CHAPTER
14
COS Cell Expression John F. Hancock
1. Introduction COS-1 cells were created by transforming an established line of monkey epithelial cells, CV1, with a defective mutant of SV40 (1). The SV40 mutant used carried a small deletion within the origin of replication and, although this construct transformed CV1 cells, which are permissive for lytic growth of SV40, no infectious virus was produced after prolonged culture. One transformed cell line, COSl, was fully characterized and found to contain the complete early region of the SV40 genome. COSl cells express nuclear large T and all proteins necessary for replication of appropriate circular genomes. This was first demonstrated by showing that COSl cells could support the replication of early region mutants of SV40. More important, however, the introduction of any plasmid containing an SV40 origin of replication into COS1 cells results in rapid replication of the plasmid to high copy number. Coincidently, of course, the transfected cells will express any gene on the plasmid that is driven by a suitable eukaryotic promoter. The combined effect of these phenomena is transient high-level expression of the encoded protein. Using a COS-cell expression system has several advantages over generating stable cell lines. First, the whole process is rapid, taking just a few days from transfection to assay. Second, no selection for transfectants is required, since a high proportion of the cells take up the plasmid DNA and express the transfected gene. Third, uniformly high expression of wild-type and mutant From: Methods in Molecular Biology, Vol. 8: Practical Molecular Virology: Viral Vectors for Gene Expression Edited by: M. Collins Q 1991 The Humana Press Inc., Clifton, NJ
153
154
Hancock
forms of proteins can be expected as long as the expressed protein is stable and nontoxic. The advantage over prokaryotic expression systems is that the expressed protein should undergo normal posttranslational processing and hence be localized normally within the cell or be secreted into the culture medium. These characteristics of COS-cell transient expression have been exploited in a wide variety of ways, for example, expression cloning of growth factor receptors and lymphocyte antigens (2,3), functional confirmation of cDNA predicted protein sequences (4), and analysis of the posttranslational pro cessing of various cellular proteins and oncogene products (5,6).
1.1. Principle The introduction of DNA into eukaryotic cells may be achieved by electroporation, calcium phosphate precipitation, or DEAEdextran transfection. Each method has advantages and disadvantages. Calcium phosphate precipitation is best employed for the generation of stably transfected cell lines, whereas DEAEdextran techniques are most applicable to transient expression sytems. A DEAEdextran method will be described here that is simple to use and has proved to give highly reproducible results. The basis of the method is to mix together plasmid DNA in the absence of any carrier DNA with a solution of DEAEdextran. It is presumed that the the DNA forms high-mol-wt complexes with the dextran that stick to the COS cells when the mixture is applied to a cell monolayer. These complexes are internalized, probably by endocytosis, although this has never been formally demonstrated. Once inside the cell, the plasmid DNA is assembled into nucleosome-containing minichromosomes (7) and is rapidly replicated. The plasmid DNA does not become integrated into the hostcell genome with any significant frequency. The efficiency of DNA uptake by the cells is much improved by including a DMSO shock step that stimulates endocytosis (8). Some investigators advocate the inclusion of a chloroquine treatment of cells to further increase the efficiency of DF,AEdextran transfections. Our experience is that this increases the toxicity of the procedure without significantly improving the proportion of successfully transfected COS cells. For example, without including such a chloroquine incubation, up to 20% of COS cells consistently take up DNA and express novel protein at high levels when assayed by immunofluorescence (9). Protein expression is detectable 24 h after the DNA is added to the cells and reaches a peak at around 66-72 h after transfection. The only constraint on plasmid design is the presence of an SV40 origin of replication. Otherwise, the SV40 enhancer/early promoter or the CMV promoter have both proven to be highly efficient for cDNA expression in
155
COS Cell Expression
COS cells (1Q 11). The DNA is applied in high concentration to the cells and the preparation should therefore be as pure as possible, preferably by banding on cesium chloride gradients (12).
2. Materials 1. 2. 3. 4. 5.
Dulbecco’s Modified Eagle’s Medium (DMEM). Donor calf serum. Phosphate-buffered saline, magnesium- and calcium-free (PBS). Trypsin (1:250) in Puck’s saline. DEAEdextran stock Thoroughly dissolve DEAE-dextran (mol wt = 500,000 Sigma D9885) to a final concentration of liOmg/mL in sterile O.lM Tris-HCl, pH 7.0. Filter-sterilize the solution and store as lOGZOO PL aliquots at -2OOC. Thaw the requisite number of aliquots immediately before use. 6. HEPESbuffered saline (HBS): For 2x HBS, make up a solution of 280 mMNaCl,50 mMHEPES, 1.5 mMNa9HP04, and adjust to pH 7.1-7.2. Sterilize the stock solution by filtration and store at -20°C in ZO-mL aliquots. 7. Shock solution: Make up a solution of 10% DMSO in lx HBS and filtersterilize. Shock solution is stable at room temperature. 8. TE Buffer: 10 mMTris-HCl, pH 7.5; 1 mMEDTA.
3. Method 1. Prior to transfection grow the COS cells in DMEM containing 10% v/v donor calf serum (DClO) so they are just confluent. A 175mm tissueculture flask grown to this density contains approx 15 x lo6 COS cells. If the cells are allowed to grow to too high a density, trypsinization becomes difficult and the cells detach in clumps. 2. On the day of transfection (day 1) trypsinize the cells from the culture flask Remove the culture medium and wash the cell monolayer twice with 25 mL of warm PBS. Then incubate the cell monolayer in 3 mL trypsin solution at 37OC until the cells are rounded and starting to float free from the flask (approx I-Zmin) . Add 7 mL of DC10 and titrate well to obtain a good single-cell suspension. 3. Count the cell suspension and plate the COS cells at 6 x ld per 60-mm tissue-culture dish in a total volume of 5 mL of DClO. Incubate the plates at 37OC for 2 h. 4. Toward the end of the incubation, make up the DNA solutions for transfection. Aliquot 1.5 mL of DMEM (containing no serum) into sterile
156
5.
6.
7.
8.
9.
10.
11.
Hancock bijoux. To this, add 5 ltg of plasmid DNA in TE buffer, mix, and then add 7.5 ltL of stock DEAEdextran and mix again. No precipitate should be visible. If concentrated (>3 pg/mL) plasmid solutions are added to DMEM containing DEAEdextran, the DNA can form large clumps. This is avoided by following the addition sequence indicated. Check that the cells have reattached and are evenly spread over the plate. Remove the DC10 and wash the cells twice with 5 mL of prewarmed DMEM (containing no serum). The cells are only weakly adherent at this stage, so wash carefully. Aspirate the final wash and add the DNA mixture. Swirl each plate gently to ensure that the mixture is equally distributed over the plate surface. Incubate at 37*C for 4 h. Ensure that the plate is completely flat in the incubator, so that the small volume of medium covers all the cells. Check that the cells are still attached. They may have started to round up during this incubation, but only a tiny minority should be free-floating, since COS cells are well able to tolerate serum-free conditions. Aspirate the DNA mixture and carefully add 1 mL of warm shock solution. DMSO is toxic at a 10% concentration and must be added very carefully to avoid washing the cells completely off the dish. Trickle the shock solution slowly down the side of the dish and rotate it once to distribute the solution over all the cells; then quickly return the dish to the incubator. The cells must be exposed to the shock solution for no longer than 2 min, therefore, start timing as soon as the first drop of shock solution is added and have the cells ready for the next step after exactly 2 min. Aspirate the shock solution and gently wash the cells once with 5 mL of warm PBS. Replace the PBS with 5 mL of DC10 and incubate at 37*C. The cells always look very sick at this stage, with many being partially rounded up rather than flat and firmly adherent. On day 2 check that the cells have recovered from the transfection procedure. The majority should now be firmly attached and have relatively normal morphology, although vacuolation can be pronounced. Expect a small proportion of free-floating dead cells; however, there should be no area of the dish that is bare of cells if the shock step was carried out correctly. There is no need to medium change the cells (see Note 4), On day 4, 64-72 h after the DMSO shock, assay the cells for protein expression. The cells should be just confluent at this stage.
COS Cell Expression
157
4. Notes 1. These transfections can readily be scaled up. For example, seed 4 x lo6 cells in 30 mL of DC10 onto a 140-mm plate and, after 2 h, replace with 12 mL of DMEM containing 40 ltg plasmid and 60 ltL of DEAEdextran stock. Keep all incubation times unchanged, but increase the volumes of the washes to 2OmL, and use 5 mL of shock solution in step 8 of Section 3. 2. If required, the COS cells may, on day 2, be reseeded to smaller tissueculture dishes or onto cover slips for immunofloresence studies. Aspirate the DClO, wash twice with PBS, trypsinize, and take the cells up in an appropriate volume of DClO, e.g., 8 mL for four 30-mm dishes. Aliquot the cells and return to the incubator. It is much easier to reseed cells in this way than to attempt to transfect COS cells directly onto very small dishes or directly onto cover slips. This procedure also ensures that nearly identical aliquots of cells are available for later assays, since any variation in transfection efficiency between plates is abolished. 3. Cells can be pulse-labeled on day 4, before lysis, and assayed or labeled overnight by switching the cells to labeling medium on day 3,48 h after the DMSO shock. In the case of secreted proteins, replace the culture medium with a small volume of fresh medium (e.g., 1.5 mL for a 60-mm dish) on day 3, 48 h after the DMSO shock, and, after an overnight incubation, assay for activity in the medium. Repeat this procedure on day 4 by replacing the overnight medium with a small volume of fresh medium and incubating for a further 8-12 h to collect a second sample for assay. 4. This protocol has been found to work well for a wide variety of constructs in different vectors. Transfection efficiencies assayed by immunofluoresence are at least lo%, and more typically 20%. If these levels are not achieved, an attempt may be made to further optimize the procedure for an individual clone of cells and expression vector. The transfection efftciency is dependent on the DNA and DEAEdextran concentrations. Therefore, set up a series of control transfections varying both the plasmid concentration (between 1-15 ug/mL) and the Iinal DFAFdextran concentration (between 10&.500 ug/mL) and assay each set of combinations. Certain investigators also add chloroquine diphosphate to the COS cells, along with the DNAdextran mixture, in an attempt to improve transfection efficiencies. If this step is to be included, add chloroquine to the DNAdextran mixture (to a final concentration of 100 pJ4) before applying the mixture to the washed COS
Hancock cell monolayer. It is then important to check the cells every hour of the subsequent 4-h incubation to ensure that excessive death is not occurring.
5. References
5 6. 7.
8.
9.
10. 11. 12.
Gluzman, Y. (1981) SV40 transformed simian cells support the replication of early SV40 mutants. CCU23,175-182. D’Andrea, A. D., Lodish, H. F., and Wong, G. G (1989) Expression cloning of the murine erythropoietin receptor. cell 57,277-285. Aruffo, A. and Seed, B. ( 1987) Molecular cloning of a CD28 cDNA by a high efi%ciency COS cell expression system. f%vc. Natl. Acad. Sn. USA 84,8573-8577. Yamasaki, X., Taga, T., Hirata, Y., Yawata, H., Bawanishi, Y., Seed, B., Taniguchi, T., Hirano, T., and Xishimoto, T. (1988) Cloning and expression of the Human Interleukin-6 @SF-2/IFN-2) Receptor Snenu 241,825-828 Jing, S. Q. and Trowbridge, I. S. (1987) Identification of the mtermolecular disulphide bonds of the human transfer+ receptor and its lipid attachment site. EMBOJ 6, 327-331. Hancock, J. F., Magee, A. I , Childs, J., and Marshall, C. J. (1989) All ras protems are polyisoprenylated but only some are pahnitoylated. Cell 57,1167-l 177. Reeves, R., Corman, C., and Howard, B. (1985) Minichromosome aasembly of non-integrated plasrmd DNA transfectcd mto mammalian cells, NucLrc Ands R.es 13, 3599-3615. Lopata, M. A., Cleveland, D. W., and Sollner-Webb, B. (1984) High level expresston of a chloramphenicol acetyltransferase gene by DEAEdextran mediated DNA transfection coupled with a dimethyl sulphoxide or glycerol shock treatment. Nucleic Ands Be.% 12,5707-5717. Hancock, J. F., Marshall, C. J., McKay, A. I., Gardner, S., Houslay, M. D , Hall, A., and Wakelam, M. J. 0. (1988) Mutant but not normal ~21” elevates inositol phosphate breakdown in two dtfferent cell systems. Oncogene 3,187-193. Miller, J. and Cermain, R. N. (1986) Efficient cell surface expression of class II MHC molecules in the absence of associated invariant chain.J. Exp. Med. 164,1478-1489. Seed, B. (1987) An WA-3 cDNA encodes a phosphohpid linked membrane protein homologous to its receptor CD2. Natun 329,840-842. Mania& T., Frltisch, E. F., and Sambrook, J. (1982) Molecular Ckmng. A Laborat Manuul (Cold Spring Harbor Laboratory, Cold Spnng Harbor, NY)
