Separation and Purification of Plasmid Mixtures by Continuous Elution Electrophoresis

Benchmarks Separation and Purification of Plasmid Mixtures by Continuous Elution Electrophoresis BioTechniques 29:1204-1206 (December 2000) Isolation and purification of cova- lently closed circular (CCC) plasmid DNA is necessary for a wide range of molecular procedures. For many appli- cations, standard plasmid isolation methods provide plasmid DNA of ade- quate quality. However, there are se- lected situations when a mixture of plasmids exist, such as with the use of supF-containing plasmids (10) and as- says or systems involving co-transfec- tions. Plasmids that contain the supF selection marker have advantages for use in particular procedures, but this se- lection system necessitates growth in bacteria containing a helper plasmid, such as p3. The fact that standard plasmid pu- rification techniques co-purify both the supF-containing and helper plasmids have been an obstacle to the wider use of these plasmids, as the separation and purification of individual plasmids from mixtures of plasmids have been problematic. We have developed a tech- nique for purifying CCC plasmid DNAs that differ in mass as part of an antitumor “naked DNA” or polynu- cleotide vaccine (PNV) strategy. This purification technique makes use of the Model 491 Prep Cell (Bio-Rad Labora- tories, Hercules, CA, USA) and contin- uous elution electrophoresis, which pu- rifies the supF-containing plasmid, free from the p3 helper plasmid and RNA. The development of PNV strategies and the promise of human clinical trials have focused attention on purification procedures for generating therapeutic DNA preparations containing only CCC plasmids. Most PNV vectors contain antibiotic resistance gene coding se- quences within the vector backbone, which have raised concerns both from immunologic and regulatory stand- points (4,12). There are established vec- tors that do not contain intact antibiotic resistance gene sequences, such as those containing the supF gene se- quence, and these supF-containing plas- 1204 BioTechniques mids require growth in concert with an appropriate helper plasmid such as p3 (10). The p3 helper plasmid, contained in a number of commercially available strains of E. coli (DH10§/p3; Life Tech- nologies, Rockville, MD, USA; and MC1038, Top 10F/p3; Invitrogen, Carlsbad, CA, USA), is a 60-kb plasmid containing an intact kanamycin resis- tance coding sequence along with tetra- cycline and ampicillin coding se- quences, each with inactivating amber stop codon point mutations. The supF gene product encodes a mutant prokary- otic tRNA that complements these mu- tations and allows for growth under dual ampicillin and tetracycline selection. To apply the supF/p3 selection sys- tem to our application, it was necessary to develop an alternative purification procedure for isolating different-sized plasmids in their native CCC state. Con- tinuous elution electrophoresis using the Model 491 Prep Cell is an electro- phorectic method originally designed for purifying proteins in polyacrylamide gels. Although restriction fragments (9) and RNA (2,6) have been separated us- ing agarose and polyacrylamide mix- tures in the Prep Cell, exposure to ethid- ium bromide and acrylamide was not desirable for a potentially therapeutic plasmid DNA product. Thus, we modi- fied the Prep Cell to run pure agarose gels. This led to the purification of high- quality CCC plasmid while avoiding exposure of the DNA to both acry- lamide and ethidium bromide, thus lim- iting the acquisition of potential toxic contaminants during the isolation and purification procedure. This continuous elution electropho- resis procedure provides an alternative isolation/purification procedure that al- lows for the separation of individual plasmid components from a plasmid mixture and makes the use of supF-con- taining plasmids for PNV and other pro- cedures more feasible. We grew and iso- lated a number of supF-containing plasmids, including pCDM8 obtained from Invitrogen, and our PNV construct designated pITL (Wetzel et al. unpub- lished) in E. coli strain DH10§/p3 pur- chased from Life Technologies. DH10§/ p3 was transformed with these plasmids using a standard heat shock procedure (8) and grown in Luria-Bertani (LB) Broth (8) under dual selection of 50 µg/mL ampicillin and 10 µg/mL tetra- cycline at 39°C–40°C. Cultures at an A 600 of approximately 0.5 were ampli- fied with 10 µg/mL chloramphenicol and grown for an additional 16–18 h. Initial plasmid mixture preparations were purified by CsCl-ethidium bro- mide density gradient centrifugation before use for evaluating continuous elution electrophoresis. Subsequently, crude lysate was prepared from the bac- teria using Triton X-100 lysis (8) and treated with RNase A/T1. Electropho- resis was carried out using the Model 491 Prep Cell. The gel casting tube was initially modified by replacing the stan- dard glass cylinder with a 1.75-inch (OD) Eastman Tenite Butyrate plastic tube obtained from Consolidated Plas- tics (Twinsburg, OH, USA). Subse- quently, more rigid lucite tubes were prepared by the onsite machine shop from standard lucite stock and used in an identical manner with equal efficacy. The agarose gels consisted of two parts. A 3–5 mm base layer of 2% SeaKem Gold agarose (BMA, Rock- land, ME, USA) was poured, and fol- lowing polymerization of this layer, an 8-cm, 1.0% or 0.8% agarose gel was poured on top of the initial base layer. Crude lysate or CsCl-ethidium bro- mide-purified plasmid preparations were loaded onto the gel and elec- trophoresed overnight at 135 V in 1× TAE buffer (0.04 M Tris-acetate and 0.001 M EDTA, pH 8.0). Sample was eluted at a rate of 0.25 mL/min using a peristaltic pump (Rainin Instrument, Woburn, NJ, USA), routed through a UV detector and collected with a frac- tion collector (both from Amersham Pharmacia Biotech, Piscataway, NJ, USA). A254 readings were recorded on a two-channel chart recorder. Sequen- tial fractions (approximately 4.5 mL) corresponding to the plasmid DNA peak on the chromatogram were col- lected. Twenty microliters per fraction were analyzed on a 0.8% agarose slab gel. Bands were visualized by staining with ethidium bromide. Initial experiments using plasmid DNA purified by CsCl density ultra- centrifugation allowed us to establish the optimal agarose concentration and gel height. We found that an 8-cm gel provided adequate separation and pu- rification of the supF-containing plas- Vol. 29, No. 6 (2000)