Introduction

As targets, we used monolayers of the immortalized murine fibroblast-like 10T1/2 cell line (Table 1) because it forms homogenous, confluent monolayers of extensively spread cells. As effectors, we used murine and human lymphokine activated killer (LAK) cells. Nylon wool nonadherent mouse spleen cells were treated with IL-2 or IL-15 for 5 days, or with IL-2 and IL-12 for 5 and 3 days, respectively (Geldhof et al, 1998). Human mononuclear cells were prepared by density-gradient centrifugation of peripheral blood from healthy donors. The cells were cultured for 3 days with recombinant IL-2 (Whiteside, 1996). Inoculated on 10T1/2 monolayers in 10 cm2 tissue culture dishes (Falcon, Sunnyvale, California) was 1 ml of LAK cell suspensions at different effector over target (E/T) ratios. Within 15 minutes after the inoculation of LAK cells on the monolayers, individual 10T1/2 cells contracted rapidly, leaving openings in the monolayer. We observed blebbing on the surface of the contracted cells and, eventually, fragmentation. The destruction of a 10T1/2 cell was completed in about 10 minutes. After 1 hour, no more 10T1/2 cells were seen to contract. The interaction could be videotaped using a microscope in a cabinet at 37° C. An animation showing these events is available at www.uscapjournals.org.

Table 1 Evaluation of Different Target Cell Populations for Use in the Morphometric Cytotoxicity Assay (Mo.C.A.)
Full size table

Routinely, the assay was stopped after 2 hours of interaction by twice adding 1 ml of Coomassie blue solution (0.2% R250 Coomassie Brilliant blue [Sigma, St. Louis, Missouri] in acetic acid/methanol/water [10/45/45: v/v/v]). After 15 minutes the dishes were abundantly rinsed with water and air-dried for storage and later analysis. Most of the dead 10T1/2 cells and their debris, as well as unattached LAK cells, disappeared during the rinses (Fig. 1, a to c). Low power microscopic video images of the stained monolayers can be imported into any commercial or public domain computer program for image analysis. We used a × 6.3 objective, and the NIH image software (National Institutes of Health, available at: http://rsb.info.nih.gov/nih-image/). After digital contrast enhancement, the gray level images were converted into a binary image, subcellular details scrapped, and the ratio between the resulting white (cleared) and black (confluent) areas (Fig. 1, d to f) used as a parameter of monolayer lysis: “% clearance” is the cleared cell-free area as a percentage of the total area of the microscope field (averaged for 10 fields per condition).

Figure 1
figure 1

Detection of lymphokine activated killer cell lysis in cultured target cell monolayer using the morphometric cytotoxicity assay (Mo.C.A.). a to c, Coomassie blue stained monolayers of 10T1/2 cells at high magnification. a, Confluent monolayer before addition of lymphokine-activated killer (LAK) cells; b, Monolayer incubated with 106 LAK (L) cells for 2 hours; c, Monolayer incubated with 2.106LAK (L) cells for 2 hours. Bar = 25 μm. d to f, Image processing of Coomassie blue stained 10T1/2 monolayers after incubation with IL-2-activated killer cells. Arbitrary microscopic fields (objective magnification × 6.3) are imported to a personal computer. d, The picture is converted to obtain a binary image; e, Clear and confluent areas are homogenized by excluding objects smaller than a cell (f). Bar = 140 μm.

Full size image

With syngeneic (C3H) IL-2 activated LAK cells, lysis of 10T1/2 cells was noticeable from an E/T ratio of 0.75/1 and reached a maximum of approximately 50% at a ratio of 6/1 (Fig. 2). The same effectors and targets were tested in parallel by the classical 111In-release method (Geldhof et al, 1995). The Mo.C.A. is clearly more sensitive, as we needed a fivefold higher E/T ratio to obtain comparable levels of lysis by the 111In-release method (30% specific release with an E/T ratio of 20/1). Murine LAK cells from different haplotypes had similar lytic activity in the Mo.C.A. Human LAK cells caused a slightly lower clearance (Fig. 2). We also evaluated the activation status of killer cells activated by different lymphokine treatments. LAKs activated with IL-2 plus IL-12 were significantly more cytolytic to 10T1/2 than IL-2- or IL-15-activated LAKs. This distinction was also seen, but less pronounced, in the 111In-release assay.

Figure 2
figure 2

Analysis of the sensitivity of 10T1/2 to different LAK populations by the Mo.C.A. The graph shows the effect of syngeneic (C3H mice, H-2k haplotype [Harlan CPB]), allogeneic (AKR, C57BL/6, and BALB/c mice; H-2k, H-2b, and H-2d haplotype, respectively [Harlan CPB]), and xenogeneic (human) effector cells. Confluent 10T1/2 monolayers were incubated with LAK cells at different E/T ratio’s for 2 hours. The remaining adherent target cells were fixed and stained with Coomassie blue. Percent clearance was measured in 20 different arbitrary microscopic fields (10 fields in duplicate dishes). Mean values and sem of 3 to 8 experiments, with linear regression plots.

Full size image

As alternative target cell lines, we used epithelial, endothelial, and fibroblast-like cell lines from different species (Table 1). We found that many types of cell lines can be used in the Mo.C.A. provided they form uniform monolayers in culture, but their sensitivity to LAK cell-mediated lysis varied widely: from 60% (BAEC) to 0% (MDCK). In mixed target cell monolayers (prepared as in Verschueren et al, 1991), LAK cells killed 10T1/2 cells very efficiently, whereas MDCK cells were left intact, showing that soluble activating factors produced by the sensitive 10T1/2 or inactivating factors produced by the resistant MDCK cells were not responsible for the difference in sensitivity. Other interesting targets included ICAM-1 deficient endothelial cell monolayers (more sensitive than the wild type) and B7-1 transfected 3T3 cells (more sensitive than the wild type).

An additional opportunity with the Mo.C.A. is that the LAK cells could be recovered from the interacting cultures by simple washings. Thus, we could show by Annexin V-FITC and propidium iodide staining, followed by fluorescent activated cell sorter analysis (Vermes et al, 1995), that the level of apoptosis in LAK cells was increased from 12% to 32% after interaction with 3T3 targets.

In conclusion, the Mo.C.A. provides for a very simple, sensitive, and versatile method for the assessment of cytolytic activity of killer cells. It may facilitate routine analysis (eg, in clinical settings), but it may also offer new opportunities for the fundamental analysis of the interaction between cytotoxic cells and their targets.