Capillary electrophoretic separation of uncharged polymers using polyelectrolyte engines. Theoretical model.

We recently demonstrated that the molecular mass distribution of an uncharged polymer sample can be analyzed using free-solution capillary electrophoresis of DNA-polymer conjugates. In these conjugates, the DNA is providing the electromotive force while the uncharged polydisperse polymer chains of the sample retard the DNA engine with different amounts of hydrodynamic drag. Here we present a theoretical model of this new analytical method. We show that for the most favourable, diffusion-limited electrophoresis conditions, there is actually an optimal DNA size to achieve the separation of a given polymer sample. Moreover, we demonstrate that the effective friction coefficient of the polymer chains is related to the stiffness of the two polymers of the conjugate, thus offering a method to estimate the persistence length of the uncharged polymer through mobility measurements. Finally, we compare some of our predictions with available experimental results.

Molar mass profiling of synthetic polymers by free-solution capillary electrophoresis of DNA-polymer conjugates.

The molar mass distribution of a polymer sample is a critical determinant of its material properties and is generally analyzed by gel permeation chromatography or more recently, by MALDI-TOF mass spectrometry. We describe here a novel method for the determination of the degree of polymerization of polydisperse, uncharged, water-soluble polymers (e.g., poly(ethylene glycol) (PEG)), based upon single-monomer resolution of DNA-polymer conjugates by free-solution capillary electrophoresis. This is accomplished by end-on covalent conjugation of a polydisperse, uncharged polymer sample (PEG) to a monodisperse, fluorescently labeled DNA oligomer, followed by electrophoretic analysis. The monodisperse, charged DNA "engine" confers to each conjugate an equal amount of electromotive force, while the varying contour lengths of the uncharged, polydisperse polymers engender different amounts of hydrodynamic drag. The balance of electromotive and hydrodynamic forces enables rapid, high-resolution separation of the DNA-polymer conjugates as a function of the size of the uncharged PEG tail. This provides a profile of the molar mass distribution of the original polymer sample that can be detected by laser-induced fluorescence through excitation of the dye-labeled DNA. We call this method free solution conjugate electrophoresis (FSCE). Theory-based analysis of the resulting electrophoresis data allows precise calculation of the degree of polymerization of the PEG portion of each conjugate molecule. Knowledge of the molecular mass of the uncharged polymer's repeat unit allows for direct calculation of the molar mass averages as well as sample polydispersity index. The results of these analyses are strikingly reminiscent of MALDI-TOF spectra taken of the same PEG samples. PEG samples of 3.4-, 5-, and 20-kDa nominal average molar mass were analyzed by FSCE and MALDI-TOF; the values of the molar mass averages, Mw and Mn, typically agree to within 5%. Measurements and molar mass calculations are performed without any internal standards or calibration. Moreover, when DNA-polymer conjugate analysis is performed in a chip-based electrophoresis system, separation is complete in less than 13 min. FSCE offers an alternative to MALDI-TOF for the characterization of uncharged, water-soluble polymers that can be uniquely conjugated to DNA.

Separating DNA sequencing fragments without a sieving matrix.

The possibility of separating appropriately labeled DNA fragments using free-flow capillary electrophoresis was predicted a few years ago based on simple theoretical arguments. Free-flow separation of double-stranded DNA (dsDNA) fragments in the 100-1000 base range was later demonstrated using a streptavidin label. In this article, we now report that end-labeled free-flow electrophoresis (ELFSE) can also be used to sequence single-stranded DNA (ssDNA). The first 100 bases of a DNA sequencing reaction were read without any sieving matrix when fractionated streptavidin was added to the 5'-end of the ssDNA fragments. These separations required only 18 min and did not require coated capillaries. An analysis of the results indicates that sample injection, analyte-wall interactions and thermal diffusion are the limiting factors at this time. Extrapolating from our data, we predict that several hundred bases could be sequenced in less than 30 min with the proper conditions. ELFSE thus offers an attractive potential alternative to polymer solutions for DNA sequencing in capillaries and microchips.

Separation of DNA sequencing fragments using an automated capillary electrophoresis instrument.

Rapid, high-resolution separation of DNA sequencing fragments by capillary gel electrophoresis using an automated, commercially available instrument is presented. The effect of column lengths and electric field strength on the resolution of sequencing fragments as well as the sensitivity of laser-induced fluorescence (LIF) detection was investigated. Using a short capillary of 20 cm length, which results in a U-shape of the capillary in the capillary cartridge, very high separation efficiency, up to 17 x 10(6) theoretical plates per m, is obtained. Analysis of the band broadening factors revealed that the resolution on the short column is predominantly determined by axial diffusion and to a minor extent by detection zone width. Presumably due to the coiling of longer capillaries in the capillary cartridge, increasing the capillary length does not increase the separation efficiency as predicted for diffusion-limited separation. The concentration limit of detection (signal-to-noise ratio = 2) is 0.2 x 10(-12) M of fluorescein-labeled oligonucleotide primer under the separating conditions for DNA sequencing samples. Increasing the electric field strength from 100 to 175 V/cm improved resolution and at the same time approximately doubled the sequencing speed. Fragments up to 500 nucleotides in length are resolved in less than 50 min.

Fine-mapping of shotgun template-libraries; an efficient strategy for the systematic sequencing of genomic DNA.

To test the effectiveness of ordering shotgun DNA-templates prior to sequence analysis, the 450 kb left arm of yeast chromosome XII was randomly subcloned into a phagemid vector. Clones were ordered by hybridisation to an average map density of one new insert every 125 bp and are currently used for sequencing the chromosomal fragment. An 11.5 kb overlap between the template map and a DNA fragment that had been sequenced earlier allowed an independent evaluation of the strategy's effectiveness. To this end, clones were selected from the map and tag-sequenced from either end, thus comparing the map position with the actual location within the 11.5 kb. Of 65 selected clones, taken mostly at random from a total of 423, 58 mapped on average about a quarter of a clone length around their predicted position, with the other seven being between 0.6 and 1.5 clone length off. 75-86 sequencing reactions on clones selected from the map would have been sufficient for completely sequencing both strands of the 11.5 kb fragment. The results demonstrate the efficacy of such template sorting, considerably assisting sequencing at relatively little cost on the mapping level.

Line scanning system for direct digital chemiluminescence imaging of DNA sequencing blots.

A cryogenically cooled charge-coupled device (CCD) camera equipped with an area CCD array is used in a line scanning system for low-light-level imaging of chemiluminescent DNA sequencing blots. Operating the CCD camera in time-delayed integration (TDI) mode results in continuous data acquisition independent of the length of the CCD array. Scanning is possible with a resolution of 1.4 line pairs/mm at the 50% level of the modulation transfer function. High-sensitivity, low-light-level scanning of chemiluminescent direct-transfer electrophoresis (DTE) DNA sequencing blots is shown. The detection of DNA fragments on the blot involves DNA-DNA hybridization with oligonucleotide-alkaline phosphatase conjugate and 1,2-dioxetane-based chemiluminescence. The width of the scan allows the recording of up to four sequencing reactions (16 lanes) on one scan. The scan speed of 52 cm/h used for the sequencing blots corresponds to a data acquisition rate of 384 pixels/s. The chemiluminescence detection limit on the scanned images is 3.9 x 10(-18) mol of plasmid DNA. A conditional median filter is described to remove spikes caused by cosmic ray events from the CCD images.

Digital chemiluminescence imaging of DNA sequencing blots using a charge-coupled device camera.

Digital chemiluminescence imaging with a cryogenically cooled charge-coupled device (CCD) camera is used to visualize DNA sequencing fragments covalently bound to a blotting membrane. The detection is based on DNA hybridization with an alkaline phosphatase(AP) labeled oligodeoxyribonucleotide probe and AP triggered chemiluminescence of the substrate 3-(2'-spiro-adamantane)-4-methoxy-4-(3"-phosphoryloxy)phenyl- 1,2-dioxetane (AMPPD). The detection using a direct AP-oligonucleotide conjugate is compared to the secondary detection of biotinylated oligonucleotides with respect to their sensitivity and nonspecific binding to the nylon membrane by quantitative imaging. Using the direct oligonucleotide-AP conjugate as a hybridization probe, sub-attomol (0.5 pg of 2.7 kb pUC plasmid DNA) quantities of membrane bound DNA are detectable with 30 min CCD exposures. Detection using the biotinylated probe in combination with streptavidin-AP was found to be background limited by nonspecific binding of streptavidin-AP and the oligo(biotin-11-dUTP) label in equal proportions. In contrast, the nonspecific background of AP-labeled oligonucleotide is indistinguishable from that seen with 5'-32P-label, in that respect making AP an ideal enzymatic label. The effect of hybridization time, probe concentration, and presence of luminescence enhancers on the detection of plasmid DNA were investigated.

Multiwavelength fluorescence detection for DNA sequencing using capillary electrophoresis.

Multiwavelength detection of laser induced fluorescence for dideoxynucleotide DNA sequencing with four different fluorophores and separation by capillary gel electrophoresis is described. A cryogenically cooled, low readout noise, 2-dimensional charge-coupled device is used as a detector for the on-line, on-column recording of emission spectra. The detection system has no moving parts and provides wavelength selectivity on a single detector device. The detection limit of fluorescently labeled oligonucleotides meets the high sensitivity requirements for capillary DNA sequencing largely due to the efficient operation of the CCD detector with a 94% duty cycle. Using the condition number as a selectivity criterion, multiwavelength detection provides better analytical selectivity than detection with four bandpass filters. Monte Carlo studies and analytical estimates show that base assignment errors are reduced with peak identification based on entire emission spectra. High-speed separation of sequencing samples and the treatment of the 2-dimensional electropherogram data is presented. Comparing the DNA sequence of a sample separated by slab gel electrophoresis with sequence from capillary gel electrophoresis and multiwavelength detection we find no significant difference in the amount of error attributable to the instrumentation.