Cytokines impact natural killer cell phenotype and functionality against glioblastoma in vitro

Abstract

Objective: Natural killer (NK) cells are a part of the innate immune system and first-line defense against cancer. Since they possess natural mechanisms to recognize and kill tumor cells, NK cells are considered as a potential option for an off-the-shelf allogeneic cell-based immunotherapy. Here, our objective was to identify the optimal cytokine-based, feeder-free, activation and expansion protocol for cytotoxic NK cells against glioblastoma in vitro.

Methods: NK cells were enriched from human peripheral blood and expanded for 16 days with different activation and cytokine combinations. The expansion conditions were evaluated based on NK cell viability, functionality, expansion rate and purity. The cytotoxicity and degranulation of the expanded NK cells were measured in vitro from co‑cultures with the glioma cell lines U‑87 MG, U‑87 MG EGFR vIII, LN-229, U-118 and DK-MG. The best expansion protocols were selected from ultimately 39 different conditions: three magnetic cell‑selection steps (Depletion of CD3+ cells, enrichment of CD56+ cells, and depletion of CD3+ cells followed by enrichment of CD56+ cells); four activation protocols (continuous, pre-activation, re-activation, and boost); and four cytokine combinations (IL-2/15, IL‑21/15, IL‑27/18/15 and IL-12/18/15).

Results: The expansion rates varied between 2-50-fold, depending on the donor and the expansion conditions. The best expansion rate and purity were gained with sequential selection (Depletion of CD3+ cells and enrichment of CD56+ cells) from the starting material and pre-activation with IL‑12/18/15 cytokines, which are known to produce cytokine-induced memory-like NK cells. The cytotoxicity of these memory-like NK cells was enhanced with re-activation, diminishing the donor variation. The most cytotoxic NK cells were produced when cells were boosted at the end of the expansion with IL-12/18/15 or IL-21/15.

Conclusion: According to our findings the ex vivo proliferation capacity and functionality of NK cells is affected by multiple factors, such as the donor, composition of starting material, cytokine combination and the activation protocol. The cytokines modified the NK cells' phenotype and functionality, which was evident in their reactivity against the glioma cell lines. To our knowledge, this is the first comprehensive comparative study performed to this extent, and these findings could be used for upscaling clinical NK cell manufacturing.

Keywords: cytokines; glioblastoma; immunotherapy; memory-like natural killer cells; natural killer cells.

Figure 1

Comparison between different selection methods and cytokine combinations. NK cells were analyzed after 16 days expansion for their expansion rate, purity, and cytotoxicity against U87-wt glioblastoma cell line. (A–C) Comparison between three selection methods [depletion of CD3+ cells (CD3-, square), enrichment of CD56+ cells (CD56+, circle), and depletion of CD3+ cells followed by enrichment of CD56+ cells (CD3-/CD56+, triangle)]. (A) Purity of CD45+/CD3-/CD56+ NK cells in the population after ex vivo expansion. (B) Cytotoxicity against U87-wt with 1:1 E:T ratio. (C) Expansion rate calculated by dividing the output number of expanded NK cells after 16 days of culture by the NK cell number on day 0. (D–L) Comparison of different expansion methods (1-10; Supplementary Table 3 ) within the selection methods. Each symbol represents one expansion, all together 25 donors were used, minimum 2 donors per condition. If one datapoint or NA the measurement was not performed due to low yield in that expansion. Numbers 1-10 represent the different expansion conditions (1.IL-21/15 Pre, 2.IL-21/15 Re, 3.IL-21/15 Boost, 4.IL-27/18/15 Pre, 5.IL-27/18/15 Re, 6.IL-27/18/15 Boost, 7.IL-12/18/15 Pre, 8.IL-12/18/15 Re, 9.IL-12/18/15 Boost, 10.IL-2/15 Cont; Pre=pre-activation, Re, re-activation; Boost, activation on day 15; Cont, Continuous). Pattern indicates the activation timing: Pre-activation = horizontal stripes, Re-activation = diagonal stripes, Boost = dotted, Continuous = blank. Color indicates the cytokine combination: IL-21/15 = purple, IL-27/18/15 = white, IL-12/18/15 = blue, IL-2/15 = green. NA, Not available. Data shown as mean ± SEM. ***p< 0.001, **p< 0.01, *p< 0.05, ns, non significant.

Figure 2

Histograms of flow cytometry data comparing expression of phenotype markers CD69, CD25, CD16, and NKp46 after CD3-/CD56+ selection (Day 0) and 16 days expansion with different expansion protocols (1-10). (A) Representative example from the gating strategy for viable (7-AAD-) CD45+/CD3-/CD56+ NK cells. Representative histograms shown from a single donor (B–E). (B) CD69 expression, (C) CD25 expression, (D) CD16 expression, (E) NKp46 expression. (F) Group details and median fluorescence intensity (MFI) of the measured markers. Pre, pre-activation; Re, re-activation; Boost, activation on day 15; Cont, Continuous; NS, Non stained.

Figure 3

NK cell functionality against the glioblastoma cell line U87-wt after expansion with different conditions (1-10). (A) NK cell cytotoxicity with 0.5:1, 1:1, or 2:1 E:T ratios. (B) CD107a expression (degranulation) on CD45+ cells after 3 hour co-culturing (1:1 E:T). (C) Expansion rate calculated by dividing the output number of expanded NK cells after 16 days of culture by the NK cell number on day 0. (D) NK cell cytotoxicity (basal) compared to cytotoxicity with additional IL-2 (20 IU/ml) or IL-15 (5 ng/ml) in the co-cultures (1:1 E:T). E:T, Effector to target ratio. Numbers 1-10 represent the different expansion conditions (1.IL-21/15 Pre, 2.IL-21/15 Re, 3.IL-21/15 Boost, 4.IL-27/18/15 Pre, 5.IL-27/18/15 Re, 6.IL-27/18/15 Boost, 7.IL-12/18/15 Pre, 8.IL-12/18/15 Re, 9.IL-12/18/15 Boost, 10.IL-2/15 Cont; Pre=pre-activation, Re, re-activation; Boost, activation on day 15; Cont, Continuous). Pattern indicates the activation timing: Pre-activation = horizontal stripes, Re-activation = diagonal stripes, Boost = dotted, Continuous = blank. Color indicates the cytokine combination: IL-21/15 = purple, IL-27/18/15 =white, IL-12/18/15 = blue, IL-2/15 =green. Data are shown from 3 donors with three technical replicates, identified by symbol shape and color. ***p< 0.001, **p< 0.01, *p< 0.05.

Figure 4

Multiple linear regression and Principal components analysis (PCA) from NK cell expansion rates, purity, cytotoxicity and phenotype markers (CD16, CD25, CD69, and NKp46), and degranulation (CD107a). (A) Multiple linear regression. The cytotoxicity of continuous IL-2/15 expanded NK cells was considered as a reference level. (B) Eigenvectors showing the variance in the input data. (C) Principal component (PC) scores indicating the expansion protocol and cytokines by color and the expansion rate by size. (D) PC indicating the cytotoxicity in color scale, yellow being the highest cytotoxicity rate and the expansion rate by size. Data collected from 3 donors. **p< 0.01, *p< 0.05, ns= nonsignificant.

Figure 5

Comparison of histograms for the functionality markers Fas-ligand (FASL), natural killer group 2D (NKG2D), TNF-related apoptosis-inducing ligand (TRAIL), and lymphocyte function-associated antigen 1 (ITGAL) expression on CD45+/CD3-/CD56+ NK cells. NK cells, CD56+/CD3- selected, were expanded for 16 days with IL-12/18/15 (pre, re, and boost) and IL-2/15 (Cont). Representative histograms shown from a single donor (A–D). (A) FASL-expression. (B) NKG2D expression. (C) TRAIL expression. (D) ITGAL expression. (E) Group details and median fluorescence intensity (MFI) of the measured markers. Pre, Pre-activation; Re, Re-activation; Boost, cytokine activation at the end of expansion. Cont, Continuously.

Figure 6

NK cell functionality against five glioma cell lines: U87-wt, U87vIII, DK-MG, U-118, and LN-229 after 16 days expansion with IL-12/18/15 (pre, re, and boost) and IL-2/15 (Cont). (A) NK cell cytotoxicity at 1:1 E:T ratio against the glioma cell lines. (B) CD107a expression (degranulation) on CD45+ cells after 3 hour co-culturing (1:1 E:T) with the glioma cell lines. Controls represent the basal expression level without co-culturing. Numbers 7-10 represent the different expansion conditions (7.IL-12/18/15 Pre, 8.IL-12/18/15 Re, 9.IL-12/18/15 Boost, 10.IL-2/15 Cont; Cont; Pre, pre-activation; Re, re-activation; Boost, activation on day 15; Cont, Continuous). E:T, Effector to target ratio. Data from two donors with three technical replicates represented as the mean ±SEM. ***p< 0.001, **p< 0.01, *p< 0.05.