Adenine photodimerization in deoxyadenylate sequences: elucidation of the mechanism through structural studies of a major d(ApA) photoproduct.

The mechanism of the photodimerization of adjacent adenine bases on the same strand of DNA has been elucidated by determining the structure of one of the two major photoproducts that are formed by UV irradiation of the deoxydinucleoside monophosphate d(ApA). The photoproduct, denoted d(ApA)*, corresponds to a species of adenine photodimer first described by P?rschke (P?rschke, D. (1973) J.Am.Chem.Soc. 95, 8440-8446). From a detailed examination of its chemical and spectroscopic properties, including comparisons with the model compound N-cyano-N1-(1-methylimidazol-5-yl)formamidine, it is deduced that d(ApA)* contains a deoxyadenosine unit covalently linked through its C(8) position to C(4) of an imidazole N(1) deoxyribonucleoside moiety bearing an N-cyanoformamidino substituent at C(5). On treatment with acid, d(ApA)* is degraded with high specificity to 8-(5-amino-imidazol-4-yl)adenine whose identity has been confirmed by independent chemical synthesis. It is concluded that the primary event in adenine photodimerization entails photoaddition of the N(7)-C(8) double bond of the 5'-adenine across the C(6) and C(5) positions of the 3'-adenine. The azetidine species thus generated acts as a common precursor to both types of d(ApA) photoproduct which are formed from it by competing modes of azetidine ring fission.     

Ultraviolet-induced 8,8-adenine dehydrodimers in oligo- and polynucleotides.

Characteristic fluorescence excitation and emission is induced by either acetone-sensitized 313 nm irradiation of mixtures of 8-bromoadenosine and adenosine or 254 nm irradiation of oligo- and polynucleotides containing adenine neighbors. The acetone-sensitized reaction involves cleavage of bromine from 8-bromoadenosine with activation of C-8, leading to formation of an 8,8-adenosine dehydrodimer. Comparable fluorescence properties arise in the unsensitized photoreaction of dApdA, pdApdA, ApA, poly(dA), poly(A), poly(dA.dT), and poly(dA.U). The previously unidentified adenine ultraviolet photoproduct described by Porschke has been isolated as several variants from solutions of pdApdA and poly(dA) irradiated at 254 nm. Based upon fluorescence spectra and mass spectra, these variants are shown to contain the 8,8-adenine dehydrodimer moiety.     

A proton magnetic resonance study of the interaction of adenylyl (3' is greater than 5') adenosine with polyuridylic acid

The interaction of adenylyl (3′ → 5′) adenosine (ApA) with polyuridylic acid in D₂O solution at neutral pD has been studied by high resolution proton magnetic, resonance spectroscopy. At temperatures above ～32°C, no evidence was obtained for the interaction of ApA with poly U. Below this temperature, a rigid triple-stranded complex involving a stoichiometry of 1 adenine to 2 uracil bases is formed, presumably via specific adenine–uracil base-pairing and cooperative base stacking of the adenine bases in a manner similar to that previously reported for the adenosine–poly U complex.     

The photochemistry of d(T-A) in aqueous solution and in ice.

When d(T-A) is irradiated at 254 nm in aqueous solution an internal photoadduct is formed between its constituent adenine and thymine bases. The resultant photoproduct, designated TA*, arises from a singlet excited state precursor; a similar photoreaction is not observed with d(C-A) or d(T-G). In contradistinction, irradiation of d(T-A) in frozen aqueous solution yields a dimeric photoproduct in which two d(T-A) molecules are coupled together by a (6-4) photoadduct linkage between their respective thymine bases. Both photoproducts have been extensively characterised by a combination of electron impact and fast atom bombardment mass spectrometry, UV, CD, 1H NMR and fluorescence spectroscopy. Acid treatment of TA* gives 6-methylimidazo[4,5-b]pyridin-5-one whose identity was established by an independent chemical synthesis involving photorearrangement of 6-methyl-imidazo[4,5-b]pyridine N(4)-oxide. A tentative mechanism is presented to account for the acid degradation of TA*. The structure of the dimeric ice photoproduct follows from its cleavage, by snake venom phosphodiesterase, to 5'-dAMP and the (6-4) bimolecular photoadduct of thymidine; on acid hydrolysis it gives adenine and 6-(5'-methyl-2'-oxopyrimidin-4'-yl) thymine.     

The reaction of 14C-labelled platinum ethylenediamine dichloride with adenine compounds and DNA

Various derivatives of adenine have been studied with regard to their rate of reaction with ¹⁴C-labelled platinum ethylenediamine dichloride, Pt(¹⁴C-en)Cl₂. The reactivities have been calculated from the “rate of disappearance” of Pt(¹⁴C-en)Cl₂ using chromatographic separation of reactants and products.Adenine and adenosine react very slowly at 37° whereas other adenine derivatives react much more readily in the order: poly A > AMP > ApA > poly d(AT). From the numerical values of the rate constants it is concluded that the presence of a phosphate group increases the reaction rate considerably. This is partly the explanation of the rapid reaction of poly A which possesses terminal phosphate groups. However adjacent adenine moieties such as those in polyadenylic acid (poly A) and adenosyl-3′5′-adenosine (ApA) may also react by another mechanism which involves the 6-NH₂ groups.The energies of activation of the second order reaction with platinum ethylenediamine dichloride (PtenCl₂) are 12.9, 18.8, 19.0 kcal/mole for poly A, AMP and ApA respectively.In DNA, no free phosphate groups are present, and the occurrence of adjacent adenines will be low. The reaction of PtenCl₂ with DNA seems to involve a rapid attack on deoxyguanosine (GdR) and a slow reaction with deoxyadenosine (AdR) and deoxycytidine (CdR).     

Search for an adenine photoproduct in DNA.

Poly(d[14C]A), p(dA)2, and [14C]adenosine-labeled DNA were irradiated at 254 nm with fluences up to 50 J/m2, and then following formic acid hydrolysis at 170 degrees C WERE SUBJECTED TO PAPER CHROMAtography using a butanol:water:acetic acid (80:30:12) solvent system. For poly(dA), up to 25% of the radioactivity appeared as fluorescent material located in the Rf 0.21-0.29 region. The hydrolysate of the purified photoproduct, p(dA)2, isolated from irradiated p(dA)2 by DEAE chromatography also had an Rf of 0.29 as well as an absorbance maximum at 310 nm. In all cases studied, however, the photoproduct yield in the Rf 0.29 region for native DNA was less than 2%. Denaturation of the DNA appeared to enhance the yield slightly, although no pronounced peak in this region of the chromatogram was discerned. Mechanistic studies indicate that the yield of the adenine photoproduct in poly(dA) is favored by base stacking, has a singlet excimer as a precursor, and is quenched by hydrogen bonding to a pyrimidine. It is concluded that the yield of the adenine photoproduct in both native and denatured DNA is considerably less than in poly (dA) and in all probability does not represent a biologically significant product.     

The photoreactivity of T-A sequences in oligodeoxyribonucleotides and DNA.

Photoaddition between adjacent adenine and thymine bases occurs, with a quantum yield of approximately 5 X 10(-4) mol einstein-1, when d(T-A), dT-A, d(pT-A), d(T-A-T), d(T-A-T-A) and poly(dA-dT) are irradiated, at 254 nm, in aqueous solution. The photoadduct thus formed is specifically degraded by acid to the fluorescent heterocyclic base 6-methylimidazo[4,5-b]pyridin-5-one (6-MIP) with retention of C(8) of adenine and the methyl group of thymine. This reaction, coupled with either spectrofluorimetric or radiochemical assay of 6-MIP isolated by high voltage paper electrophoresis, has been used to demonstrate formation of the adenine-thymine photoadduct on UV irradiation of poly(dA-dT).poly(dA-dT) and both native and denatured DNA from calf thymus and E. coli. Estimated quantum yields for this new type of photoreaction in DNA show that it is substantially quenched by base pairing. Possible biological implications of the photoreaction are discussed.     

The binding of Ni2+ to adenylyl-3',5'-adenosine and to poly(adenylic acid)

Studies of the binding of Ni²⁺ to adenylyl-3',5'-adenosine (ApA) at pH 6-0 by ultraviolet spectrophotometry indicate the formation of a 1:1 complex in the presence of a large excess of metal ion. At 25 °C. and ionic strength μ = 0.5 M, the stability constant of Ni(ApA) is evaluated to be K = 2.6 (±0.6) M^?1. The low stability is taken as evidence that the predominant complex species is one in which the ApA acts as a monodentate ligand, mainly through the adenine group. The rate constants for complex formation and dissociation, k_f = 1430 M^?1 s^?1 and k_b = 665 s^?1 (25°C. μ = 0.5M). determined by the temperature-jump relaxation technique, are consistent with this interpretation. The binding strength of Ni²⁺ to poly(adenylic acid) [poly(A)] has been studied at pH 7.0 using murexide as an indicator of the concentration of free Ni²⁺. Within the concentration range [Ni²⁺ = 1 × 10^?5 × 10^?3 M the data can be represented in the form of a linear Scatchard plot. i.e., the process can be described as the binding of Ni²⁺ to one class of independent binding sites. The number of binding sites per monomer is 0.26, and the stability constant K = 8.2×10³ M^?1 (25°C μ = 0.1 M). In kinetic studies of the reaction of Ni²⁺ with poly(A), two relaxation effects due to complex formation were detected, one with a concentration-independent time constant of about 0.4 ms, the other with a concentration-dependent time constant in the millisecond range. The concentration dependence of the longer relaxation time can be accounted for by a three-step mechanism which consists of a fast second-order association reaction followed by two first-order steps. There is evidence, however, that the overall process is more complicated than expressed by the three-step mechanism.     

Conformational properties of adenylyl-3' leads to 5'-adenosine in aqueous solution.

A detailed 220-MHz NMR study has been made of the conformational properties for the homodinucleotide adenylyl-3' leads to 5'-adenosine, ApA, in D2O. Unambiguous signal assignments of all proton signals were made with the aid of selectively deuterated nucleotidyl units, ApA, ApA, and D-8ApA, and complete, accurate sets of NMR parameters were derived by simulation-iteration methods. Sets of limiting chemical shifts and coupling values were also obtained for ApA and constituent monomers 3'-AMP and 5'-AMP at infinite dilution and at identical ionization states for assessment of dimerization effects. Conformational properties were evaluated quantitatively for most of the conformational bonds of ApA and these are consistent with two compact folded dynamically averaged structures, a base-stacked right helical structure, I, characterized as anti, C3'-endo, g-, w,w' (320,330 degrees), g'g', gg, C3'-endo, anti, and a more loosely base-stacked loop structure, II, with anti, C3'-endo, g-, w,w' (80 degrees, 50 degrees), g'g', gg, C3'-endo, anti orientations. Dimerization produces a number of nucleotidyl conformational changes including a shift in ribose equilibrium C2'-endo (S) in equilibrium C3'-endo (N) in favor of C3'-endo in both Ap- and -pA (60:40 vs. 35:65 in monomers), a change in glycosidic torsion angle chiCN toward 0 degrees, and a greater locking-in of rotamers along bonds involved in the phosphodiester backbone. Moreover, there is clear evidence that the transitions from S leads to N forms and chiCN leads to 0 degrees are directly related to base stacking in ApA. Finally, ApA exists in solution as an equilibrium between I, II and an unstacked form(s) with as yet undetermined conformational features. Since C4'-C5', C5'-O5', and C3'-O3' bonds possess exceptional conformational stabilities, it is proposed that destacking occurs primarily by rotation about P-O5' and/or O3'-P. Predominant factors influencing the overall ApA conformation are thus base-base interaction and flexibility about P-O5' and O3'-P, with change of ribose conformation occurring in consequence of an alteration of chiCN, the latter in turn being governed by the need for maximum eta overlap of stacked adenine rings.     

High-resolution crystal structure of the intramolecular d(TpA) thymine-adenine photoadduct and its mechanistic implications

A high-resolution crystal structure is reported for d(TpA)*, the intramolecular thymine–adenine photoadduct that is produced by direct ultraviolet excitation of the dinucleoside monophosphate d(TpA). It confirms the presence of a central 1,3-diazacyclooctatriene ring linking the remnants of the T and A bases, as previously deduced from heteronuclear NMR measurements by Zhao et al. (The structure of d(TpA)*, the major photoproduct of thymidylyl-(3′-5′)-deoxyadenosine. Nucleic Acids Res., 1996, 24, 1554–1560). Within the crystal, the d(TpA)* molecules exist as zwitterions with a protonated amidine fragment of the eight-membered ring neutralizing the charge of the internucleotide phosphate monoanion. The absolute configuration at the original thymine C5 and C6 atoms is determined as 5S,6R. This is consistent with d(TpA)* arising by valence isomerization of a precursor cyclobutane photoproduct with cis–syn stereochemistry that is generated by [2 + 2] photoaddition of the thymine 5,6-double bond across the C6 and C5 positions of adenine. This mode of photoaddition should be favoured by the stacked conformation of adjacent T and A bases in B-form DNA. It is probable that the primary photoreaction is mechanistically analogous to pyrimidine dimerization despite having a much lower quantum yield.     

Adenine-adenine interaction in adenylyladenosine and polyadenylic acid measured with emission spectroscopy

Dynamical behavior of a melting of stacking interaction in adenylyladenosine (ApA) and polyadenylic acid (polyA) has been investigated with steady state and time resolved emission spectroscopy. Temperature dependence of monomer-like emission and excimer emission shows that the melting behavior of ApA can be analyzed by a simple two-state model whereas that of polyA exhibits the influence of neighboring adenine molecules which can interact in the excited state.     

An efficient synthesis of acyclic N7- and N9-adenine nucleosides via alkylation with secondary carbon electrophiles to introduce versatile functional groups at the C-1 position of acyclic moiety

The introduction of versatile functional groups, allyl and ester, at the C-1 position of the acyclic chain in acyclic adenine nucleosides was achieved for the first time directly by alkylation of adenine and N6-potected adenine. Thus, the C-1'-substituted N9-adenine acyclic nucleoside, adenine-9-yl-pent-4-enoic acid ethyl ester (11), was prepared by direct alkylation of adenine with 2-bromopent-4-enoic acid ethyl ester (6), while the corresponding N7-regioisomer, 2-[6-(dimethylaminomethyleneamino)-purin-7-yl]-pent-4-enoic acid ethyl ester (10), was obtained in one step by the coupling of N, N-dimethyl-N'- (9H-purin-6-yl)-formamidine (9) with 2-bromopent-4-enoic acid ethyl ester (6). The functional groups, ester and allyl, were converted to the desired hydroxymethyl and hydroxyethyl groups, and subsequently to phosphonomethyl derivatives and corresponding pyrophosphorylphosphonates.     

Halogenated nucleic acids: effects of 5-fluorouracil on the conformation and properties of a polyribonucleotide and its constituents

The conformational properties of 5-fluorouracil derivatives are compared to uracil derivatives. FUrd, 5′-FUMP, and poly(FU) are studied as a function of pH and temperature by ¹⁹F- and ¹H-nmr spectroscopy, and the corresponding uracil derivatives by ¹H-nmr spectroscopy. FUrd exhibits no significant conformational changes with solution pH (5–10). In contrast, at low pH (6–7) 5′-FUMP and 5′-UMP show similar conformational features, while at high pH (9) 5′-FUMP shows significant conformational alterations. Also, poly(U) and poly(FU) are conformationally similar at low pH, but increasing pH induces changes in poly(FU). These changes are observed in the backbone [γ(C4′-C5′)], furanose, and furanose-base conformations. The apparent pK_a of N3-H ionization of the FUra base is determined by ¹H- and ¹⁹F-nmr to range from 7.5 to 8.2 [FUrd < 5′-FUMP < 5′-FUDP < poly(FU)]. These observations are interpreted as a result of electrostatic interactions generated between the ionized phosphate group and the negatively charged base moiety as the pH is raised. The interaction properties of poly(FU) with ApA are studied by ¹H- and ¹⁹F-nmr spectroscopy, and these properties compared to those published for poly(U). Poly(FU) forms a complex with ApA inducing upfield ¹H-shifts in both components, and downfield ¹⁹F- shifts in poly(FU). The base stoichiometry of the complex for poly(U)·ApA is 2U:1A at various U/A ratios. In contrast, the base stoichiometry of the poly(FU)·ApA complex appears to be dependent on the FU/A ratio. At high FU/A ratio, the complex is 2FU:1A, and as the FU/A ratio approaches unity the complex becomes 1FU:1A.     

The solution structure of the intramolecular photoproduct of d(TpA) derived with the use of NMR and a combination of distance geometry and molecular dynamics.

One and two dimensional NMR techniques have been used together with molecular modelling to obtain the solution structure for the photoproduct d(TpA)*. The NMR data confirm that the cyclobutane linkage is formed between the bonds thymine C6-C5 and adenine C5-C6. The 2D NOE data are used as constraints in a distance geometry calculation. The structures obtained show a trans-syn cyclobutane linkage and the glycosidic angles are SYN and ANTI for thymidine and deoxyadenosine, respectively. The coupling constant data are used to check the backbone torsion angles of the obtained structures. Typical torsion angles are a gamma+ and beta t for the deoxyadenosine residue. A free molecular dynamics simulation of a trans-syn d(TpA) photoproduct confirmed all these structural characteristics.     

Novel non-specific DNA adenine methyltransferases

The mom gene of bacteriophage Mu encodes an enzyme that converts adenine to N(6)-(1-acetamido)-adenine in the phage DNA and thereby protects the viral genome from cleavage by a wide variety of restriction endonucleases. Mu-like prophage sequences present in Haemophilus influenzae Rd (FluMu), Neisseria meningitidis type A strain Z2491 (Pnme1) and H. influenzae biotype aegyptius ATCC 11116 do not possess a Mom-encoding gene. Instead, at the position occupied by mom in Mu they carry an unrelated gene that encodes a protein with homology to DNA adenine N(6)-methyltransferases (hin1523, nma1821, hia5, respectively). Products of the hin1523, hia5 and nma1821 genes modify adenine residues to N(6)-methyladenine, both in vitro and in vivo. All of these enzymes catalyzed extensive DNA methylation; most notably the Hia5 protein caused the methylation of 61% of the adenines in λ DNA. Kinetic analysis of oligonucleotide methylation suggests that all adenine residues in DNA, with the possible exception of poly(A)-tracts, constitute substrates for the Hia5 and Hin1523 enzymes. Their potential 'sequence specificity' could be summarized as AB or BA (where B = C, G or T). Plasmid DNA isolated from Escherichia coli cells overexpressing these novel DNA methyltransferases was resistant to cleavage by many restriction enzymes sensitive to adenine methylation.     

Identification and characterization of a Treponema pallidum subsp. pallidum gene encoding a DNA adenine methyltransferase

The nucleotide sequence of a DNA adenine methyltransferase gene (dam) from Treponema pallidum has been determined. Southern blot analysis of T. pallidum chromosomal DNA indicated that this gene is present as a single copy. The dam gene encodes a 303 amino acid protein whose deduced sequence has significant homology with DNA (N⁶-adenine) methyltransferases. T. pallidum Dam can be assigned to group α DNA amino methyltransferases based on the order of nine conserved motifs that are present in the protein. Digests of T. pallidum chromosomal DNA performed with isoschizomer restriction endonucleases (Sau3AI, DpnI, and MboI) confirmed the presence of methylated adenine residues in GATC sequences (Dam⁺ phenotype).     

Study of the interaction of polynucleotide chains with oligomers by means of chromatography. II. Effect of magnesium ions and noncomplementary bases on the stability of complexes

The interaction of the oligonucleotides ApA, ApApA, ApApC and ApApU with poly(U) and (Ip)₅I and (Ip)₆ with poly(C) has been studied by means of equilibrium gelfiltration through Sephadex.From sorption isotherms the free energies, energies and entropies of complexing have been computed for different concentrations of magnesium ions in the medium.The stoichiometric ratio of polymers to oligomers has been measured and found equal to 2 in the case of ApApA and ApApC. This shows that the cytidylic acid residue is included in the ternary complex. But in the case of ApApU the noncomplementary base is partly squeezed out of the complex.The stacking free energy of neighbouring oligomers has been found to be in the range 1000–3000 $\text{cal}\text{mole}$ depending on the conditions.The stoichimetric ratio has been found to be 1 in the case of poly(C): oligo(I), the stacking energy is equal to 1.2 $\text{kcal}\text{mole}$. The effect of magnesium is somewhat different in the case of double and triple helices and probably reflects the formation of coordination compounds with the nitrogen bases of nucleotides.     

Sequence and conformational effects on imino proton exchange in A.T- and A.U-containing DNA and RNA duplexes

¹H-nmr relaxation has been used to study the effect of sequence and conformation on imino proton exchange in adenine–thymine (A · T) and adenine–uracil (A · U) containing DNA and RNA duplexes. At low temperature, relaxation is caused by dipolar interactions between the imino and the adenine amino and AH2 protons, and at higher temperature, by exchange with the solvent protons. Although room temperature exchange rates vary between 3 and 12s^?1, the exchange activation energies (E_α) are insensitive to changes in the duplex sequence (alternating vs homopolymer duplexes), the conformation (B-form DNA vs A-form RNA), and the identity of the pyrimidine base (thymine vs uracil). The average value of the activation energy for the five duplexes studied, poly[d(A-T)], poly[d(A) · d(T)], poly[d(A-U)], Poly[d(A) · d(U)], and poly[r(A) · r(U)], was 16.8 ± 1.3 kcal/mol. In addition, we find that the average E_α for the A.T base pairs in a 43-base-pair restriction fragment is 16.4 ± 1.0 kcal/mol. This result is to be contrasted with the observation that the E_α of cytosine-containing duplexes depends on the sequence, conformation, and substituent groups on the purine and pyrimidine bases. Taken together, the data indicate that there is a common low-energy pathway for the escape of the thymine (uracil) imino protons from the double helix. The absolute values of the exchange rates in the simple sequence polymers are typically 3–10 times faster than in DNAs containing both A · T and G · C base pairs.     

Calorimetric study of 2U:1A three-stranded complexes formed between poly(U) and adenine dinucleotides: ApA and diastereoisomers of nonionic adenine dideoxyribonucleoside methyl phosphonate

Melting parameters of 2U:1A complexes formed by polyuridylic acid [poly(U)] and three adenine dinucleotides, diribonucleoside monophosphonate ApA and diastereoisomers of dideoxyribonucleoside methyl phosphonate [(dApA)₁ and (dApA)₂], in 1M NaCl and at a number of dinucleotide concentrations were obtained from differential scanning microcalorimetric data and interpreted in terms of the theory of helix–coil equilibrium in oligonucleotide–polynucleotide systems. The apparent binding constant, 1/c_m, at 39°C and melting temperatures, T_m, at 1 × 10^?3 M dinucleotide concentration indicate the following order of thermodynamic stability of the complexes: 2 poly(U) · (dApA)₂ (2.27 × 10³M^?1, 44.2°C) > 2 poly(U) · (dApA)₁ (9.9 × 10²M¹, 39.2°C) > 2 poly(U) · (ApA) (5.9 × 10²M^?1, 35.8°C). Corresponding calorimetric enthalpies of melting, ΔH_m: 13.5, 12.7, and 12.8 kcal/mol (UUA base triplets) were found to be considerably lower than the van't Hoff enthalpies, ΔH_app: 29.4, 16.2, and 16.2 kcal/mol, respectively, evaluated from the dependence of the melting temperatures on dinucleotide concentration. Self-association of dinucleotides and their simultaneous binding as monomers, dimers, and higher-order associated species is suggested as the most probable cause of the differences between ΔH_m and ΔH_app values. The differences in thermodynamic properties of the complexes formed by (dApA)₁ and (dApA)₂ diastereoisomers are discussed in connection with their known conformational properties. The higher and essentially enthalpic stability of the 2 poly(U) · (dApA)₂ complex correlates with a lower degree of intramolecular stacking of the (dApA)₂ isomer. The hydrophobically enhanced strong self-association of the latter greatly influences the thermodynamics of its complex formation with poly(U) and results in ΔH_app/ΔH_m = 2.3.     

Solution-state structure of the Dewar pyrimidinone photoproduct of thymidylyl-(3'----5')-thymidine

The preparation, spectroscopic investigation, structure determination, conformational analysis, and modeling of the Dewar pyrimidinone photoproduct of thymidylyl-(3'----5')-thymidine, previously referred to as TpT3 [Johns, H. E., Pearson, M. L., LeBlanc, J. C., & Heilleiner, C. W. (1964) J. Mol. Biol. 9, 503-524], is described. TpT3 was prepared in quantitative yield by photolysis of an aqueous solution of the (6-4) photoproduct of TpT with Pyrex-filtered medium-pressure mercury arc light. TpT3 was analyzed by FAB MS, IR, UV, and 1H, 13C, and 31P NMR spectroscopy. The spectroscopic data led to the conclusion that TpT3 results from the photoisomerization of the pyrimidinone ring of the (6-4) product of TpT to its Dewar valence isomer. Torsion angle and interproton distance information derived from coupling constants and NOE data was used to constrain ring conformation searches by utilizing the SYBYL molecular modeling program subroutine SEARCH. Sets of angles derived from the ring search procedure were then used to construct structures whose geometries were optimized by the energy-minimization subroutine MAXIMIN. A two-state model for the solution-state structure of the Dewar photoproduct was chosen which was energetically sound, fit the experimental coupling constants with an RMS deviation of 1.15 Hz, and was consistent with the NOE data. The model for the Dewar photoproduct was compared to a model for the (6-4) photoproduct and the TpT subunits of the Dickerson dodecamer structure by a least-squares fitting procedure. It was concluded that the Dewar photoproduct more closely resembles a B-form TpT unit than does the (6-4) photoproduct.