Adenine-rich sequences (A-tracts) occur strategically in regions of curvature [Sinden, 1994] and draw continued interest because the origin and nature of bending (as well as interpretation of crystal and solution data) remain unresolved. To explain dissimilar bend directions in A-tract crystals, versus unique bend directions in solution, we simulated two A-tracts whose structures had also been determined crystallographically: CGCGA₆CG and CGCA₆GCG.

Our simulations captured a subtle conversion (from initially dissimilar to similar bends equivalent to minor groove compression) [Strahs & Schlick, 2000]. The subtlety of the changes from the crystallographic structures together with development of bend directions as predicted in solution experimental studies suggested that we could reconcile data from various studies through common bend-stabilizing motifs in our simulations [Young et al., 1995; Young & Beveridge, 1998; Sprous et al., 1999; Sherer et al., 1999].

These motifs include large amounts of propeller twisting in AT base pairs, consistent sugar puckering differences between A and T sugars, a minor-groove water spine with high occupancy and long lifetimes stabilized by a narrow minor groove, and large rolling on the 5' side in relatively straight A-tract moieties. We explain [Strahs & Schlick, 2000] part of the differing interpretations by program-dependent [Lu & Olson, 1999, Lu et al., 1999] measurements of curvature. We avoid this sensitivity in our study by defining global bend angles -- global roll and tilt -- that depend on all base pair bends rather than only on external values, as generally done.

Our calculated average and maximum lifetimes of the water spine molecules are in excellent agreement with various NMR studies. Intriguingly, the geometrical requirements of the water spine stabilize the propeller twist and perhaps the sugar puckering differences between the A and T sugars.

• TATA DNA Sequence/Activity Correlation
  
  
• Software -- MadBend
  
  
• Article -- "A-tract Bending: Insights into Experimental Structures by Computational Models"
  
  
• Article -- "Dynamic Simulations of 13 TATA Variants Refine Kinetic Hypotheses on Sequence/Activity Relationships"
  
  
• Article -- "Computational Challenges in Simulating Large DNA over Long Times"
  
  
• Article -- "Computational Modeling Predicts the Structure and Dynamics of Chromatin Fiber"
  
  
• Art Gallery -- "Dominant DNA Motion (Movie)"
  
  
• Art Gallery -- "Hypothetical Deformation Toward DNA Structure in Complex (Movie)"
  
  
• Dan Strahs' Home Page
  
  

Click to go back to Research