The read length for DNA sequencing using capillary electrophoresis and replaceable linear polyacrylamide (LPA) solutions has been extended to more than 1000 bases with a run time of 80 min. This result was successfully achieved through the combined use of cycle sequencing with dye-labeled primers, improved matrix and separation conditions, and enhanced base-calling software. The influences of LPA molecular weight and concentration on separation were investigated. Additionally, the separation buffer, column temperature, and electric field were adjusted to increase the number of resolvable DNA fragments per run while maintaining an enhanced separation speed. Using low concentrations [2% (w/v)] of high molecular weight LPA polymers (> 5.5 x 10(6) Da), elevated column temperature (50 degrees C) and moderately high field (150 V/cm), rapid sequencing analysis for more than 1000 bases on a model ssM13mp18 template was obtained with 96.8% accuracy.

Long, accurate reads are an important factor for high-throughput de novo DNA sequencing. In previous work from this laboratory, a separation matrix of high-weight-average molecular mass (HMM) linear polyacrylamide (LPA) at a concentration of 2% (w/w) was used to separate 1000 bases of DNA sequence in 80 min with an accuracy close to 97% (Carrilho, E.; et al. Anal. Chem. 1996, 68, 3305-3313). In the present work, significantly improved speed and sequencing accuracy have been achieved by further optimization of factors affecting electrophoretic separation and data processing. A replaceable matrix containing a mixture of 2.0% (w/w) HMM (9 MDa) and 0.5% (w/w) low-weight-average molecular mass (50 kDa) LPA was employed to enhance the separation of DNA sequencing fragments in CE. Experimental conditions, such as electric field strength and column temperature, as well as internal diameter of the capillary column, have been optimized for this mixed separation matrix. Under these conditions, in combination with energy-transfer (BigDye) dye-labeled primers for high signal-to-noise ratio and a newly developed expert system for base calling, the electrophoretic separation of 1000 DNA sequencing fragments of both standard (M13mp18) and cloned single-stranded templates from human chromosome 17 could be routinely achieved in less than 55 min, with a base-calling accuracy between 98 and 99%. Identical read length, accuracy, and migration time were achieved in more than 300 consecutive runs in a single column.

In a previous paper, a 2% w/w replaceable high molecular mass linear polyacrylamide solution (high molecular mass LPA) was used to achieve long read-lengths for DNA sequencing by capillary electrophoresis (E. Carrilho et al., Anal. Chem. 1996, 68, 3305-3313). In that work, the polymer was prepared by polymerization in water at 6% w/w, followed by dilution to 2% w/w. In this study, an improved method for preparation of high molecular mass LPA was developed, based on inverse emulsion polymerization. With this polymerization procedure, the LPA results in a molecular mass of approximately 9 MDa, with characteristics of a fine powder of high purity and practically unlimited shelf life. Using size exclusion chromatography (SEC) and viscosity measurements to characterize the polymer, good batch-to-batch reproducibility was found. It was observed that the viscous polymer solutions made from these high molecular mass polymers require careful preparation and handling because the method of dissolution could affect the molecular mass distribution and the resultant separation of DNA components. Solutions containing 2% w/w of LPA made by emulsion polymerization were simple to prepare, resulting in excellent performance as a replaceable matrix for DNA sequencing by capillary electrophoresis. The viscosity of the polymer decreased exponentially when pressure was applied, allowing easy replacement from a capillary using a syringe. With a properly prepared matrix, a read-length of more than 1000 bases in 80 min with an accuracy better than 97%, and better than 99% for the first 800 bases, could be achieved.

An automated system for loading samples into a microcoil NMR probe has been developed using segmented flow analysis. This approach enhanced 2-fold the throughput of the published direct injection and flow injection methods, improved sample utilization 3-fold, and was applicable to high-field NMR facilities with long transfer lines between the sample handler and NMR magnet. Sample volumes of 2 microL (10-30 mM, approximately 10 microg) were drawn from a 96-well microtiter plate by a sample handler, then pumped to a 0.5-microL microcoil NMR probe as a queue of closely spaced "plugs" separated by an immiscible fluorocarbon fluid. Individual sample plugs were detected by their NMR signal and automatically positioned for stopped-flow data acquisition. The sample in the NMR coil could be changed within 35 s by advancing the queue. The fluorocarbon liquid wetted the wall of the Teflon transfer line, preventing the DMSO samples from contacting the capillary wall and thus reducing sample losses to below 5% after passage through the 3-m transfer line. With a wash plug of solvent between samples, sample-to-sample carryover was <1%. Significantly, the samples did not disperse into the carrier liquid during loading or during acquisitions of several days for trace analysis. For automated high-throughput analysis using a 16-second acquisition time, spectra were recorded at a rate of 1.5 min/sample and total deuterated solvent consumption was <0.5 mL (1 US dollar) per 96-well plate.