[Univ. of Illinois]  [School of Chemical Sciences]  [Dept. of Chemistry]  [Organic Chemistry]  [Chemical Biology]

[Journals List]  [Group Meetings]  [Group Pictures]  [CV & Contact Info]

[SKS Lab Home]  [SKS Lab People]  [SKS Lab Publications]

DNA as a macromolecular conformational constraint

At least three properties of double-stranded DNA make it an excellent candidate for a nanoscale construction element: (1) Duplex DNA is rigid on the nanometer scale (10 DNA base pairs = 1 helix turn = 34 Å = 3.4 nm). (2) DNA oligonucleotides are easily prepared in various lengths by standard solid-phase synthesis. (3) DNA nucleosides incorporating chemical modifications are readily prepared and used in solid-phase DNA synthesis. We capitalize on these favorable chemical and physical properties by using DNA as a structural constraint to control the conformations of other molecules. Our efforts to date have used DNA to control RNA folding, in part because we can take advantage of our substantial expertise in the area, and in part because this allows us to study RNA folding and catalysis using the DNA constraints. For a description of our research efforts on RNA folding and catalysis, see below and also our webpage on RNA folding & catalysis.

For a recent review article that describes the use of DNA as a macromolecular conformational constraint, including both our own work and that of other labs, see Silverman, Mol. Biosyst. 2007, 3, 24-29 [#53].

Context of DNA constraints relative to DNA nanotechnology

Others have used DNA as a building block for what has been termed "DNA nanotechnology". For any device or application, the use of DNA may be classified by considering two separate questions: (1) Is the DNA static or dynamic during operation? (2) Does the device or application involve DNA only, or are other large molecules involved? Efforts from other laboratories achieve three of the four possible combinations of these questions (see matrix below). Arguably the most interesting combination, in which DNA dynamically controls the structure of another molecule, has not been reported by others, although it seems that any practical DNA nanotechnology requires this goal to be achieved. As described here, our laboratory has recently accomplished experimentally a simple form of this goal (see Miduturu & Silverman, J. Am. Chem. Soc. 2005, 127, 10144-10145 [#39] on the Full publications listing). We have followed up in a subsequent study that expands our ability to modulate the DNA constraints (see Miduturu & Silverman, Angew. Chem. Int. Ed. 2006, 45, 1918-1921 [#45]).

Choice of the macromolecule whose structure is controlled

For the macromolecule whose conformation is controlled, we initially chose the P4-P6 domain of the Tetrahymena group I intron RNA, which is a structurally well-characterized large RNA (160 nucleotides, molecular weight ~51000; see Fundamental studies of RNA folding and catalysis). Because the X-ray crystal structure of P4-P6 is known (green structure below), we can choose our DNA constraint lengths and locations rationally, and we can predict the structural consequences on the folding of the RNA macromolecule. For example, at the two nucleotides (U107 and C240) whose 2'-positions are marked in blue below, simple measurement predicts that a 10-base pair (bp) DNA constraint is too short to span the two sites in the correctly folded RNA conformation, and thus formation of the DNA constraint should induce the RNA to misfold. That is, the DNA constraint is a "mismatch" with the RNA conformation. Conversely, a 20-bp DNA constraint is long enough to span the two sites without distorting the RNA conformation, and thus formation of the DNA constraint should not induce the RNA to misfold; the DNA constraint is a "match".

Construction of a DNA constraint on a macromolecule

To use DNA for controlling the conformation of another molecule, we must have the ability to attach the DNA covalently to other molecules. Attachment of single-stranded DNA to RNA by reductive amination is shown schematically below. Synthesis of the illustrated diol-phosphoramidite has been published (see Miduturu & Silverman, J. Org. Chem. 2006, 71, 5774-5777 [#49]). The multi-step assembly route necessary to build the large P4-P6 RNA with two attached DNA strands is shown below the synthetic scheme.

Experimentally demonstrating the DNA constraint effect

After developing the synthetic procedures necessary to prepare large RNA molecules like P4-P6 with attached DNA strands, we showed that DNA constraints can control RNA folding. Experimental demonstration of the DNA constraint effect was achieved using several techniques that are appropriate for examining the structures of folded macromolecules. One such technique is nondenaturing polyacrylamide gel electrophoresis (native PAGE). In this method, the shape and compactness of the RNA molecule controls its migration rate through a highly cross-linked polyacrylamide gel; more compact molecules run more quickly through the gel under the influence of an electrical gradient. Because RNA folding requires Mg²⁺, the average compactness of P4-P6 is a sensitive function of the Mg²⁺ concentration. When P4-P6 has no DNA constraint attached, the Mg²⁺ midpoint of its folding as assessed by native PAGE is ~0.7 mM. When a DNA constraint is attached by the methods described above, the Mg²⁺ midpoint is observed to shift, but only if the constraint is predicted to distort the RNA conformation. For example, a 10-bp ("mismatched") DNA constraint attached at nucleotides U107-C240 causes a tremendous shift in the Mg²⁺ midpoint, whereas a 20-bp ("matched") constraint has almost no effect. These measurements are readily converted to free energy effects; the 10-bp constraint increases the folding free energy by DDG° >6 kcal/mol, but the 20-bp constraint increases the folding free energy by only DDG° = 0.4 kcal/mol. These values are entirely as expected from the simple geometrical considerations depicted above, which indicate that only the 10-bp constraint is structurally incompatible with the correctly folded P4-P6 conformation.

A second technique useful for investigating RNA conformation is chemical probing using dimethyl sulfate (DMS). In this method, DMS is added to a solution of RNA, which leads to methylation of certain functional groups such as the N¹ positions of adenine nucleobases. However, this methylation occurs at each location only if the relevant RNA functional group is not buried within the interior of the folded RNA structure, which would "protect" it from reaction with DMS. Because DMS methylation depends on the folded state of the RNA in this way, DMS probing may be used to determine the folded RNA state as a function of solution conditions such as the Mg²⁺concentration. In practice, the extent of methylation at each RNA position that may be affected (such as each adenine N¹)  is assayed by reverse transcription (RT) using the DMS-treated RNA as a template and an added DNA oligonucleotide as the primer. Extension of the DNA primer is observed at RNA adenosine nucleotides that are not N¹-methylated, whereas a reverse transcription abort band is observed at N¹-methylated adenosines, which must have been accessible to DMS at the time of probing. The extent of accessibility (and therefore the intensity of the abort band on PAGE) correlates with the extent of RNA folding. When this assay was performed for the same DNA-constrained P4-P6 variants that were examined above by native PAGE, the plotted data look almost identical. This very close correlation between two extremely different physical methods, native PAGE and DMS probing, provides great confidence that we are truly examining an effect of the DNA constraint upon RNA folding, rather than some artifact of a particular experimental technique.

Modulating the DNA constraint effect

The ability to distort RNA conformation in a rational manner using a DNA constraint is only half of the story. Additionally, we would also like to modulate this DNA constraint effect; i.e., to control the integrity of the DNA constraint by external means. This could be done in a number of ways, which include the following: (1) Using complementary DNA oligonucleotides to separate the interacting DNA strands of the constraint. (2) Adding a DNA-cleaving enzyme to destroy the constraint. This enzyme may either be nonspecific for any double-stranded DNA, such as deoxyribonuclease (DNase), or specific for certain DNA sequences, such as a restriction enzyme. (3) Adding a chemical reagent that cleaves the RNA-DNA linkage. (4) Using light as a reagent for photocleavage of the RNA-DNA linkage. We have achieved the first three of these methods experimentally, and we are working on the fourth.

Representative data for method (1), which uses oligonucleotides to modulate the constraint, are shown below. These data highlight the reversibility of using DNA oligonucleotides to modulate the constraint effect.

We have also achieved the use of fluorescence to monitor modulation of the DNA constraint, instead of using native PAGE or DMS probing. A pyrene chromophore that is covalently attached to the RNA can report directly on the RNA's folded state via changes in the pyrene fluorescence emission intensity (see our webpage on Fundamental studies of RNA folding and catalysis). We have combined the use of pyrene fluorescence to monitor RNA folding with the use of DNA oligonucleotides to modulate the DNA constraint effect. Note that the constraint effect is reversible over many cycles of "complement" and "rescue" strand addition. The use of a practical detection method like fluorescence in this fashion offers hope that DNA constraints may be integrated into practical applications such as sensor devices.

Using DNA constraints to control ribozyme activity: Hammerhead ribozyme

A clear objective for applying DNA as a constraint is to move beyond merely controlling macromolecular structure by controlling macromolecular function. Because large RNA molecules can have catalytic activity as ribozymes, we chose to apply DNA constraints for controlling ribozyme catalysis. We began by studying the hammerhead ribozyme, which is an often-studied natural self-cleaving RNA originally found in viroids and satellite RNAs of plant viruses. The hammerhead ribozyme is the simplest RNA capable of self-cleavage at biologically relevant rates. Because the hammerhead ribozyme has been extensively studied biochemically and because the structure of the catalytically active conformation was recently determined by X-ray crystallography, we felt that this ribozyme is ideal for our initial effort in applying DNA constraints to control catalysis.

The hammerhead ribozyme can be studied in a two-stranded form: one "enzyme" strand with 43 nucleotides, and one "substrate" strand with 20 nucleotides that contains the phosphodiester linkage to be cleaved. Using a deoxyribozyme obtained by in vitro selection that attaches DNA to RNA, thereby making branched RNA-DNA conjugates (see our webpage on DNA as a catalyst), we attached two DNA strands at separate positions of the ribozyme. When the attached DNA strands were complementary, the ribozyme structure was disrupted by the DNA constraint, and catalysis was greatly suppressed. Moreover, addition of a free oligonucleotide complementary to one of the constraint strands removed the constraint and restored catalytic function. These experiments demonstrated that that DNA constraints can be applied to control RNA catalysis and not only RNA structure (see Zelin & Silverman, ChemBioChem 2007, 8, 1907-1911 [#60]).

Using DNA constraints to control ribozyme activity: Group I intron RNA

We have recently expanded DNA constraints to control the folding and catalysis of the Tetrahymena group I intron ribozyme, which has 388 nucleotides. When two attached DNA strands were complementary, nearly complete abolition of catalysis was observed. Activity was then restored to wild-type level upon separation of the constraint strands via addition of free complementary DNA. We intend to use this DNA constraint approach to prepare discrete misfolded states of the intron and compare these to previously reported and biologically relevant misfolded states (kinetic traps).

Photocleavable DNA constraints

Some of our current efforts are focused on combining DNA constraints with our photochemical approach to study RNA folding. This will allow time-resolved studies of RNA folding when initiated from DNA-constrained misfolded conformations. Please see our webpage on Fundamental studies of RNA folding and catalysis for a brief description of the general photochemical approach to study RNA folding.

Click here to return to the Silverman lab overview of research

Click here for a full publications list

Research description last updated January 26, 2008