MethodsDesign, data-processing and modelling: DNA sequences were designed using our own algorithms based on sequence symmetry minimization implemented in Matlab and C (available at http://www.dna.caltech.edu/DNAdesign/). Curve fits for persistence length data and models for lattice strain energies were calculated in Matlab.Molecular models were constructed and visualized using a combination of NAMOT, RasMol, and PyMol (scripts and coordinates at http://www.dna.caltech.edu/SupplementaryMaterial/).The molecular models are for visualization only and have not been subjected to molecular dynamics calculations.DNA sample preparation: Lyophilized HPLC-or PAGEpurified DNA oligonucleotides were purchased from Integrated DNA Technologies (Coralville, IA), resuspended in water, quantitated by UV absorbance at 260 nm, and stored at -20• C. All samples were prepared in a 1X Tris-Acetate-EDTA (TAE) buffer with 12.5 mM magnesium acetate (pH=8.3). An equimolar mixture of strands (5 strands if one tile, 10 strands if two) was annealed from 95• C to 25• C (fluorescence microscopy) steps in a PCR machine (Eppendorf Mastercycler). For AFM, each strand was present at 200 nM, for fluorescence microscopy the total concentration of tiles was kept at 400 nM. For fluorescence microscopy, a single fluorescein-labeled strand was incorporated into each tile; the position of the dye was varied from the 5 end of the #3 strand to the 5 end of the #5 strand with no apparent effect. AFM of REp+SEp(3:FAM) was similar to that of REp+SEp.Preparation of PVP coated glass: Adapted from.1 Microscope slides and coverslips were washed in 1M NaOH for 1 hour, rinsed thoroughly with de-ionized (DI) water and immersed in 1% v/v acetic acid solution for 2 hours. Then, they were rinsed again with DI water and silanized in a 1% v/v 3-(trimethoxysilyl)propylmethacrylate (Aldrich) in 1% v/v acetic acid for 36 hours. For polymer coating, 500 mL of a 4% w/v Mw = 360, 000 poly(vinylpyrrolidone) (PVP, USB Corp.) solution with 2.5 mL of 10% w/w ammonium persulfate solution and 250 µl of N,N,N ,N -tetramethylethylenediamine (TEMED, Acros) was prepared. Slides and coverslips were incubated in the PVP solution at 80• C for 18 hours. They were then rinsed and stored in DI water. Coating was stable for at least 2 weeks.Preparation for fluorescence microscopy: Samples were left overnight at room temperature after annealing. Immediately prior to use, a PVP-coated microscope slide and coverslip were rinsed with ethanol and dried. Then, 2.6 µl of solution containing DNA tubes and oxygen scavenging system (0.035 mg/ml catalase, 0.2 mg/ml glucose oxidase, 4.5 mg/ml glucose, 5% β-mercaptoethanol) was deposited onto the slide, covered with the coverslip and sealed with epoxy or parafin. The distance between slide and coverslip was ≈5 µm and the thickness of sample solution was typically narrowed to ≈3 µm by the PVP coating.Fluorescence microscopy: Samples were imaged on an inverted microscope (IX 70, Olympus) with 100X/1.40 NA oil immersion and 40X/0.75 NA air objectives. Blue lig...

We present the first direct observations of tile-based DNA self-assembly in solution using fluorescent nanotubes composed of a single tile. The nanotubes reach tens of microns in length by end-to-end joining rather than by sequential addition of single tiles. Their exponential length distributions withstand dilution but decay via scission upon heating, with an energy barrier Esc approximately 180kBT. DNA nanotubes are thus uniquely accessible equilibrium polymers that enable new approaches to optimizing DNA-based programming and understanding the biologically programmed self-assembly of protein polymers.

The standard model for the structure of collagen in tendon is an ascending hierarchy of bundling. Collagen triple helices bundle into microfibrils, microfibrils bundle into subfibrils, and subfibrils bundle into fibrils, the basic structural unit of tendon. This model, developed primarily on the basis of x-ray diffraction results, is necessarily vague about the cross-sectional organization of fibrils and has led to the widespread assumption of laterally homogeneous closepacking. This assumption is inconsistent with data presented here. Using atomic force microscopy and micromanipulation, we observe how collagen fibrils from tendons behave mechanically as tubes. We conclude that the collagen fibril is an inhomogeneous structure composed of a relatively hard shell and a softer, less dense core.

The diffusion of fluoresceinated probes inside single collagen fibrils was investigated by imaging the migration of fluorescence along the fibrils in oil and by monitoring fluorescence recovery after photobleaching (FRAP). Probes were excluded from the fibrils according to their size. Probes that were not excluded diffused in the fibrils, but FRAP occurred 6 x 10(-4) times more slowly than in water due to binding interactions between collagen and the probes. The dissociation constant of the fluorescein-collagen complex was determined (K(D)=1.8+/-0.1 microM).

The dynamical behavior of the neutral polymer (dextran, M(w)=2 x 10(6)) is investigated during DNA electrophoresis in a dilute solution. Using a fluorescence recovery after photobleaching setup, we measured the velocity of fluorescein-labeled dextran induced by the migration of the DNA. We found that each DNA molecule drags a large number of dextrans with it. We show that DNA-dextran interactions are not only binary but long range and indirect. We conclude that the DNA-dextran complex creates a hydrodynamic field that entrains polymers far from the DNA during electrophoresis.

No abstract