Search PubMedSearch

Biomedical subjects

M Sundaralingam

Publications and source records attributed to M Sundaralingam.

At least 19 recordsLinked to original sources

1.76 A structure of a pyrimidine start alternating A-RNA hexamer r(CGUAC)dG.

The crystal structure of the alternating RNA r(CGUAC)dG with a 3'-terminal deoxy G residue has been determined at 1.76 A resolution. The crystal belongs to the orthorhombic space group C2221, unit-cell dimensions a = 29.53, b = 44.61 and c = 94.18 A, with two independent duplexes (I and II) per asymmetric unit. The structure was solved by the molecular-replacement method. The final R factor was 18.8% using 4757 reflections in the resolution range 8.0-1.76 A. The model contains a total of 496 atoms and 85 solvent molecules. The two duplexes form the repeating unit and stack in the usual head-to-tail (5',3'/5',3') fashion into a pseudocontinuous helical column. Almost all of the 2'-hydroxyl groups are engaged in the three modes of water-mediated interactions to the base N3/O2 atoms, the sugar O4' atoms and the backbone phosphates. Thus, the 2'-hydroxyl group of RNA is probably contributing to the stability of the duplexes.

Biopolymers

Structure of the complex of bovine pancreatic phospholipase A2 with a transition-state analogue.

The 1.89 A resolution structure of the complex of bovine pancreatic phospholipase A2 (PLA2) with the transition-state analogue L-1-O-octyl-2-heptylphosphonyl-sn-glycero-3-phosphoethanolamine (TSA) has been determined. The crystal of the complex is trigonal, space group P3121, a = b = 46.58 and c = 102.91 A and isomorphous to the native recombinant wild type (WT). The structure was refined to a final crystallographic R value of 18.0% including 957 protein atoms, 88 water molecules, one calcium ion and all 31 non-H atoms of the inhibitor at 1.89 A resolution. In all, 7 726 reflections [F>2sigma(F)] were used between 8.0 and 1.89 A resolution. The inhibitor is deeply locked into the active-site cleft and coordinates to the calcium ion by displacing the two water molecules in the calcium pentagonal bipyramid by the anionic O atoms of the phosphate and phosphonate group. The hydroxyl group of Tyr69 hydrogen bonds to the second anionic O atom of the phosphate group while that of the phosphonate group replaces the third water, 'catalytic' water, which forms a hydrogen bond to Ndelta1 of His48. The fourth water which also shares Ndelta1 of His48 is displaced by the steric hinderance of the inhibitor. The fifth conserved structural water is still present in the active site and forms a network of hydrogen bonds with the surrounding residues. The structure is compared to the other known TSA-PLA2 complexes.

Amino Acid Sequence

1.72 A resolution refinement of the trigonal form of bovine pancreatic phospholipase A2.

The trigonal crystal structure of the recombinant bovine pancreatic phospholipase A2 has been re-refined at a slightly higher resolution (1.72 A). The crystals are trigonal, space group P3121, unit-cell parameters a = b = 46.78 and c = 102.89 A and are isomorphous to the previous structure. The structure was refined to a final crystallographic R value of 19.5% (Rfree = 28.4%) using 10 531 reflections. A total of 106 solvent molecules were included in the refinement compared with the earlier refinement which contains only 85 water molecules and 8 925 reflections at 1.8 A resolution. The root-mean-square deviation from the ideal bond lengths and bond angles is considerably better in the present refinement. The active site is extended ( approximately 14 A) from Ala1 to the calcium. The three catalytic residues (Asp99, His48 and the catalytic water) are connected by the conserved structural water and the N-terminal Ala1 on one side, and by the calcium through an equatorial water on the other. The water molecules play a role in the activity of the enzyme PLA2. The Ala1 end of the extended active site performs the activation of the phospholid membranes while the opposite end performs the hydrolysis of the monomeric phospholids.

Amino Acid Sequence

Crystal structure of the complex of bovine pancreatic phospholipase A2 with the inhibitor 1-hexadecyl-3-(trifluoroethyl)-sn-glycero-2-phosphomethanol,.

The structure of recombinant bovine pancreatic phospholipase A2 (PLA2) complexed with the competitive inhibitor 1-hexadecyl-3-(trifluoroethyl)-sn-glycero -2-phosphomethanol (hereafter MJ33), a phospholipid analogue without the sn-3 phosphodiester group, has been determined. The crystals are trigonal, space group P3121, a = b = 46.36 A and c = 102.56 A, and isomorphous to the recombinant PLA2 with one molecule in the asymmetric unit. The structure was refined using 8082 reflections between 8.0 and 1.91 A resolution to a final R-value of 18.4% [Rfree = 28.0%]. The model includes 957 protein atoms, 86 water molecules, one calcium ion, and 26 non-hydrogen atoms of the inhibitor MJ33. The overall tertiary fold of the complex is very similar to that of the inhibitor-free recombinant PLA2 with a root mean square deviation of 0.32 A for all the backbone atoms. The electron density of the surface loop residues 62-66 is clear and ordered, unlike the other trigonal bovine PLA2 structures done to date. This structural change could be responsible for the interfacial allosteric activation, which thermodynamically relates the enhanced binding of the substrate mimic to the active site of the enzyme. MJ33 is tightly bound in the active-site cleft, dislodging the equatorial coordinated calcium water (W5), the putative catalytic water W6, and the neighboring water W7. The axial coordinated calcium water is missing; thus the hexacoordinated calcium is a monocapped pentagonal pyramid. Although MJ33 is a sn-2 tetrahedral mimic, its phosphate binds to PLA2 differently from the sn-2 phosphonate analogue of phospholipids, another tetrahedral mimic. The knowledge of the active-site geometry of MJ33 would be useful in the design of more useful therapeutic agents for PLA2.

Animals

Crystal structure of r(GUGUGUA)dC with tandem G x U/U x G wobble pairs with strand slippage.

To better understand the frequent occurrence of adjacent wobble pairs in ribosomal RNAs we have determined the crystal structure of the RNA duplex, r(GUGUGUA)dC with the 3'-terminal deoxy C residue. Two different crystal forms of the duplex were obtained and both belong to the rhombohedral space group, R3. Crystal form I has hexagonal unit cell dimensions, a = b = 40.82 A and c = 66.09 A and diffracts to 1.58 A resolution, while crystal form II has a = b = 47.11 A and c = 59.86 A, diffracting only to 2.50 A resolution. Both structures were solved by the molecular replacement method using different starting models. In spite of the large differences in the cell dimensions the overall structures in both crystals are similar. Instead of the expected blunt-end duplex with four consecutive G x U pairs, the slippage of the strands resulted in two different tandem G x U/U x G wobble pairs involving two of the central and two of the 5' overhang bases, still yielding a total of four wobble pairs. These tandem wobble pairs are flanked by two Watson-Crick pairs. The A-type duplexes stack in the familiar head-to-tail fashion forming a pseudocontinuous helix. The wobble pairs of the present motif II (G x U/U x G) structure stack with a low twist angle of 25.3 degrees in contrast to that of motif I (U x G/G x U), 38.1 degrees. The four wobble pairs are characteristically heavily hydrated in both the grooves accounting for their stability.

Crystallography, X-Ray

Methylation of the Z-DNA decamer d(GC)5 potentiates the formation of A-DNA: crystal structure of d(Gm5CGm5CGCGCGC).

It is well known that methylation of alternating Py x Pu sequences potentiates the formation of Z-DNA. However, we have now observed that methylation of the alternating Z-DNA oligomer d(GCGCGCGCGC), which starts with a 5'-purine, unexpectedly stabilizes the A-DNA conformation. The double methyl derivative d(Gm5CGm5CGCGCGC), which crystallizes as duplex A-DNA in the hexagonal space group P6(1)22, a = b = 39.33 A and c = 77.93 A with one strand per asymmetric unit and six duplexes in the unit cell, refined to an R factor of 19.1 for 204 DNA atoms and 43 solvent molecules. This is the first report of a DNA sequence crystallized in both right and left-handed conformations, allowing structural comparisons not previously possible and, more importantly, this is the first time that methylation has been shown to potentiate the formation of A-DNA from a sequence known to crystallize as Z-DNA. From this study, ten base-pairs appear to be the critical length in determining the handedness of d(GC)n-type sequences in the crystalline state. Because methylation of nuclear DNA is linked to a number of cellular processes, including transcriptional inactivation, this study has important implications for the role of A-DNA in methylated regions of genomic DNA and, thus, in the regulation of gene expression. In this context, the structure of d(Gm5Cm5CGCGCGC) will be compared with that of the alternating A-DNA decamer d(GCACGCGTGC) and the alternating Z-DNA decamer d(GCGCGCGCGC) and discussed in terms of the forces that govern the handedness of duplex DNA oligomers.

Base Composition

Crystal structure of an alternating octamer r(GUAUGUA)dC with adjacent G x U wobble pairs.

The crystal structure of the RNA duplex, r(GUAUGUA)dC, with a 3'-terminal deoxy C residue, has been determined at 1.38 A resolution. The r(GUAUGU) hexameric consensus sequence is present at the exon-intron junction in pre-mRNAs of yeast and higher eukaryotic organisms. The crystal belongs to the rhombohedral space group R3. The hexagonal unit cell dimensions are a = b = 39.71 A, c = 68.15 A and gamma = 120 degrees with one duplex in the asymmetric unit. The structure was solved using the molecular replacement method. The final model contains 332 atoms of the duplex and 67 solvent molecules. The R-factor is 17.6% (Rfree of 23.1%) for 4035 reflections with F > or = 1.5sigma(F) in the resolution range 10.0 to 1.38 A. The duplex is of the A-type with a pseudodyad relating the two strands. The RNA helix is slightly distorted, in spite of the presence of two adjacent G x U wobble base-pairs located at the center of the helix. The twist angle between the wobble pairs, 38.1 degrees, is above the average value and those between the wobble base-pairs and the flanking Watson-Crick base-pairs, 26.7 degrees and 26.3 degrees, respectively, are lower than the average values. The twist between the junction base-pairs are about 24 degrees. The G x U wobble pairs are bridged by water molecules and solvated in the grooves. G x U base-pairs are as stable as the Watson-Crick A x U pairs and only slightly less stable than the G x C pairs accounting for their frequent occurrence in RNA.

Crystallography, X-Ray

Crystal structures of the side-by-side binding of distamycin to AT-containing DNA octamers d(ICITACIC) and d(ICATATIC).

To understand the recognition interactions between AT-containing alternating DNA and minor groove binding drugs, the crystal structures of the side-by-side binding of two distamycin molecules to the DNA octamers d(ICITACIC)2 and d(ICATATIC)2, referred to here as TA and ATAT, respectively, have been determined at 1.6 A and 2.2 A, respectively. Compared to the previous 2:1 all-IC d(ICICICIC)2-distamycin complex, the substitutions of the I x C base-pairs by the A x T base-pairs enable the interactions of the drug with its natural target to be studied. Both complexes assume side-by-side drug binding, isomorphous to the all IC counterpart in the tetragonal space group P4(1)22 (a = b = 28.03 A, c = 58.04 A and a = b = 27.86 A, c = 58.62 A, respectively). The ATAT complex also crystallized in a new polymorphic monoclinic space group C2 (a = 33.38 A, b = 25.33 A, c = 28.11 A and beta = 120.45 degrees) and was solved at 1.9 A resolution. The structures of the three double drug x DNA complexes are very similar, characterized by systematic hydrogen bonding and van der Waals interactions. Each drug hydrogen bonds with the bases of the proximal DNA strand only and stacks with the sugar moiety, while the side-by-side drugs themselves exhibit pyrrole ring-peptide stacking. The pyrrole-peptide interaction is crucial for the side-by-side binding mode of the distamycin/netropsin family of drugs. The purine-pyrimidine alternation is probably responsible for the striking alternation in the helical and backbone conformations. The structures are conserved between the pure IC complex and the AT substituted complexes but further details of the side-by-side binding to DNA are provided by the 1.6 A resolution structure of TA.

Antiviral Agents

Nine polymorphic crystal structures of d(CCGGGCCCGG), d(CCGGGCCm5CGG), d(Cm5CGGGCCm5CGG) and d(CCGGGCC(Br)5CGG) in three different conformations: effects of spermine binding and methylation on the bending and condensation of A-DNA.

The A-DNA decamer d(CCGGGCCm5CGG) crystallizes in the presence of spermine in three polymorphic forms and with one duplex in the asymmetric unit: hexagonal (P6(1)), unit cell of 55.0 A x 55.0 A x 45.9 A; orthorhombic (P2(1)2(1)2(1)), unit cell of 24.8 A x 44.6 A x 48.0 A, and a second orthorhombic (P2(1)2(1)2(1)), unit cell of 23.6 A x 40.8 A x 43.4 A. The reduction in cell volume among the three different forms is accompanied by a large reduction in solvent content (67% versus 40% versus 24%) and a significant reduction in volume per base-pair (2005 A(3) versus 1325 A(3) versus 1048 A(3)). There is also a concomitant increase in the number of bound spermine molecules per duplex (0 versus 1 versus 2) as well as an increase in DNA bending (10 degrees versus 16 degrees versus 31 degrees), which correspond to major groove widths of 8.0 A versus 4.5 A versus 1.3 A, respectively. The P6(1) crystal form, which represents a new space group for A-DNA decamers, supports one of the most hydrated and extended DNA duplexes to date, while the second orthorhombic form supports one of the least-hydrated and most-condensed non-Z-DNA duplexes. The unmethylated analogue d(CCGGGCCCGG), the double-methyl derivative d(Cm5CGGGCCm5CGG) and the bromine derivative d(CCGGGCC(Br)5CGG) also crystallize in at least two of the aforementioned conformations, and all nine crystal structures were determined. We report, in detail, on the three crystal structures of d(CCGGGCCm5CGG) and the effects of methylation and spermine binding on A-DNA conformation.

5-Methylcytosine

Phospholipase A2 engineering. Structural and functional roles of the highly conserved active site residue aspartate-99.

The aspartate-99 of secreted phospholipase A2 (PLA2) has been proposed to be critical for the catalytic mechanism and interfacial activation of PLA2. Aspartate-99 connects the catalytic machinery (including the catalytic diad, the putative catalytic waters W5 and W6, and the calcium cofactor) to the hydrogen-bonding network. The latter involves Y52, Y73, the structural water, and the N-terminal region putatively required for the interfacial activation. A triple mutant of bovine pancreatic PLA2 with substitutions aspartate plus adjacent tyrosine residues (Y52,73F/D99N) was constructed, its X-ray structure was determined, and kinetic characteristics were analyzed. The kinetic properties of the D99N mutant constructed previously were also further analyzed. The X-ray structure of the Y52,73F/D99N mutant indicated a substantial disruption of the hydrogen-bonding network including the loss of the structural water similar to that seen in the structure of the D99N mutant published previously [Kumar, A., Sekharudu, Y. C., Ramakrishnan, B., Dupureur, C. M., Zhu, H., Tsai, M.-D., & Sundaralingam, M. (1994) Protein Sci. 3, 2082-2088]. Kinetic analysis demonstrated that these mutants possessed considerable catalytic activity with a k(cat) value of about 5% compared to WT. The values of the interfacial Michaelis constant were also little perturbed (ca. 4-fold lower for D99N and marginally higher for Y52,73F/D99N). The results taken together suggest that the hydrogen-bonding network is not critically important for interfacial activation. Instead, it is the chemical step that is perturbed, though only modestly, in the mutants.

Amino Acid Sequence

Comparison of major groove hydration in isomorphous A-DNA octamers and dependence on base sequence and local helix geometry.

The family of ten isomorphous tetragonal A-DNA octamers provides a unique opportunity to examine major groove hydration in terms of base sequence and local parameter effects. The presence of a severely underwound central py.pu base step (average = 24.1 degrees), which lies on a crystallographic 2-fold in the unit cell, provides a sharp change in the local environment in which to study and separate the effects of base sequence and local helix geometry on major groove hydration. For this reason, and to avoid bias secondary to end effects, hydration analysis was restricted to the central four dyad-related base paris. This study finds that d(CG) base pairs are better hydrated than d(TA) base pairs, 2.5 H2O vs 1.3 H2O; steps with high twist angles are better hydrated than steps with low twist angles, 6.9 H2O vs 0 H2O; negative roll angles are better hydrated than positive roll angles, 2.8 H2O vs 1.8 H2O; and flanking base pairs are better hydrated than central base pairs, 2.6 H2O vs 2.0 H2O, a phenomenon which is sequence independent, occurring for both d(CG) and d(TA) base pairs. The twist angle and base roll combine to significantly affect the pattern and degree of major groove hydration in this family of octamers. A previous study of A-DNA octamers and their helix parameters established a strong dependency on crystal packing forces with little or no dependence on the base sequence [Ramakrishnan & Sundaralingam (1993) J. Biomol. Struct. Dyn. 11. 11-26]. We find that the degree and pattern of major groove hydration are strongly influenced by the local helix parameters, implying an indirect, but significant, relationship between major groove hydration and environmental forces, i.e., crystal packing, drug binding, and protein-DNA interactions.

Crystallography, X-Ray

Crystal structures of A-DNA duplexes.

All crystal structures of A-DNA duplexes exhibit a typical crystal packing, with the termini of one molecule abutting the shallow grooves of symmetry related neighbors, while all other forms (B, Z, and RNA) tend to form infinitely stacked helices. The A-DNA arrangement leads to the formation of shallow groove base multiples that have implications for the structure of DNA in compacted states. The characteristic packing leaves big solvent channels, which can be sometimes occupied by B-DNA duplexes. Comparisons of the structures of the same oligomer crystallizing in two different space groups and of different sequences crystallizing in the same space group show that the lattice forces dominate the A-DNA conformation in the crystals, complicating the effort to elucidate the influence of the base sequence on the structures. Nevertheless, in both alternating and nonalternating fragments some sequence effects can still be uncovered. Furthermore, several studies have started to define the minimal sequence changes or chemical modifications that can interconvert the oligomers between different double-helical conformers (A-, B-, and Z-form). Overall, it is seen that the rigid nucleotide principle applies to the oligomeric fragments. Besides the structures of the naked DNAs, their interactions with water, polyamines, and metal ions have attracted considerable attention. There are conserved patterns in the hydration, involving both the grooves and the backbone, which are different from those of B-DNA or Z-DNA. Overall, A-DNA seems to be more economically hydrated than B-DNA, particularly around the sugar-phosphate backbone. Spermine was found to be able to bind exclusively to either of the grooves or to the phosphate groups of the backbone, or exhibit a mixed binding mode. The located metal cations prefer binding to guanine bases and phosphate groups. The only mispairs investigated in A-DNA are the wobble pairs, yielding structural insight into their effects on helix stabilities and hydration. G.T wobble pairs have been determined in various sequence contexts, where they differentially affect the conformations and stableness of the duplexes. The structure of a G.m5C base pair, which surprisingly also adopted the wobble conformation, suggests that a similar geometry may transiently exist for G.C pairs. These results from the crystalline state will be compared to the solution state and discussed in relation to their relevance in biology.

Base Sequence

Crystal structure of d(GCGCGCG) with 5'-overhang G residues.

The crystal structure of the DNA heptamer d(GCGCGCG) has been solved at 1.65 A resolution by the molecular replacement method and refined to an R-value of 0.184 for 3598 reflections. The heptamer forms a Z-DNA d(CGCGCG)2 with 5'-overhang G residues instead of an A-DNA d(GCGCGC)2 with 3'-overhang G residues. The overhang G residues from parallel strands of two adjacent duplexes form a trans reverse Hoogsteen G x G basepair that stacks on the six Z-DNA basepairs to produce a pseudocontinuous helix. The reverse Hoogsteen G x G basepair is unusual in that the displacement of one G base relative to the other allows them to participate in a bifurcated (G1)N2 . . . N7(G8) and an enhanced (G8)C8-H . . . O6(G1) hydrogen bond, in addition to the two usual hydrogen bonds. The 5'-overhang G residues are anti and C2'-endo while the 3'-terminal G residues are syn and C2'-endo. The conformations of both G residues are different from the syn/C3'-endo for the guanosine in a standard Z-DNA. The two cobalt hexammine ions bind to the phosphate groups in both GpC and CpG steps in Z(I) and Z(II) conformations. The water structure motif is similar to the other Z-DNA structures.

Base Composition

Correlation of hydrophobicity and packing in A-DNA oligonucleotides.

A-DNA oligomers pack in a slanted fashion with the terminal base pairs abutting into the minor groove of neighboring molecules unlike the other forms of DNA which pack by vertically stacking one over the other into helical columns. To explain the differences in packing we have advanced a hypothesis that the orientation of the sugar-phosphate backbone is different in A-DNA from that in the other forms of DNA, mainly due to the differences in the sugar puckering.

DNA

Phospholipase A2 engineering. Deletion of the C-terminus segment changes substrate specificity and uncouples calcium and substrate binding at the zwitterionic interface.

It has been suggested [Dijkstra, B. W., Drenth, J., & Kalk, K. H. (1981) Nature 289, 604-606] that the interfacial binding site of phospholipase A2 (PLA2) involves a large number of residues, including a cluster at the N-terminus and another cluster at the C-terminus. The approaches of multiple mutation and deletion were used to test the roles of the C-terminal residues of bovine pancreatic PLA2 overexpressed in Escherichia coli. A double mutant K120A/K121A and a deletion mutant delta 115-123/ C27A were constructed, and structural and functional analyses were performed on both mutants. The double mutant showed little perturbation in the global structure on the basis of proton NMR and X-ray crystallographic analyses. The proton NMR analysis of the deletion mutant suggested that a few residues at the active site, the hydrophobic channel, and the calcium binding loop are perturbed, but the global conformation is not changed. The mutants were then characterized for catalytic and binding properties by use of various kinetic and spectroscopic methods. The double mutant behaved in a manner similar to that of the wild type (WT) PLA2 in every property examined. The deletion mutant was found to show an interesting change of substrate specificity. The kcat,app of the zwitterionic DC8PC micelles but not the anionic DC8PM micelles decreased by a factor of > 100; however, the activity of DC8PC was restored upon addition of 4 M NaCl. The results of fluorescence spectroscopic studies indicate that the deletion mutant behaves in a manner similar to that of WT in the binding to anionic vesicles and to zwitterionic neutral diluent. Thus, the binding affinity of the enzyme to the interface (the E to E* step) should not be the main cause for the change in substrate specificity. The cause lies at least partially in the binding of substrate or inhibitor to the active site of the enzyme at the interface, i.e., the E* to E*L step, as revealed by the results of equilibrium binding studies. The equilibrium dissociation constants of ligands are generally higher for the deletion mutant (relative to WT) at the zwitterionic interface but not at the anionic interface. The cause for the low affinity of an active site-directed ligand to the active site at the zwitterionic interface could be related to the inability of Ca2+ to enhance ligand binding for the deletion mutant. This is in contrast to the WT PLA2 for which Ca2+ binding enhances binding of the substrate to the active site. Overall, the structural and functional perturbations caused by deleting the C-terminal segment are modest, but the changes in substrate specificity and the uncoupling between substrate and calcium binding are interesting and significant.

Amino Acid Sequence

Crystal structure of the self-complementary 5'-purine start decamer d(GCGCGCGCGC) in the Z-DNA conformation. I.

Alternating self-complementary oligonucleotides starting with a 5'-pyrimidine usually form left-handed Z-DNA; however, with a 5'-purine start sequence they form the right-handed A-DNA. Here we report the crystal structure of the decamer d(GCGCGCGCGC) with a 5'-purine start in the Z-DNA form. The decamer crystallizes in the hexagonal space group P6(5)22, unit cell dimensions a = b = 18.08 and c = 43.10 A, with one of the following four dinucleotide diphosphates in the asymmetric unit: d(pGpC)/d(GpCp)/d(pCpG)/d(CpGp). The molecular replacement method, starting with d(pGpC) of the isomorphous Z-DNA hexamer d(araC-dG)3 without the 2'-OH group of arabinose, was used in the structure analysis. The method gave the solution only after the sugar-phosphate conformation of the GpC step was manipulated. The refinement converged to a final R value of 18.6% for 340 unique reflections in the resolution range 8.0-1.9 A. A result of the sequence alternation is the alternation in the nucleotide conformation; guanosine is C3'-endo, syn, and cytidine is C2'-endo, anti. The CpG step phosphodiester conformation is the same as ZI or ZII, whereas that of the GpC step phosphodiester is "intermediate" in the sense that zeta (O3'-P bond) is the same as ZII but alpha (P-O5' bond) is the same as ZI. The duplexes generated from the dinucleotide asymmetric unit are stacked one on top of the other in the crystal to form an infinite pseudocontinuous helix. This renders it a quasi-polymerlike structure that has assumed the Z-DNA conformation further strengthened by the long inner Z-forming stretch d(CG)4. An interesting feature of the structure is the presence of water strings in both the major and the minor grooves. In the minor groove the cytosine carbonyl oxygen atoms of the GpC and CpG steps are cross-bridged by water molecules that are not themselves hydrogen bonded but are enclosed by the water rings in the mouth of the minor groove. In the major groove three independent water molecules form a zigzagging continuous water string that runs throughout the duplex.

Base Sequence

Crystal structure of the self-complementary 5'-purine start decamer d(GCACGCGTGC) in the A-DNA conformation. II.

The crystal structure of the alternating 5'-purine start decamer d(GCGCGCGCGC) was found to be in the left-handed Z-DNA conformation. Inasmuch as the A.T base pair is known to resist Z-DNA formation, we substituted A.T base pairs in the dyad-related positions of the decamer duplex. The alternating self-complementary decamer d(GCACGCGTGC) crystallizes in a different hexagonal space group, P6(1)22, with very different unit cell dimensions a = b = 38.97 and c = 77.34 A compared with the all-G.C alternating decamer. The A.T-containing decamer has one strand in the asymmetric unit, and because it is isomorphous to some other A-DNA decamers it was considered also to be right-handed. The structure was refined, starting with the atomic coordinates of the A-DNA decamer d(GCGGGCCCGC), by use of 2491 unique reflections out to 1.9-A resolution. The refinement converged to an R value of 18.6% for a total of 202 nucleotide atoms and 32 water molecules. This research further demonstrates that A.T base pairs not only resist the formation of Z-DNA but can also assist the formation of A-DNA by switching the helix handedness when the oligomer starts with a 5'-purine; also, the length of the inner Z-DNA stretch (d(CG)n) is reduced from an octamer to a tetramer. It may be noted that these oligonucleotide properties are in crystals and not necessarily in solutions.

Base Sequence