[14 Fluorescence Activated Cell Sorting of Hybrid and Transfected Cells

By Michael E. Kamarck

The introduction of chromosomes and DNA into somatic cells by whole cell hybridization and transfection has supported studies on gene mapping, cloning, and function. The selection of host cells containing donor genetic material most often has depended on the complementation of defective genes, antibiotic resistance, or altered phenotype. Genes coding for cell surface exposed molecules also have been introduced into somatic cells, and their expression assessed with specific antibodies and indirect immunofluorescence. Using the fluorescence-activated cell sorter (FACS) individual cells have been isolated based on the expression of these surface antigen genes.

FACS selection of hybrid and transfected cells has facilitated a genetic analysis of the cell surface. The expression of surface antigens on hybrid cells allows the mapping of surface antigen genes and the development of hybrid cell lines homogeneous for the presence of a specific human chromosome and its surface antigens.1,2 Using transfection protocols, it is possible to introduce cloned DNA sequences into recipients and to assess their contribution to the surface antigen array on the host cell. This approach has been useful in identifying functional members of large multi-

1 M. E. Kamarck, j. A. Barbosa, and F. H. Ruddle, Somatic Cell Genet. 8, 385 (1982).

2 P. G. M. Peters, M. E. Kamarck, M. Hemler, J. Strominger, and F. H. Ruddle, Somatic Cell Genet. 8, 825 (1982).

gene families, such as the major histocompatibility complex.3,4 Finally, the entire genome of a donor can be introduced into host cells by transfection, resulting in a "library" of transfectants containing the introduced genome in randomized fragments. Sorter selection of the recipient cells expressing a surface antigen allows the antigen-coding genes to be isolated in host cells,5,6 and ultimately cloned.7"9 This chapter details methods which we have used for the selection of hybrid and transfected cells by flow cytometry and sorting.

Cell Lines and Biochemical Selection

Donor human chromosomes and DNA are introduced into derivatives of the mouse L cell10 with specific biochemical defects in purine or pyrimi-dine biosynthesis. For the formation of human X mouse cell hybrids we have used the cell line LA9," which is defective in the expression of the enzyme hypoxanthine phosphoribosyltransferase (HPRT, EC 2.4.2.8). Cloned genes or genomic DNA is introduced into the cell line Ltk—,12 which is defective in the expression of the thymidine kinase gene (TK; ATP:thymidine 5'-phosphotransferase, EC 2.7.1.21).

L cell lines are maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco, Grand Island, NY) supplemented with 10% fetal calf serum (FCS) (K.C. Biological, Lenexa, Kansas) without antibiotics. Both L cell lines grow in monolayer culture, are removed from flasks using phosphate-buffered saline (PBS) lacking Ca2+ and Mg2+ with 0.03% EDTA, and are passed at dilutions of 1:10 to 1:20. These cell lines can be established as suspension cultures in the same culture medium when passed at cell concentrations of 2 X 105 cells/ml into glass spinner flasks. Stock lines are regularly tested for the presence of mycoplasm using a commercial test kit (Gen-Probe, San Diego, CA). Following cell hybridization or transfection,

3 G. A. Evans, D. H. Margulies, R. D. Camerini-Otero, K. Ozato, and J. G. Seidman, Proc. Natl. Acad. Sci. U.S.A. 79, 1994(1982).

4 J. A. Barbosa, M. E. Kamarck, P. A. Biro, S. M. Weissman, and F. H. Ruddle, Proc. Natl. Acad. Sci. U.S.A. 79, 6327 (1982).

5 P. Kavathas and L. A. Herzenberg, Proc. Natl. Acad. Sci. U.S.A. 80, 524 (1983).

6 L. C. Kuhn, J. A. Barbosa, M. E. Kamarck, and F. H. Ruddle, Mol. Biol. Med. 1, 335 (1983).

7 P. Kavathas, V. P. Sukhatme, L. A. Herzenberg, and J. R. Parnes, Proc. Natl. Acad. Sci. U.S.A. 81, 7688 (1984).

8 L. C. Kuhn, A. McClelland, and F. H. Ruddle, Cell 37, 95 (1984).

9 D. R. Littman, Y. Thomas, P. J. Maddon, L. Chess, and R. Axel, Cell 40, 237 (1985).

10 W. R. Earle, J. Natl. Cancer Inst. 4, 165 (1943).

11 J. W. Littlefield, Exp. Cell Res. 41, 190 (1966).

12 S. Kit, D. Dubbs, L. Pierarski, and T. Hsu, Exp. Cell Res. 31, 297 (1963).

the antibiotics penicillin (100 U/ml) and streptomycin (100 /¿g/ml) are added to culture medium until unique lines have been expanded, and frozen stocks established. The elimination of antibiotics in standard culture work serves to monitor tissue culture technique. Regular use of antibiotics also obscures the presence of mycoplasm infection. For permanent storage, cells are suspended in 90% standard tissue culture medium, with 10% dimethyl sulfoxide (DMSO) (ATCC, Rockville, MD). Cells are aliquoted into freezing vials at concentrations of 5 X 106 cells/ml and placed in the vapor phase of liquid nitrogen.

The L cell lines should be cloned before experimental use. Extensive aneuploidy and heterogeneity of L cell populations have been observed. A unique parental genotype should therefore be established by cloning before hybrid fusion. For transfection experiments, subclones should be established and transfection efficiencies assessed with a selectable plasmid prior to experimental use. Transfection frequencies of one in 300 cells per 50 pg of selectable marker can be achieved with Ltk- clones.6

Both HeLa cells and primary cell fibroblasts have been used as a source of chromosomes and genomic DNA for these experiments. HeLa cells should be cloned and maintained identically to the mouse parental cell lines. Primary cell fibroblasts will grow readily in the standard culture media when passed at low dilution, but are difficult to clone and adapt to spinner flask.

The selectable biochemical marker for the formation of whole cell hybrids is the human enzyme HPRT whose gene is present on the distal tip of the human X-chromosome at Xq26-28. Following fusion, hybrids are selected with medium containing HAT (1 X 10-4 M hypoxanthine, 4 X 10-7 Maminopterin, and 1.6 X 10~5 Mthymidine) and 0.05 mMouabain. All surviving hybrids contain the human X-chromosome, or a fragmented X-chromosome which includes Xq26-28. Hybrids can also be selected on the basis of the thymidine kinase deficiency of Ltk- cells. Human X Ltk" hybrids which survive HAT and ouabain treatment will express the human thymidine kinase gene located at 17q21-22.

Genomic and cloned DNA is introduced into cells by cotransfer with a plasmid containing a selectable marker.13 The selectable plasmid for the cotransfection of Ltk- cells, pTKX-1, contains the herpes simplex virus-thymidine kinase gene (HSV-tk) 3.5 kb BamR\ fragment cloned in pBR322.14 Cells stably transfected with this plasmid will survive treatment with HAT medium. L cell transfectants can also be obtained using the

13 M. Wigler, S. Silverstein, L.-S. Lee, A. Pellicer, T. Cheng, and R. Axel, Cell 16, 777 (1979).

14 L. W. Enguist, G. F. Van de Woude, M. Wagner, J. R. Smiley, and W. C. Summers, Gene 7, 335 (1979).

plasmid pSV2-«eo, which carries the bacterial neomycin resistance gene.15 L cells transfected with pSV2-«eo are selected in medium containing 1 -1.2 mg/ml of the antibiotic G418 (GIBCO, Grand Island, NY).

Hybrid Fusion

Mouse LA9 (HPRT-) and human primary fibroblasts (1 X 107 of each) are removed from flasks and washed in DMEM without serum. The cells are mixed together in a 15 ml centrifuge tube and pelleted at 1500 rpm for 10 min. To the cell pellet add 1 ml of warmed 45% polyethylene glycol 1450 (J. T. Baker, Jackson, TN) in DMEM. This addition takes place over

1 min with gentle disruption of the pellet by the pipet tip. The tube is warmed at 37° for 1 min and the pellet is again diluted with 1 ml of DMEM. After 1 min at 37°, slowly add 10 ml of warm DMEM. Centrifuge the cells and resuspend in DMEM with 10% FCS at a cell concentration of

2 X 106 cells/ml. Add 0.1 ml of cells to each well of a 96-well plate. The next day add 0.1 ml of HAT medium plus 0.05 mM ouabain to each well. Feed these wells at 4 day intervals. Aliquots from hybrids should be expanded for freezing as quickly as possible and cloned. Human chromosome segregation is extremely rapid in the first month after hybrid formation. Characterization of human chromosomes present in the hybrids is performed by methods detailed in this volume.16,17

DNA Mediated Gene Transfection

DNA Preparation

The isolation of plasmid and genomic DNA for use in transfection protocols has been extensively described in other volumes of this series18 and in this volume.19

Cotransfection with Selectable Plasmid and Cloned DNA

Twenty-four hours prior to gene transfer the cells should be plated at a density of 1.2 X 106 cells per 75 cm2 tissue culture flask. For the expression of a cloned gene two to four flasks are used in each transformation. This

15 F. Colbere-Garapin, F. Horodniceanu, P. Kourilsky, and A. Garapin, J. Mol. Biol. 150, 1 (1981).

17 J. M. Trent and F. H. Thompson, this volume [20],

18 A. McClelland, M. Kamarck, and F. H. Ruddle, this series, in press (1986).

19 C. M. Fordis and B. H. Howard, this volume [21].

ensures the generation of a variety of independent transfectants, even if cotransfection frequencies happen to be low. For each flask 100 ng of plasmid pTKX-1 DNA is mixed with 2.5 fig of cloned DNA and 20 fig of mouse L cell carrier DNA. The donor DNA solution is diluted with 250 mM calcium chloride, 25 mM HEPES, pH 7.12 to twice the final DNA concentration desired. The DNA sample should comprise between 1 and 5% of this volume. The DNA is carefully mixed and is added dropwise to an equal volume of 280 mMNaCl, 25 mMHEPES, 1.5 mMNa2HP04, pH 7.12. Constant mixing is achieved by bubbling a gentle stream of air through the solution using a 1 ml sterile plastic pipet attached to a filtered air source. A DNA-calcium phosphate precipitate forms immediately, but the tube should be left undisturbed for 30 min at 20° to allow additional precipitate to form. The precipitate is dispersed by brief shaking and is added directly to the culture medium in the recipient cell flask and incubated at 37°. Good results are obtained with the addition of 1.5 ml of precipitate to each T75 flask.

The medium is replaced with fresh nonselective medium after 20 hr. At 60 hr after transfection, selective medium is added and is changed every 2-3 days. If immediate assessment of cloned gene expression is desired, it may be made between 60 and 84 hr posttransfer. At this time gene expression has been observed in between several percent and 50% of the trans-fected cells.20 Cell cultures remaining under selection will display visible colonies at 8 to 10 days after transfer, and after 2 weeks should be detached and replated in new flasks. This will considerably reduce the problems caused by cellular debris during the cell sorter selection. Frozen cell stocks should be established as soon as possible following cell expansion.

Cotransfection with Selectable Plasmid and Genomic DNA

Twenty-four hours prior to gene transfer the cells should be plated at a density of 2.6 X 106 cells per 150 cm2 tissue culture flask. For each flask 200 ng of plasmid pTK-Xl DNA is mixed with 80 fig of human genomic DNA and diluted into 1.5 ml of 250 mM calcium chloride, 25 M HEPES at pH 7.12. The DNA is carefully mixed and is added dropwise to an equal volume of 280 mM NaCl, 25 mM HEPES, 1.5 mM Na2HP04, pH 7.12. Mixing of these solutions and precipitate formation proceed exactly as described for cotransfection with plasmid and cloned DNA. Following precipitate formation, it is hand shaken and 3 ml added directly to the T150 flask at 37°. Selection, transfer, and maintenance of transfected cells are performed exactly as described for the cotransfer of cloned genes. It is

20 J. A. Barbosa, Ph.D. thesis. Yale University, New Haven, Connecticut, 1983.

crucial to immediately establish frozen stock of the transfectants to minimize the impact of faster growing colonies on the population.

The genomic DNA used for transfection consists of contiguous fragments greater than 100 kilobase pairs (kbp) in size. The ratio of 200 ng of plasmid to 80 of genomic DNA, a molar ratio of approximately 1 to 40, ensures that the majority of tk+ colonies contain unselected genomic DNA. The amount of genomic DNA transferred into an individual cell may range from 500 to 50,000 kilobase pairs.6 If each transfected cell clone contains a random fragment of 1000 kbp, then the human haploid genome (3 X 106 kbp) would be covered by 3000 HSV-tk positive recipient cells. A total of 20 to 25 flasks are used in each transfection experiment to ensure the generation of greater than 20,000 independent clones, which cover the genome with some redundancy.

Independent pools of HAT resistant mass populations should be immediately established if the recovery of distinct positive transfectants is desirable. We commonly pool 3000 to 5000 tk+ colonies for frozen stock and cell sorter analysis.

Fluorescence Labeling

Hybrid cells and transfectants generated by the introduction of cloned gene segments contain antigen-positive subpopulations of significant size (greater than 10% of the population). These cells can be split and passaged at relatively high dilution without danger of eliminating an important cell population. In contrast, genomic cell transfectants must be passaged at relatively low dilution.

Hybrid and transfectant cells are expanded to approximately 107 cells in T-flasks prior to sorter analysis. The cells are detached from the flask by incubation in PBS without Ca2+ and Mg2+ containing 0.03% EDTA at 37°. The cells are washed twice in ice cold DMEM containing 2% FCS. Cells (1-3 X 106) are incubated with saturating amounts of monoclonal antibody or antiserum in 50 fil DMEM with 2% serum for 1 hr at 4° with regular resuspension. Control staining of 1 X 106 cells is performed with irrelevant hybridoma antibody or normal serum as appropriate. The cells are washed twice and incubated for 1 hr at 4° with a fluorescein conjugated second antibody directed against the first antibody (Cooper Biomedical, Malvern, PA). Excellent specificity for mouse monoclonal antibodies can be obtained with fluoresceinated anti-heavy-chain subclass reagents (Southern Biotechnology Associates, Birmingham, AL). If propidium iodide (PI) is added for the detection of dead cells, it is included at 0.5 /zg/ml during this incubation. Finally, the cells are centrifuged through 3 ml of horse serum and resuspended at 2 X 106 cells/ml in 12 X 75 mm test tubes for the FACS. The cells are stored at 4° prior to analysis and will maintain high viability for up to 6 to 8 hours if necessary.

Microscopic examination before FACS evaluation serves to monitor cell viability and staining intensity, and is critical in the interpretation of FAC results. Cells are examined by microscopy using epifluorescence at magnifications of approximately 500X. Excellent sensitivity and resolution of FITC staining are achieved using oil immersion with a Planapochromat 63/1.3 objective (Carl Zeiss, Inc. Thornwood, NY).

FACS Parameters for Analysis and Sorting

Fluorescence analysis and sterile cell sorting is performed on a FACS IV (Becton-Dickinson FACS Systems, Mountain View, CA). Fluorescein and propidium iodide dyes are excited by an argon-ion laser producing 400 mW at 488 nm on light mode. A Zeiss 580 nm dichroic mirror (46 63 05) is used to separate fluorescence signals to two photomultiplier tubes (PMT). Signals to the fluorescein PMT (QL30, EMI Gencon Inc., Plain-view, NY) also pass through a 530/30 dichroic filter (Becton-Dickinson), and signals to the PI PMT (QL20, EMI) through a 625/35 dichroic filter (Becton-Dickinson). Logarithmic conversion of PMT voltage outputs for histogram display are performed with integral FACS electronics. Stream alignment is optimized by the use of standardized fluorescence microspheres of 5.2 fim diameter (Polysciences, Inc., Warrington, PA). In addition to maximizing the fluorescence signal, optimal scatter alignment is critical for distinguishing live and dead cells (see below).

Cell analysis and sorting are performed at rates of 2000 to 2500 cells/ sec with an 80 nm nozzle at transducer frequencies of approximately 20 kHz. Use of this larger nozzle will reduce problems caused by debris plugging the nozzle orifice. Following an initial analysis of 40,000 to 100,000 cells, sorting windows are established. Subpopulations of greater than 1%, as found in hybrids and cloned DNA transfectants, are sorted from the center of a peak to minimize contamination from overlapping cell populations. Special considerations in the sorting of subpopulations consisting of less than 0.5% of the population are detailed below. Three consecutive droplets are sorted for each selected cell, one on either side of the droplet predicted to contain the cell of interest. Electronics which abort stream charging if a second cell is contained within the three droplet grouping are routinely used, unless the desired cells make up less than 0.5% of the population.

Each nozzle used for sorting will have a characteristic break-off point, droplet shaping at break off, and droplet delay requirement. These parameters should be closely monitored during an experiment and from day to day as a measure of the integrity of the nozzle orifice. The optimal drop delay should be determined for each nozzle. This is best done by assessing recovery of standardized beads which have been sorted using single drop charging.

Sterile Cell Sorting

Viable cell sorting requires the sterile maintenance of all FACS tubing carrying air or fluid into the open system. Gas pressure is maintained by a standing N2 tank with a downline 0.22 fim air filter. Sheath fluid is delivered from a 10L pressure vessel filled with sterile PBS lacking Mg2+ and Ca2+ and delivered through a 0.22 fim filter. Before and after every experiment the system is flushed with a 1:250 dilution of 7X detergent (Row Laboratories, McLean, VA) delivered from a separate pressure vessel. Before samples are run through the machine, sample delivery tubing is extensively washed with 70% ETOH.

Cells are sorted initially into DMEM, with 10% FCS, penicillin, streptomycin, and garamycin (GIBCO, Grand Island, NY) into 35 X 10-mm petri dishes. Cells rapidly sink in these petri plates and attach to the bottom surface. Microscopic observations of sorted cells should reveal a "landing pad" of several millimeters into which all the cells have settled. This indicates accurate phase alignment during sorting, but will produce a cell density problem if a large number of cells are deposited into each petri dish. For this reason it is useful to disperse the cells following sorting. Multiple petri plates are sorted for each sample and sorted cells can be expanded into T-flasks within a week. Cells are maintained in medium containing antibiotics until aliquots are placed in liquid N2 storage.

It is critical to extensively wash tubing between samples, as a limited number of cells are lodged in the tubing during an analysis and may reappear in a subsequent sample. This cross-sample contamination is an important consideration when sorting exceptionally rare cells, particularly if positive controls are run at the outset of an experiment.

Live/Dead Cell Discrimination

Cell viabilities of greater than 98% can be standardly achieved with the protocols outlined above. Dead cells in the population, however, will nonspecifically take up the fluoresceinated second antibody and produce a fluorescently positive peak of cells. These dead cells are readily distinguished from specifically stained cells by microscopy, but are indistinguishable on machine analysis. This does not pose a practical problem in sorting hybrids or transfectants generated from cloned material because of

21 J. L. Dangl, D. R. Parks, V. T. Oi, and L. A. Herzenberg, Cytometry 2, 395 (1982).

Relative Fluorescence

Fig. 2. Indirect immunofluorescence and FACS analysis of unsorted L-cell tiansfectants. Cells were stained with a mixture of anti-HLA-A, B, C (W6/32) and anti-4F2 antibodies and fluoresceinated second antibody. Detection of fluorescence was made either on live cell gated or ungated populations. FACS parameters are described in the text and cell numbers are presented on a log scale for easier visualization. The sorting window defined by the arrow includes 0.5% of the gated cells and approximately 5% of the ungated cells. The gated population is the population of cells which does not show the secondary peak of fluorescence.

Relative Fluorescence

Fig. 2. Indirect immunofluorescence and FACS analysis of unsorted L-cell tiansfectants. Cells were stained with a mixture of anti-HLA-A, B, C (W6/32) and anti-4F2 antibodies and fluoresceinated second antibody. Detection of fluorescence was made either on live cell gated or ungated populations. FACS parameters are described in the text and cell numbers are presented on a log scale for easier visualization. The sorting window defined by the arrow includes 0.5% of the gated cells and approximately 5% of the ungated cells. The gated population is the population of cells which does not show the secondary peak of fluorescence.

can be accomplished by analyzing only those cells within the region between the horizontal lines and left of the arrows labeled "A." But, it is clear from Fig. 1 that live/dead discrimination of L cells can be based on scatter alone (area between the horizontal lines bounded by the arrows at "B"). By eliminating propidium iodide staining from the analysis a fluorescence photomultiplier tube is available for quantitation of additional antigens. A comparison of gated and ungated transfected cells analyzed by indirect immunofluorescence is shown in Fig. 2. It demonstrates that the effect of removing dead cells from the analysis is to decrease the background in the region of the curve where signals from specifically stained cells are expected.

Recovery of Rare Cells in Sorting

Even with efficient live cell gating, antigen-positive cells which occur at frequencies of 0.01-0.1% in the population do not appear as a distinct peak on the first analysis (Fig. 2). It is therefore necessary to sort the brightest 0.5% of the cells to assure recovery of antigen-positive cells, and to obtain sufficient cells so that cell expansion and reanalysis can be performed rapidly. Cells to the right of the arrow in the gated population of Fig. 2 represent the 0.5% of the population sorted in this experiment.

Figure 3a presents the analysis of this population of transfectants following the initial enrichment sort. In this example, the original genomic population was stained and sorted for the expression of both human HLA and 4F2 antigens. It can be seen that two groups of antigen-positive cells are visible after the first sort, and that each is present in a different fraction of the population (Fig 3a). These cells were sorted a second time, again using a mixture of both antibodies. Following this second sort almost all the cells are positive for transfected antigens (Fig. 3b), a larger fraction expressing 4F2 rather than HLA (compare Fig. 3c and d). The data presented in Fig. 3 demonstrate that antigen expressing transfectants can be recovered with an initial frequency of 0.05% or less. In addition it illustrates the simultaneous selection of different "primary" antigen transfectants by using mixtures of antibodies. This considerably improves the overall efficiency of the procedure since the tissue culture expansion of sorted cells is particularly time consuming.

Was this article helpful?

0 0

Post a comment