state of the art of protein/DNA adducts recognition, depending on the major or minor groove orientation of the DNA bonding as well as on the molecular consequences in terms of double-stranded DNA maintenance. It reviews the implication of proteins from both DNA repair, transcription, replication and chromatin maintenance in selective DNA adduct recognition. The main section of the manuscript is focusing on the implication of the moonlighting protein GAPDH in DNA adduct recognition with the model of the peculiar DNA minor groove alkylating and destabilizing drug S23906-1. The mechanism of action of S23906-1 alkylating drug and the large variety of GAPDH cellular functions are presented prior to focus on GAPDH direct binding to S23906-1 adducts.

Minor groove orientation of the KWKK peptide tethered via the N-terminal amine to the acrolein-derived 1,N2-gamma-hydroxypropanodeoxyguanosine lesion with a trimethylene linkage

DNA−protein conjugates are potentially repaired via proteolytic digestion to DNA−peptide conjugates. The latter have been modeled with the amino-terminal lysine of the peptide KWKK conjugated via a trimethylene linkage to the N²-dG amine positioned in 5′-d(GCTAGCXAGTCC)-3′·5′-d(GGACTCGCTAGC)-3′ (X = N²-dG−trimethylene link−KWKK). This linkage is a surrogate for the reversible linkage formed by the γ-OH-1,N²-propanodeoxyguanosine (γ-OH-PdG) adduct. This conjugated KWKK stabilizes the DNA. Amino acids K²⁶, W²⁷, K²⁸, and K²⁹ are in the minor groove. The W²⁷ indolyl group does not intercalate into the DNA. The G⁷N² amine and the K²⁶ N-terminal amine nitrogens are in the trans configuration with respect to the C_α or C_γ of the trimethylene tether, respectively. The structure of this DNA−KWKK conjugate is discussed in the context of its biological processing.

DNA-Destabilizing Agents as an Alternative Approach for Targeting DNA: Mechanisms of Action and Cellular Consequences

DNA targeting drugs represent a large proportion of the actual anticancer drug pharmacopeia, both in terms of drug brands and prescription volumes. Small DNA-interacting molecules share the ability of certain proteins to change the DNA helix's overall organization and geometrical orientation via tilt, roll, twist, slip, and flip effects. In this ocean of DNA-interacting compounds, most stabilize both DNA strands and very few display helix-destabilizing properties. These types of DNA-destabilizing effect are observed with certain mono- or bis-intercalators and DNA alkylating agents (some of which have been or are being developed as cancer drugs). The formation of locally destabilized DNA portions could interfere with protein/DNA recognition and potentially affect several crucial cellular processes, such as DNA repair, replication, and transcription. The present paper describes the molecular basis of DNA destabilization, the cellular impact on protein recognition, and DNA repair processes and the latter's relationships with antitumour efficacy.

Orientation of Dye Molecules in DNA-Based Films with Chain Alignment and Judgment of Their DNA-Binding Modes

Novel composite films of chain-oriented DNA, which contain the DNA-binding dyes aligned in specific orientation, were successfully prepared by drying the solution under a horizontal magnetic field. Most of the dye-DNA composite films showed linear dichroism, as revealed by polarized ultraviolet-visible (UV-vis) spectroscopy. The intercalators, ethidium bromide and acridine orange, were fixed in chain-oriented DNA films in a similar binding manner as in solutions. Also, Hoechst 33258 and 4',6-diamidino-2-phenylindole were found to be aligned along the minor groove, even in the solid films. Thus, our new method of preparing dye-DNA composite films with chain orientation is useful for aligning small molecules, and it will provide views of the novel anisotropic materials expected in various application fields. We used this method to prepare composite DNA films with newly designed original compounds. Seven of nine dyes were judged to bind obviously to DNA as intercalators by polarized UV-vis spectroscopy. The DNA-binding manners were further analyzed by fluorescence anisotropy measurements. On the basis of the curves for the rotational angle dependence of the anisotropy, we were able to estimate the angles between the transition-dipole moments of dyes and the aligned chain axis of DNA. Interestingly, two original compounds were found to be in the tilted forms with regard to the plane of base pairs. We emphasize here that the method using aligned dye-DNA films is very convenient for identifying the binding modes of the compounds for double-stranded DNA.

Molecular Chain Orientation of DNA Films Induced by Both the Magnetic Field and the Interfacial Effect

DNA films showing highly homogeneous orientation of molecular chains were successfully prepared by drying a semidiluted solution in a horizontal magnetic field. Most of the molecular chain elements in the obtained film were found to be one-dimensionally oriented, as shown by X-ray diffraction, polarization microscopy, and linear dichroism spectroscopy. Because a DNA chain is theoretically expected to orientate only in divergent directions perpendicular to a magnetic field, this result suggests that the DNA chains were aligned not only by a magnetic field but also by the interfacial effect that induced the chains to fit along the air-liquid interface. The descent speed of an air-liquid interface by evaporation was faster than the estimated diffusion rate of DNA, suggesting an emergence of a concentrated layer near the surface. As proved by polarization microscopy, this emergence led to the transitional formation of a nematic-like liquid crystalline phase, which resulted in a DNA film with good chain alignment and unitary orientation. This mechanism underlying chain alignment was supported by molecular weight dependency, in which higher molecular weight DNA is more likely to evince chain alignment that exhibits a higher degree of birefringence. Low molecular weight components have such high thermal motility that it would be difficult to fit them along the air-liquid interface in the early stage of drying. For chain alignment, it was preferable to use an initial concentration of DNA lower than a critical concentration for liquid crystal formation so that the possible diffusion and assembly in a diluted solution would be essential for chain alignment. The DNA film exhibited obvious linear dichroism, indicating the potential for further applications.

Binding of Mg2+, Cd2+, and Ni2+ to Liquid Crystalline NaDNA: Polarized Light Microscopy and NMR Investigations

The interaction of the divalent metal ions Mg(2+), Cd(2+), and Ni(2+) with liquid crystalline NaDNA solutions (molar ratios Me(2+)/DNA-phosphate </=0.050) was investigated by polarized light microscopy and multinuclear (31)P, (2)H, and (23)Na NMR. Our findings show that the state of the cholesteric NaDNA phase at equal MgCl(2), CdCl(2), or NiCl(2) concentration is affected in a different way and to a different extent by the nature of the divalent cation. Indeed, we found that the occurrence of an isotropic phase is markedly favored by Mg(2+), and to a lower extent by Cd(2+), and that the pitch of the cholesteric phase decreases in the presence of Cd(2+) or Ni(2+). (23)Na NMR spectroscopy also shows differences between the binding behavior of Mg(2+) and the transition metal ions in the counterion atmosphere around DNA. The results are discussed in terms of different binding modes of the metal ions.

Melting of cross-linked DNA IV. Methods for computer modeling of total influence on DNA melting of monofunctional adducts, intrastrand and interstrand cross-links formed by molecules of an antitumor drug

A theoretical method is developed for calculation of melting curves of covalent complexes of DNA with antitumor drugs. The method takes into account all the types of chemical modifications of the double helix caused by platinum compounds and DNA alkylating agents: 1) monofunctional adducts bound to one nucleotide; 2) intrastrand cross-links which appear due to bidentate binding of a drug molecule to two nucleotides that are included into the same DNA strand; 3) interstrand cross-links caused by bidentate binding of a molecule to two nucleotides of different strands. The developed calculation method takes into account the following double helix alterations at sites of chemical modifications: 1) a change in stability of chemically modified base pairs and neighboring ones, that is caused by all the types of chemical modifications; 2) a change in the energy of boundaries between helical and melted regions at sites of chemical modification (local alteration of the factor of cooperativity of DNA melting), that is caused by all the types of chemical modifications, too; 3) a change in the loop entropy factor of melted regions that include interstrand cross-links; 4) the prohibition of divergence of DNA strands in completely melted DNA molecules, which is caused by interstrand cross-links only. General equations are derived, and three calculation methods are proposed to calculate DNA melting curves and the parameters that characterize the helix-coil transition.

Melting of cross-linked DNA. III. Calculation of differential melting curves

In our previous papers I and II (D. Y. Lando et al, J. Biomol. Struct. Dynam. (1997) v. 15, N1, p. 129-140, p. 141-150), two methods were developed for calculation of melting curves of cross-linked DNA. One of them is based on Poland's and another on the Fixman-Freire approach. In the present communication, III, a new theoretical method is developed for computation of differential melting curves of DNAs cross-linked by anticancer drugs and their inactive analogs. As Poland's approach, the method allows study of the influence of the loop entropy factor, delta(n), on melting behavior (n is the length of a loop in base pairs). However the method is much faster and requires computer time that inherent for the most rapid Fixman-Freire calculation approach. In contrast to the computation procedures described before in communications I and II, the method is suitable for computation of differential melting curves in the case of long DNA chains, arbitrary loop entropy factors of melted regions and arbitrary degree of cross-linking including very low values that occur in vivo after administration of antitumor drugs. The method is also appropriate for DNAs without cross-links. The results of calculation demonstrate that even very low degree of cross-linking alters the DNA differential melting curve. Cross-linking also markedly strengthens the influence of particular function delta(n) upon melting behavior.

Melting of cross-linked DNA: II. Influence of interstrand linking on DNA stability

In the previous paper (D.Y. Lando, J. Biomol. Struct. Dynam, 15, 129-140 (1997)) the melting of cross-linked DNA with N base pairs and omega interstrand cross-links has been considered theoretically. In the present study on the basis of these results, two simple schemes are developed for the computation of melting curves of cross-linked DNA. The investigation of influence of interstrand linking on DNA stability has been carried out by computer simulation. It is shown that the relative concentration of cross-links, CCT = omega/N, their distribution along a DNA molecule, and particular values of the entropy factors of small loops formed by cross-links in melted regions strongly affect the DNA melting temperature, Tm. On the contrary, for DNA without cross-links, a ten-fold increase or decrease in the entropy factors of small loops does not cause the Tm variation. The comparison of the results of calculation with experimental data suggests that the majority of types of cross-link neither maintain ordered parallel orientation of bases in melted regions nor increase considerably the thermostability of cross-linked base pairs. Four different ways of influence of interstrand cross-linking on the DNA double helix stability are considered. It is shown that cross-linking significantly enhances the influence of single strand stiffness in melted regions on DNA melting behavior.

Melting of cross-linked DNA: I. Model and theoretical methods

Covalent and strong coordination binding to DNA of a large number of antitumour drugs and other compounds leads to interstrand cross-link formation. To investigate cross-link influence on double helix stability, two methods are developed for the calculation of melting curves. The first method is based on Poland's approach. It requires computer time proportional to u.N, where u is the average distance (in base pairs) between neighboring cross-links and N is the number of base pairs in the DNA chain. The method is more suitable when u is not large, and small loops formed by interstrand cross-links in melted regions strongly affect DNA melting. The computer time for the second method, based on the Fixman-Freire approach, does not depend on the number of cross-links and is proportional to I.N (I is the number of exponential functions used for a decomposition of the loop entropy factor). It is more appropriate when N and u are large, and therefore particular values of the entropy factors of small loops do not influence DNA melting behavior.

Recognition of DNA adducts by human nucleotide excision repair. Evidence for a thermodynamic probing mechanism

The mechanism by which mammalian nucleotide excision repair (NER) detects a wide range of base lesions is poorly understood. Here, we tested the ability of human NER to recognize bulky modifications that either destabilize the DNA double helix (acetylaminofluorene (AAF) and benzo[a]pyrene diol-epoxide (BPDE) adducts, UV radiation products) or induce opposite effects by stabilizing the double helix (8-methoxypsoralen (8-MOP), anthramycin, and CC-1065 adducts). We constructed plasmid DNA carrying a defined number of each of these adducts and determined their potential to sequester NER factors contained in a human cell-free extract. For that purpose, we measured the capacity of damaged plasmids to compete with excision repair of a site-directed NER substrate. This novel approach showed differences of more than 3 orders of magnitude in the efficiency by which helix-destabilizing and helix-stabilizing adducts sequester NER factors. For example, AAF modifications were able to compete with the NER substrate approximately 1740 times more effectively than 8-MOP adducts. The sequestration potency decreased with the following order of adducts, AAF > UV >/= BPDE > 8-MOP > anthramycin, CC-1065. A strong preference for helix-destabilizing lesions was confirmed by monitoring the formation of NER patches at site-specific adducts with either AAF or CC-1065. This comparison based on factor sequestration and repair synthesis indicates that human NER is primarily targeted to sites at which the secondary structure of DNA is destabilized. Thus, an early step of DNA damage recognition involves thermodynamic probing of the duplex.

Orientation isomers of the mitomycin C interstrand cross-link in non-self-complementary DNA. Differential effect of the two isomers on restriction endonuclease cleavage at a nearby site

Reductively activated mitomycin C (MC) forms DNA interstrand cross-links between two guanines at CG.CG sequences. It is predictable that such cross-links should occur in two isomeric strand orientations in duplex DNA (except when located in the center of a self-complementary duplex). This was verified by the isolation and characterization of a pair of two isomeric oligonucleotides in each case of five non-self-complementary duplexes of 8-bp length, cross-linked by MC. Isomer separation was accomplished by reverse-phase HPLC. The isomers in a pair were formed in approximately 1:1 proportion. Their structures were rigorously characterized by a two-step cross-linking procedure: first, 1''-monoalkylation of each strand, followed by conversion to a cross-linked duplex by annealing the monoalkylated strand to its complement in the presence of a reducing agent. The resulting individual authentic orientation isomers were used as standards for identification of the two isomers formed in the original (one-step) cross-linking reactions. A 16-bp duplex oligonucleotide was synthesized featuring the AluI cognate sequence, separated from a MC cross-link site by only 1 bp. Its two MC cross-linked isomers were prepared separately, and their rate of cleavage by AluI was determined using HPLC. Cleavage of both the unmodified and cross-linked duplexes was nonsymmetrical. The isomer in which the 2''-NH3+ of MC is oriented toward the AluI site was cleaved essentially at the same rate as the control duplex, while cleavage of the isomer with the MC indoloquinone group oriented toward the AluI site was inhibited 2-fold at the faster-cleaved strand.(ABSTRACT TRUNCATED AT 250 WORDS)

Mechanism of action of the Escherichia coli UvrABC nuclease: clues to the damage recognition problem

During the process of E. coli nucleotide excision repair, DNA damage recognition and processing are achieved by the action of the uvrA, uvrB, and uvrC gene products. The availability of highly purified proteins has lead to a detailed molecular description of E. coli nucleotide excision repair that serves as a model for similar processes in eukaryotes. An interesting aspect of this repair system is the protein complex's ability to work on a vast array of DNA lesions that differ widely in their chemical composition and molecular architecture. Here we propose a model for damage recognition in which the UvrB protein serves as the component that confers enhanced specificity to a preincision complex. We hypothesize that one major determinant for the formation of a stable preincision complex appears to be the disruption of base stacking interactions by DNA lesions.

Recent advances in the synthesis and structure determination of site specifically psoralen-modified DNA oligonucleotides

We have developed novel methods for the preparation of multimicromole quantities of extremely pure, uniquely photoadducted psoralen-DNA cross-links, furan-side monoadducted DNA and pyrone-side monoadducts. Psoralen cross-linked and furan-side monoadducted DNA were produced by employing high intensity argon ion and krypton ion lasers as light sources. Pyrone-side monoadducts were prepared by base-catalyzed photoreversal of psoralen cross-links. The various psoralen-adducted DNA oligomers were efficiently purified by high performance liquid chromatography. These methods have permitted us to synthesize 4 mumol each of a self-complementary 8-mer d(GCGTACGC) 4'-(hydroxymethyl)-4,5',8-trimethylpsoralen (HMT) furan-side monoadduct and HMT cross-link. Preliminary nuclear magnetic resonance (NMR) data on the HMT cross-linked 8-mer d(GCGTACGC) have been obtained which confirmed the presence of the diadducted psoralen at the unique 5'TpA3' site. NMR data obtained from the 8-mer furan-side monoadduct revealed that the psoralen molecule is intercalated into the DNA double helix. Preliminary crystals of 8-mer cross-linked DNA molecule have been grown. Conditions for the growth of X-ray diffraction-quality crystals and the further analysis of these crystals are now in progress.

Phage nucleoprotein-psoralen interaction: quantitative characterization of dark and photoreactions

The irradiation of the phage T7 system containing psoralen as photosensitizer causes many processes, each of them leading to phage inactivation. These processes include the UV-induced photoreactions in the phage nucleic acid, and photoreactions in the nucleic acid sensitized by either psoralen or psoralen photobreakdown products. In addition the intercalation of the psoralen molecule itself in the phage nucleic acid as well as the psoralen photobreakdown products cause phage inactivation. Under appropriate experimental conditions these reactions can be studied and characterized separately. The quantitative characteristics (e.g. inactivation cross-section, action spectra and index for dark genotoxicity) are demonstrated for different linear and angular psoralens. Some theoretical and practical consequences of the results obtained are discussed.

Post-PCR sterilization: a method to control carryover contamination for the polymerase chain reaction

We describe a photochemical procedure for the sterilization of polynucleotides that are created by the Polymerase Chain Reaction (PCR). The procedure is based upon the blockage of Taq DNA polymerase when it encounters a photochemically modified base in a polynucleotide strand. We have discovered reagents that can be added to a PCR reaction mixture prior to amplification and tolerate the thermal cycles of PCR, are photoactivated after amplification, and damage a PCR strand in a manner that, should the damaged strand be carried over into a new reaction vessel, prevent it from functioning as a template for the PCR. These reagents, which are isopsoralen derivatives that form cyclobutane adducts with pyrimidine bases, are shown to stop Taq polymerase under conditions appropriate for the PCR process. We show that effective sterilization of PCR products requires the use of these reagents at concentrations that are tailored to the length and sequence of the PCR product and the level of amplification of the PCR protocol.

Photoactivatable antisense DNA: suppression of ampicillin resistance in normally resistant Escherichia coli

Antisense oligodeoxyribonucleotides complementary to a segment of the beta-lactamase gene and containing psoralen monoadducts at specific sites were examined for their ability to make normally resistant bacteria sensitive to ampicillin. Irradiation of oligonucleotides and psoralens with long-wavelength ultraviolet radiation (380-400 nm) produced monoadducted antisense molecules. High-performance liquid chromatography was used to purify microgram quantities of photoactivatable antisense DNA. Escherichia coli transformed with a plasmid containing the gene for beta-lactamase were used to test a series of oligonucleotides containing psoralen monoadducts after additional exposure to the photoactivating effects of long-wavelength ultraviolet radiation (320-400 nm). Normally resistant bacteria treated with this photoactivatable form of antisense DNA (0.4 microM) were specifically sensitized to ampicillin. The reduction in colony formation ranged from 31 to 79% in comparison to control oligonucleotides which did not contain photoactivatable monoadduct moieties. Bacteria treated in a similar manner but in the presence of tetracycline instead of ampicillin were not affected. The activity of beta-galactosidase, whose gene is located on the same plasmid as beta-lactamase, was not affected.

Duplex oligodeoxyribonucleotides cross-linked by mitomycin C at a single site: synthesis, properties, and cross-link reversibility

Oligodeoxyribonucleotides cross-linked by reductively activated mitomycin C (MC) were prepared and purified for the first time. The cross-linked products were structurally characterized by nucleoside and MC-nucleoside adduct analysis. Optimal conditions were established for the cross-linking reaction, resulting in high yields, typically in the 20-50% range. Nuclease digests of the cross-linked oligonucleotides yielded the same bifunctional MC-deoxyguanosine adduct as that previously isolated from DNA exposed to MC in vitro and in vivo [Tomasz et al. (1987) Science 235, 1204]. The cross-linked oligonucleotides displayed broad thermal melting profiles, greatly increased Tm, and complex circular dichroism spectra. Phosphodiester linkages at the cross-link were resistant to spleen exonuclease, nuclease P1, and TaqI and ClaI restriction endonucleases; snake venom diesterase action was uninhibited. The cross-links are stable to heat at neutral pH but are removed by treatment in hot piperidine or by the reducing agents Na2S2O4 and dithiothreitol. Mechanisms are proposed for these reactions. These studies define optimal methods for introducing mitomycin cross-links into DNA fragments at a specific site, providing a versatile tool to study the effects of the MC cross-links on DNA structure and function.

Monofunctional angular furocoumarins: sequence specificity in DNA photobinding of 6,4,4'-trimethylangelicin and other angelicins

The sequence specificity in the photoreaction (365 nm) of 6,4,4'-trimethylangelicin (TMA) with DNA fragments of the lac I gene of Escherichia coli was studied by using DNA sequencing methodology. In order to map the sites of TMA photoaddition, we took advantage of the (3'-5') exonuclease activity associated with T4 DNA polymerase, which is blocked by bulky adducts, such as furocoumarin photoadducts. A quantitative analysis of the sites of photoaddition is reported. TMA was demonstrated to photoreact with thymine and, to a lower extent, to cytosine. AT-rich sequences and TTT sites in a GC context are the most reactive sites towards TMA whereas TA, AT, CA, AC sites are weaker sites with similar reactivity. Cytosines in alternated CG sequences are also targets of TMA photobinding. We observed a less pronounced sequence specificity of TMA than that of other psoralen derivatives already studied (Sage and Moustacchi, 1987; Boyer et al., 1988). A comparison with other furocoumarins 4,4'-dimethylangelicin (4,4'-DMA), 4'-methylangelicin (4'-MA), angelicin, 4,5',8-trimethylpsoralen (TMP) and 8-methoxypsoralen (8-MOP) is also reported. The role of flanking sequence and consequently of the local conformation at the various sites of photoaddition is discussed. A preferential orientation of the TMA molecule during the intercalation in the dark is suggested. Hot alkali treatment of TMA-modified DNA did not reveal any DNA strand breakage due to photooxidized bases.

Specificity of site directed psoralen addition to RNA

We describe the attachment of a psoralen derivative (site specific psoralen, SSP) to the 5' end of a DNA oligonucleotide and the hybridization and the photoreaction of this reagent with a complementary target site on an RNA molecule. SSP was coupled to a variety of DNA oligonucleotides to investigate the structural requirements for addition to the RNA. Efficient SSP photoadducts were made on specific uridines by designing an intercalation site at an unpaired nucleotide in the RNA strand within the heteroduplex region. The optimal location for this site was five nucleotides from the oligonucleotide 5' end and just 5' to the target uridine residue. Because the attachment of the SSP to the oligonucleotide is through a disulfide bond, the DNA oligonucleotide can be removed with reduction to leave SSP attached to the RNA strand. The SSP adduct made in this way will be useful for subsequent biochemical and biophysical experiments.

UVA-induced binding of 8-methoxypsoralen to cells of Saccharomyces cerevisiae: separation and characterization of DNA photoadducts

We present methods for the determination of UVA-induced binding of 8-methoxypsoralen (8-MOP) to nucleic acids and protein and for a quantitative assay of radioactively labelled 8-MOP plus UVA induced DNA photoproducts in the yeast Saccharomyces cerevisiae. For the dose range up to 60 kJ m-2, with a wild-type survival of 1% or higher, binding to DNA is 100-fold and to RNA 10- to 20-fold more efficient than that to protein. Between 20% and 65% of the 8-MOP binds to macromolecules that are neither nucleic acids nor protein. The number of DNA-bound 8-MOP molecules for the haploid genome rises from 14 (unirradiated control) to 349 at the highest UVA exposure dose (60 kJ m-2). Gel chromatography reveals three types of DNA thymine photoproduct, the pyrone-side monoadducts, the furan-side monoadducts and the diadducts. Among these, pyrone-side monoadducts always constitute the smallest fraction, regardless of whether the treatment is with in vitro or in vivo 8-MOP plus UVA.

Determination of secondary structure in the initiation region of ovalbumin mRNA

We have analyzed the secondary structure in the region surrounding the initiation codons of both cellular and synthetic versions of ovalbumin mRNA. RNase V1 cleavage sites and structure-dependent, chemically modified bases in cellular ovalbumin mRNA were determined by reverse transcription of hen poly A(+) RNA using ovalbumin-specific, synthetic DNA primers. These results indicate an extensive region of unpaired nucleotides preceding the initiation codon and a region of base-paired nucleotides including and following the initiation codon. A synthetic ovalbumin mRNA (SP65.OV) was prepared by run-off transcription of a cloned ovalbumin cDNA (pSP65.OV). Identical regions of hen ovalbumin and SP65.OV mRNAs gave identical patterns of structure-dependent base modifications. A computer program for determining RNA secondary structure was used to find a 5'-region structure for ovalbumin mRNA that is consistent with our data.

Mutagenesis induced by site specifically placed 4'-hydroxymethyl-4,5',8-trimethylpsoralen adducts

Closed circular double stranded M13mp19 DNA containing a site-specifically placed HMT (4'-hydroxymethyl-4-5'-8-trimethylpsoralen) monoadduct or crosslink was synthesized in vitro. The damaged DNA were scored for loss of infectivity by transfection into repair proficient or deficient E. coli and into SOS induced E. coli. Mutant phages were detected by the loss of alpha-complementation between the viral and the host Lac Z genes or by the acquisition of resistance to kpn I digestion. Our results indicate that HMT mutagenesis is targeted and that deletion or transversion of the modified thymidine is the predominant sequence change elicited by a monoadduct or a crosslink. Transfection of the monoadducted DNA into a Uvr A deficient strain did not change the mutation pattern but did increase the respective mutation frequencies. Transfection of the crosslinked DNA into a SOS induced host resulted in the appearence of other types of mutations attributable to an increase in both targeted and untargeted mutations.

Evidence for structural deformation of the DNA helix by a psoralen diadduct but not by a monoadduct

We have investigated the structural change in a double-stranded DNA helix caused by covalent addition of a psoralen. A synthetic double-stranded DNA was constructed to contain either a psoralen furan-side monoadduct or an interstrand diadduct at a specific site. When the unmodified and psoralen modified DNAs were examined by electron microscopy in the presence of distamycin, which stiffens the DNA helix, the DNA containing the psoralen interstrand diadduct appeared bent (or kinked), whereas the furan-side monoadducted DNA appeared similar to the unmodified DNA. RecA protein from E. coli has been shown to preferentially bind UV (ultra violet) irradiated DNA presumably due to alterations in the normal DNA helical structure. Using a nitrocellulose filter binding assay, we have found that the psoralen interstrand diadduct enhances the binding of recA protein to the double-stranded DNA, whereas a furan-side monoadduct has little effect. Thus both the recA protein binding and the electron microscopic data suggest that a psoralen diadduct causes deformation of a DNA helix, most likely by kinking the helix, and that a monoadduct has little effect on the DNA helix structure.

Photoreactivities and thermal properties of psoralen cross-links

We have studied the photoreaction of 8-methoxypsoralen (8-MOP), 4,5',8-trimethylpsoralen (TMP), and 4'-(hydroxymethyl)-4,5',8-trimethylpsoralen (HMT) with a pair of 18-base-long oligonucleotides in which a 14-base region is complementary. Only one 5'TpA site, favored for both monoadduct and cross-link formation with psoralen, is present in this oligonucleotide pair. We have used this model system to demonstrate, for the first time, strand specificity in the photoreaction of psoralen with DNA. We found that the two types of cross-links which form at this site have large differences in thermal stabilities. In addition, the denaturation of each cross-link isomer duplex occurred in at least three stages, which can be visualized as three bands in thermal equilibrium under the conditions of a denaturing polyacrylamide gel. This novel observation suggests that there are several domains differing in thermal stability in a psoralen cross-link.

Interaction of Escherichia coli RNA polymerase with DNA in an elongation complex arrested at a specific psoralen crosslink site

We have probed the interaction of Escherichia coli RNA polymerase with DNA in an elongation complex arrested by a site-specifically placed psoralen crosslink using DNase I footprinting techniques. The psoralen derivative 4'-hydroxymethyl-4,5',8-trimethylpsoralen was first placed at a specific site in the middle of a chemically synthesized double-stranded DNA fragment containing an E. coli RNA polymerase promoter at one end. The psoralen molecule was photochemically attached to two adjacent thymidine residues on opposite strands as a diadduct. Using this crosslinked DNA as the template for transcription, we found that the E. coli RNA polymerase was blocked at the psoralen diadduct, yielding a transcript 29 nucleotides long. The arrested elongation complex inhibited DNase I digestion of both the coding strand and the non-coding strand from about 22 nucleotides upstream to 15 nucleotides downstream from the diadduct. These results, which suggest that the unwindase and the catalytic sites of the polymerase are very close to each other, have been incorporated into a model of the transcription elongation complex.

Structure of a psoralen cross-linked DNA in solution by nuclear magnetic resonance

One- and two-dimensional nuclear magnetic resonance (NMR) methods were used to determine a three-dimensional model of an eight-base-pair DNA fragment (d-GGGTACCC) cross-linked with psoralen in solution. Two-dimensional nuclear Overhauser effect experiments were used to assign the spectrum and estimate distances for 171 proton pairs in the cross-linked DNA. The NMR-derived model shows a 53 degree bend into the major groove that occurs primarily at the site of drug addition and a 56 degree unwinding that spans the eight-base-pair duplex.

Solution hybridization of crosslinkable DNA oligonucleotides to bacteriophage M13 DNA. Effect of secondary structure on hybridization kinetics and equilibria

Several DNA oligonucleotides have been photochemically modified with the furocoumarin 4'-hydroxymethyl-4,5',8-trimethylpsoralen (HMT) such that each contained a single HMT furan side monoadduct to thymidine at a unique 5' TpA 3' sequence. When these oligonucleotides were hybridized to their respective complements, the HMT adduct could be driven to form an interstrand crosslink by irradiation of the hybrid with 360 nm light. The ability to crosslink probe-target complexes has allowed us to determine the kinetics and the extent of hybridization in solution between these oligonucleotides and their complementary sequences in single-stranded bacteriophage M13 DNA. Our data indicate that these parameters are strongly influenced by the existence of local as well as global secondary structure in the viral DNA. During hybridization, rearrangement of this secondary structure so as to expose the target sequence can be rate-limiting. Upon attainment of equilibrium, only a portion of the target sequence may be hybridized to the probe with the remainder involved in intrastrand base-pairing. Using crosslinkable oligonucleotide probes hybridized and irradiated near the melting temperature of the respective probe-target complex one can partially overcome these secondary structure effects.

The effects of covalent additions of a psoralen on transcription by E. coli RNA polymerase

Synthetic DNA substrates containing a site-specifically engineered psoralen monoadduct or diadduct were used to characterize the response of the E. coli RNA polymerase elongation complex to these lesions. The psoralen derivative HMT (4'-hydroxymethyl-4,5', 8-trimethylpsoralen) was site specifically placed into two synthetic double-stranded DNA fragments each of which contained an E. coli RNA polymerase promoter at one end. The HMT molecule was attached to the middle of the DNA fragments as either a furan-side monoadduct or an interstrand diadduct. Transcription off the HMT crosslinked DNA templates showed that E. coli RNA polymerase terminated at the HMT diadduct site, i. e., one nucleotide before the modified thymidine residue on the template strand. The furan-side monoadduct when on the template strand also blocked transcription by the polymerase. However, no effect on transcription was observed when the monoadduct was located on the non-template strand.

Reverse Southern hybridization

A DNA oligomer 25 nucleotides long which contained an HMT (4'-hydroxymethyl-4,5', 8-trimethylpsoralen) furan side monoadduct to thymidine at a 5'-TpA-3' site was used as a probe for the polylinker sequence present in single-stranded M13 mp19 DNA and in double-stranded pUC 19 DNA. Hybridization and photofixation were carried out simultaneously in solution under conditions approximating the melting temperature of the probe-target hybrid. Use of probe concentrations greater than 10(-8) M permitted hybridization times of a few minutes. Irradiation with near ultraviolet light converted the HMT monoadduct present in hybrid complexes into an interstrand crosslink. Efficient photofixation removed hybrid from the equilibrium distribution and resulted in the formation of additional probe-target complex. After removal of excess probe by centrifugation through a semi-permeable membrane (Centricon-30), samples were electrophoresed through an alkaline agarose gel which was analyzed by autoradiography. When using an HMT-modified 25-mer probe end-labeled with 3,000 Ci/mmole 32P, 0.015 ng (3.8 X 10(6) copies) of M13 DNA could be detected. With this same probe 10 micrograms of denatured human DNA (corresponding to 3.0 X 10(6) copies) did not give a signal.