Jerry Donohue

Base pairing in DNA.

On the basis of recent structural information which extends somewhat the range of possible dimensions for bases and base pairs, five new geometrically acceptable hydrogen-bonded base pairs which might be formed by the bases in nucleosides have been suggested. The dimensions of one of these new pairs, adenine-guanine, are such that this pair would apparently fit in the Watson-Crick DNA helix. The evidence cited by Spencer against the possible existence in DNA of a pairing adenine-thymine guanine-eytosine which, differs from that formulated by Watson & Crick has been examined and is considered to be inconclusive. Particular attention has been given to the conformational relationship between the base and the sugar, and a torsional angle, (φCN, has been defined. This is the angle between the trace of the plane of the base and the projection of the C-O bond of the furanose ring when viewed along the glycosidic C-N bond. Values of this angle in known structures and proposed models have been tabulated, and the implications concerning possible conformations are considered.

Crystal structure studies of amino acids and peptides.

Publisher Summary This chapter discusses the crystal structure studies of amino acids and peptides. It describes the structures, as elucidated by single-crystal X-ray investigations, of the crystalline amino acids and peptides. The chapter also deals with the powers and limitations of present-day X-ray diffraction techniques. This chapter illustrates that in earlier times only two amino acids and DL-alanine, one Dipeptide, and one cyclic dipeptide has been investigated, and these by X-ray diffraction techniques that were crude by present-day standards. Moreover, the conclusions drawn from these early investigations—that resulted in the formulation of the stable configurations of polypeptide chains—including the a-helix and the pleated sheet and in the derivation of detailed structures for many fibrous proteins are valid. Moreover, the advent of high-speed digital computers has been responsible for a tremendous improvement in the accuracy of crystal-structure investigations; moreover, it has resulted in a parallel increase in the complexity of the molecules.

On NH···S hydrogen bonds

Abstract A survey of crystals in which NH···S interactions occur leads to the result that most of these systems should be called hydrogen bonds in the sense that this term is usually used. The NH···S distances are in the range 3.25 to 3.55 A. More important, however, are the observed shortenings of 0.2 to 0.5 A of the H···S distance under the van der Waals radius sum of 2.75 A. Small deviations from linearity, i.e. up to about 25 °, of the three atoms involved, may be expected. The possibility of NH···S bonds accordingly should not be overlooked in the construction of models of molecules in which both NH groups and sulfur are present.

Fundamental dimensions of the bases in nucleic acids

Abstract Revised dimensions, together with their standard errors, for the purine and pyrimidine bases of DNA and RNA have been derived from the results of 21 recent crystal structure determinations. These new dimensions are more precise than those which have been published previously. It is pointed out that the two purines do not have identical dimensions in their rings, nor do the two pyrimidines. It is also pointed out that in many of the pairs of external bond angles the two are not equal, contrary to previous assumptions.

Andrographolide: an X-ray crystallographic analysis

The structure of andrographolide, isolated fromAndrographis paniculata Nees, has been established by means of a single crystal X-ray analysis. The crystals have the space groupP21, witha=6.550(1),b=8.005(2),c=17.991(7) Å,β=97.36(2)°, andZ=2. The structure was refined toR=7.3%. The molecular stereochemistry, bond distances, bond angles, and hydrogen bonding scheme have all been determined. The systematic name is 3-[2-[decahydro-6-hydroxy-5-(hydroxymethyl)-5,8a-dimethyl-2-methylene-1-naphthalenyl]ethylidine]dihydro-4-hydroxy-2(3H)-furanone.

Fourier Series and Difference Maps as Lack of Structure Proof: DNA Is an Example

duce an average purine-pyrimidine pair for DNA. The atomic form factors are from International Tables for Crystallography; they are multiplied by weighting factors of 1⁄2/2 or 1⁄44 where appropriate. The temperature factor has B =6A2 since this is more appropriate for DNA. (Those familiar with the diffraction from DNA models are aware that much of the sparseness of DNA data with periodicities less than 3 A is not due only to attenuation factors that increase continuously with diffraction angle but to the fact that the molecular transform itself is small between 3 and 2 A.) However we have repeated the experiments with B = 15 A and 30 A and reach the same conclusions as when B = 6 A2. 9. M. H. F. Wilkins, in Biological Structure and Function (Academic Press, New York, 1960), vol. 1, pp. 13-32. 10. D. A. Marvin, M. H. F. Wilkins, L. D. Hamilton, Acta Cryst. 20, 663 (1966).

The structure of β-uranium

During a review of the various structures of uranium, one of us discovered indexing errors that affected a number of the investigations of the β allotrope. These errors have been corrected, and least-squares refinements performed in the three possible space groups, P4_2/mnm, P4_2nm, and P4n2, with resulting R_I values of 0.28, 0.24, and 0.28, respectively. It is concluded that the β-uranium powder data (Thewlis, 1952) cannot be used to determine the correct space group.

Crystal structures of two local anesthetics: Dibucaine·HCl·H2O and Dimethisoquin·HCl·H2O

The crystal structures of the tertiary amine local anesthetics dibucaine hydrochloride monohydrate (C20H29N3O2·HCl·H2O) and dimethisoquin hydrochloride monohydrate (C17H24N2O·HCl·H2O) have been determined. Dibucaine forms crystals inP21/c witha=10.669(3),b = 35.981(5),c= 11.417(2) Å,β= 89.654(3) ° andZ= 8. Dimethisoquin crystallizes in the space groupP¯1 witha = 5.338(6),b=7.579(8),c= 22.336(24) Å,a = 85.513(1), β = 87.447(1), γ=84.863(1) °, andZ=2. Intensity data were collected with an automatic diffractometer, using CuKα for dibucaine and MoKα for dimethisoquin. The structures were solved using direct methods and refined by full-matrix least-squares procedures to final values ofR= 12.8% for dibucaine andR = 8.8% for dimethisoquin. The structures of both local anesthetics feature planar fused ring systems and extended side chains, one of which contains the tertiary amine that forms hydrogen bonds with other local anesthetic molecules and the chlorine and solvent atoms. The possible sites of action of the anesthetics in the cell membrane are discussed on the basis of these structures.

The crystal structure of a tricarbonyliron-substituted cyclobutadiene complex: Tricarbonyl[1,2,2a,12a-η-5,10-dimethyldibenzo[a,c]cyclobuta[f]cyclooctene-3,12-dione] iron

Abstract The crystal structure of tricarbonyl[1,2,2a,12a-η-5,10-dimethyldibenzo[a,c] cyclobuta[f] cyclooctene-3,12-dione] iron, C20O2H14Fe(CO)3, has been determined by the single crystal X-ray diffraction technique using data collected with a fully automated diffractometer. The unit cell is monoclinic, space group P21/c with a = 12.025(2), b = 23.204(3), c = 14.614(2) A, β = 102.16(2)°, and contains eight molecules (two per asymmetric unit). The structure was elucidated to study the coordination of the iron atom and to correlate the conformation of the eight-membered ring with an NMR study. The final structure was obtained by Patterson-superposition and Fourier techniques and refined by full-matrix least-squares to a crystallographic residual of 0.070. In both independent molecules the iron atom is 1.77 A from the cyclobutadiene ring to which it is coordinated. The eight-membered ring is flattened from a boat conformation and the six-membered rings are twisted from coplanarity by an angle of 74°. The twist is such that if one ring points above the eight-membered ring, the other ring will point below it with concomitant pointing in the opposite sense by neighboring carbonyl groups. Such a structure has conformational chirality. It is postulated that the stability of each enantiometer is due to the particularly high energy of the transition state through which each would pass to achieve inter-conversion.

On the molecular structures ofcyclo-heptasulfur andcyclo-decasulfur

The observed conformations ofcyclo-S7 andcyclo-S10 may be successfully predicted by assuming idealized molecular symmetries ofm (Cs) and 222 (D2), respectively, with bond angles all equal to the natural value of 106.4 °. These assumptions lead to widely different torsion angles, all of which are close to the observed values.