(689e) The Sequence-Dependence of the Persistence Length of DNA

While the sequence-dependence of the DNA bending rigidity for short molecules has been studied for many decades, how this short-scale behavior is connected to the persistence length of the chain at long length scales remains unclear. Most studies involving long DNA presume that the persistence length is independent of sequence, based on the assumption that the sequence-dependence is averaged out at larger length scales. We test this conventional wisdom by measuring the extension of human DNA in nanochannels. By labeling the DNA with a sequence-specific probe, we are able to map these molecules back to the reference genome and thus determine the DNA sequence between the probes. The resulting data set consists of more than 10 million such probes, separated by distances ranging from 2.5 kilobase pairs (kbp) to 500 kbp, and containing a wide range of GC content. By converting the fractional extension of these chains to a persistence length by Odijk’s theory, we found that persistence length was positively correlated with GC content, with a variation from 55 nm to 70 nm for sequences containing 31.25% GC and 56.25% GC, respectively. These results, obtained using human DNA and 41 nm nanochannels, were corroborated by additional competitive binding experiments using T4 bacteriophage DNA (35% GC) and lambda-phage DNA (49% GC) in 100 nm nanochannels. We will discuss the importance of these results in the context of sequence-dependent theories for the DNA bending rigidity.