A reliable method for retrieving plasmid DNA from tissue culture cells

We have developed a reliable method of isolating an individual positively stained COS cell using cloning cylinders. Transfection, visualisation and plasmid retrieval can now all be carried out on the original screening plate, using immunohistochemical methods alone. There is no requirement for radiolabelling of probes, reducing the cost and time for each screen. Horst described an adaptation of the Seed et al. expression cloning technique (1) which allows single step-cloning of antigens irrespective of their subcellular localization (2). COS cells were transfected using cDNA inserted into the expression vector pcDM8, grown in polystyrene tissue culture dishes and those cells that produced the protein of interest were visualised using an immunoperoxidase technique. Positive cells were picked by scraping with a hand-held fine tip of a Gilson pipetman (2). Numerous attempts to repeat this procedure were unsuccessful and the use of a small amount of water in the tip to remove the cell from the dish was also inconsistent. We were keen to utilise this method to clone genes encoding proteins recognised by monoclonal antibodies produced in our laboratory. Cloning rings are used routinely to preserve specific cell types with specialised properties and remove them from an overgrowth of unspecialised cells in tissue culture (4). Very fine Teflon cylinders were made (2.5 mm internal diameter, 7 mm external diameter, 15 mm length), vacuum grease applied to one end and a ring was placed over each positive cell. Culture medium was removed from within the cylinder using a drawn-out Pasteur pipette and 50 gl PBS, 10 mM EDTA, 0.6% SDS, 0.025% trypsin and 10 ,ug Proteinase K was added to the cylinder and incubated at 37°C for 1 h. COS cells bind very tightly to polystyrene dishes and are even more adherent post DEAE/dextran transfection. The above cocktail and incubation conditions were found to be the most reliable method of removal. Other combinations and less harsh conditions did not result in viable plasmid (data not shown). The solution was removed using a sterile glass Pasteur pipette, transferred to a microfuge tube and subjected to HIRT (3) treatment. Using this procedure, we can reliably grow 30-70 colonies/cell isolate. We have successfully cloned the gene encoding a protein which is specifically recognised by one of our monoclonal antibodies using this method (in preparation). From a screen of 106 COS cells, six positive cells were picked and DNA was isolated using the method described above. Competent E.coli cells (MC1061/P3) were generated using Seed's calcium chloride method (1) (transformation efficiency 6 x 107/,ug) and DNA transformed into them yielded >400 colonies. Twenty were picked and grown individually, one ofwhich proved to be positive when screened by transfection. Cloning cylinders add reliability and consistency to this excellent method of cloning genes using monoclonal antibodies.