Optimizing NK-92 serial killers: gamma irradiation, CD95/Fas-ligation, and NK or LAK attack limit cytotoxic efficacy

Abstract

Background: The NK cell line NK-92 and its genetically modified variants are receiving attention as immunotherapies to treat a range of malignancies. However, since NK-92 cells are themselves tumors, they require irradiation prior to transfer and are potentially susceptible to attack by patients' immune systems. Here, we investigated NK-92 cell-mediated serial killing for the effects of gamma-irradiation and ligation of the death receptor Fas (CD95), and NK-92 cell susceptibility to attack by activated primary blood NK cells.

Methods: To evaluate serial killing, we used ⁵¹Cr-release assays with low NK-92 effector cell to target Raji, Daudi or K562 tumor cell (E:T) ratios to determine killing frequencies at 2-, 4-, 6-, and 8-h.

Results: NK-92 cells were able to kill up to 14 Raji cells per NK-92 cell in 8 h. NK-92 cells retained high cytotoxic activity immediately after irradiation with 10 Gy but the cells surviving irradiation lost > 50% activity 1 day after irradiation. Despite high expression of CD95, NK-92 cells maintained their viability following overnight Fas/CD95-ligation but lost some cytotoxic activity. However, 1 day after irradiation, NK-92 cells were more susceptible to Fas ligation, resulting in decreased cytotoxic activity of the cells surviving irradiation. Irradiated NK-92 cells were also susceptible to killing by both unstimulated and IL-2 activated primary NK cells (LAK). In contrast, non-irradiated NK-92 cells were more resistant to attack by NK and LAK cells.

Conclusions: Irradiation is deleterious to both the survival and cytotoxicity mediated by NK-92 cells and renders the NK-92 cells susceptible to Fas-initiated death and death initiated by primary blood NK cells. Therefore, replacement of irradiation as an antiproliferative pretreatment and genetic deletion of Fas and/or NK activation ligands from adoptively transferred cell lines are indicated as new approaches to increase therapeutic efficacy.

Keywords: Adoptive cell transfer; Fas/CD95; Limitations; Lymphokine activated killer; NK; NK-92; Radiation; Serial killing; Therapeutic efficacy.

Fig. 1

NK-92 cell cytotoxicity and serial killing frequencies of Raji, Daudi and K562 targets. Each colored line/symbol represents %specific release at a different E:T. Dashed lines represent the expected % dead targets if one target was killed by one effector at an E:T. A Raji cells as ‘targets’. B Daudi cells as ‘targets’. C K562 cells as ‘targets’. The three targets were assayed concurrently. The standard deviations for each data point were less than 2% specific ⁵¹Cr release. The cytotoxicity towards Raji cells is significantly greater than Daudi and K562 cells (E:T 1:32 linear regression p < 0.01 & p < 0.001, respectively)

Fig. 2

Effects of irradiation on serial killing. NK-92 cells were irradiated on A the day of experiment or B the day before the experiment and cultured with IL-2. Cytotoxicity was measured after 8 h. The slopes of linear regression for B were 40.4 for the non-irradiated cells and 42.5, 37.6, and 27.5 for the 2.5, 5 and 10 Gy irradiated cells, P < 0.001 for 10 Gy. The KFs for the 1:16 E:Ts are indicated in the two boxes in the middle of the figure. E:Ts are graphed on a log₁₀ scale. [***p-value < 0.001 via regression analysis]

Fig. 3

Expression of CD95 (Fas) by NK-92 cells and effects of Fas-ligation on their viability. Fas-sensitive Jurkat cells were used as positive controls for Fas expression and death after a day of Fas ligation. Dead cells are indicated by binding of FITC-labeled annexin V to phosphatidyl serine that is externalized in the plasma membranes of dying cells. A Surface expression of Fas/CD95. 1 Expression of Fas by Jurkat cells. 1a Unlabeled cells. 1b Cells labeled with mAb anti-CD95 Fas. 2 Expression of Fas by non-irradiated NK-92 cells. 2a&2b are as indicated for 1a & 1b. 3 Expression of Fas by NK-92 cells irradiated with 10 Gy. 3a&3b are as indicated for 1a &1b. B Induction of death with control IgM or IgM anti-Fas in overnight culture. By Overton subtraction, the conversion from annexin V-low to annexin V-high cells (apoptotic and necrotic) was 52% for Jurkat cells, 5.8% for non-irradiated NK-92 cells, and 9.2% for the irradiated NK-92 cells (*p < 0.05). C Changes in cellular size (forward scatter) in response to irradiation and Fas-ligation. 1 Non-irradiated cells cultured with media, IgM isotype, or anti-Fas IgM. 2 Cells 1 day after 10 Gy irradiation, cultured with media, IgM isotype, or anti-Fas IgM

Fig. 4

Effects of CD95 (Fas)-ligation on serial killing by non-irradiated and irradiated NK-92 cells. Non-irradiated or 10 Gy irradiated cells were cultured for 1 day with nothing, 1 ug/ml control IgM or IgM anti-Fas. Because K562 cells lack Fas, they were used as tumor ‘target’ cells to prevent addition of anti-Fas-initiated “suicide” to target cell killing by NK-92 cells. Cytotoxic activity towards K562 cells was measured after 8 h. E:T ratios are approximate. A Anti-Fas ligation limited to non-irradiated NK-92 cells. Insert: LU₅₀/1 M non-irradiated NK-92 cell, by treatment. B Anti-Fas ligation of 10 Gy irradiated NK-92 cells. Insert: LU50/1 M irradiated NK-92 cell, by treatment. [***p-value < 0.001 via regression analysis]. KFs A, B The KFs calculated for a 1:1 E:T for A, the non-irradiated control alone, with IgM or with anti-Fas were 0.36, 0.34, and 0.22, respectively. For B, the KFs for the irradiated cells alone, with IgM or with anti-Fas were 0.30, 0.30 and 0.03 respectively. The loss of KF for the irradiated and anti-Fas combination exceeded both additive and synergistic effects estimated based on the losses from single treatments

Fig. 5

Killing of NK-92 cells by either resting or activated primary blood NK cells. The susceptibility of NK-92 tumor cells to attack by primary peripheral blood NK cells from healthy donors was investigated using NK-92 cells that were either non-irradiated or were irradiated with 10 Gy and then cultured for 1 day. The primary NK cells were either resting or activated by 3 days culture with 1000 u/ml IL-2. Cytotoxicity was measured using 4-h ⁵¹Cr-release assays; the gray dashed line reflects the 10% threshold for positive killing. The results illustrated represent concurrent assays made with E:Ts of viable effector NK cells from a single donor SC-0975 and are representative of 4 experiments each with a different donor. A Killing of NK-92 cells by resting primary NK cells. LU₅₀s were excluded for unstimulated NK cells because killing was over 10% for only the highest E:T. B Killing of NK-92 cells by activated primary NK cells. For both bNKs and LAKs, the cytotoxicity towards non-irradiated versus irradiated NK-92 cells was significantly different at the highest E:T (student t-test, *p < 0.03, **p < 0.005). Insert for B LU50/1 M LAK bNK calculated using TruCOUNT™ beads (Gating—Additional file 2: Fig. S1)