Micelle-Tagging Electrophoresis in High Electric Fields: Optimal Design of Microfluidic Devices

Forensics, genetics, and medicine rely on the information from rapid DNA separations. We present improvements to DNA separation speed with the use of microchips to increase the throughput of micelle tagging electrophoresis (MTE). MTE separates DNA through the transient interactions of an alkane modification and a micelle buffer. These interactions impart a small amount of drag which allows for separation without overly slowing down the mobility of DNA. Microfluidics allow for higher electric fields, shorter capillary lengths, and access to spatial detection modes. These characteristics increase the theoretical throughput of microchip electrophoresis over conventional benchtop setups.

This increased speed does however come at the cost of resolving power. As the separation speeds increase, the distance between bands of DNA decreases and there is the potential for band broadness sources not observed at lower electric fields. Fortunately, the diffusion and injection limited separations at low electric fields seem to remain present at 1350 V/cm. The separation of a custom polymerase chain reaction ladder and a forensically relevant short tandem repeat DNA ladder verifies the diffusion and injection limitation. These separations however come short of resolving 500 bases, which is a benchmark read length for short tandem repeat applications.

Electroosmotic counter flow and hybrid detection methods are potential methods of extending read length while maintaining rapid separations. A weakness of standard MTE separations is the reversal of the elution order. Since the largest fragments elute first, they have less time to resolve and the shorter resolved fragments are slow, which increases runtime. Electroosmotic counterflow allows for the reversal of MTE elution order, which extends the read length. Using finish-line detection in combination with a spatial detection mode allows for the detection of DNA as it elutes and reduces the wait time after resolution of the longest length occurs. A direct enumeration of the sample space predicts the optimal dimensions, electroosmotic flow, and micelle size for MTE with electroosmotic flow.

To control the electroosmotic flow needed for optimal microchip separations, buffer additives modify the electroosmotic mobilities in the capillary. While additives like salts and polymers can tune electroosmotic flow, they each have shortcomings. Salts suffer from increasing Joule heating from the increased ionic strength. Many of the polymers fully suppress electroosmotic flow, which makes tuning electroosmotic flow difficult. Some promising additives were ethylene glycol, SB3-12, and hexamethonium bromide. Hexamethonium bromide was the most successful in controlling electroosmotic flow because there was low run to run variability and there was a strong concentration dependence for tunability. A MTE run with added hexamethonium bromide was successful in separating a 26 and 100 base oligo.

The surfactant buffer forms wormlike micelle networks. These networks have the potential to sieve the DNA, which acts as another separation mechanism during MTE. A reptation in a reversible gel model had previously described the separation mechanism, but the experiments at high electric fields suggest that the mechanism resembles that of solid post collisions. This model implies the micelles impede the DNA, and the DNA unwinds to unhook from the micelle. As the electric field increases this unhooking time decreases, which increases the mobility with the electric field. This separation mode separates a HindIII restriction digest of lambda DNA in under 10 seconds using microfluidics.

DNA secondary structure reduces the mobility of the DNA in MTE. Experiments with structured and unstructured DNA revealed there was a 1.4% reduction in mobility without denaturing conditions. G-quadruplexes formed tight secondary structures that were ideal for testing the differences in electrophoretic mobility. An increase in temperature and the addition of urea were successful in removing most of the DNA structure, which returned the mobility to within 0.04% of a linear oligo. This suggests that denaturing conditions are necessary to prevent misidentification of peaks.

The sum of these projects aims to provide a framework for optimal microchip design for MTE.