The Autophagy Inhibitor Bafilomycin Inhibits Antibody-Dependent Natural Killer Cell-Mediated Killing of Breast Carcinoma Cells

Abstract

The resistance of breast cancer cells to therapeutic antibodies such as anti-HER2 trastuzumab can be overcome by engaging natural killer (NK) cells for killing antibody-binding tumor cells via antibody-dependent cellular cytotoxicity (ADCC). Here, we investigated how autophagy modulation affects trastuzumab-mediated ADCC in HER2-positive JIMT1 breast cancer cells and NK cells. Autophagy inducers (rapamycin and resveratrol) had no significant impact, but the inhibitor bafilomycin nearly abolished ADCC. Protection occurred when either cancer or NK cells were pretreated, indicating dual effects. Bafilomycin reduced phosphatidylserine externalization, the loss of plasma membrane integrity, caspase-3/7 activity, and DNA fragmentation. It downregulated pro-apoptotic BAK1 and BAX without altering BCL-2. Additionally, bafilomycin decreased HER2 surface expression, impairing trastuzumab binding, and modulated immune regulators (STAT1, CD95, and PD-L1) in NK and/or in the cancer cells. Bafilomycin disrupted HER2 trafficking and induced HER2 internalization, leading to its accumulation in cytoplasmic vesicles. These findings show that autophagy inhibition by bafilomycin confers ADCC resistance by altering apoptosis, immune signaling, and HER2 dynamics. The study underscores autophagy's role in antibody-based cancer therapy efficacy.

Keywords: antibody dependent cell-mediated cytotoxicity; apoptosis; autophagy; natural killer cell; trastuzumab.

Figure 1

Bafilomycin A1 effectively inhibits antibody-dependent cellular cytotoxicity. Live cell imaging of JIMT1 cells in ADCC with NK92CD16 cells at a 2:1 effector–target ratio for 3 h. (A) JIMT1 cells in 96-well HCS plates were pretreated with 500 nM rapamycin, 500 nM bafilomycin A1, or 50 μM resveratrol for 20 h, then stained with Calcein-AM. NK cells were added in the presence of trastuzumab, and images were taken immediately after NK addition (0 h) and 3 h later. (Green: Calcein-AM staining of JIMT1 cells). (B) JIMT1 cells were seeded into 96-well HCS plates. NK92CD16 cells were pretreated with 500 nM rapamycin, 500 nM bafilomycin A1, or 50 μM resveratrol for 20 h. JIMT1 cells were stained with Calcein-AM before NK cells and trastuzumab were added, and images were taken immediately after NK addition (0 h) and 3 h later. (Green: Calcein-AM staining of JIMT1 cells). (C) JIMT1 cells were pretreated with autophagy modulators and cell confluence was determined by image analysis at 0 h and 3 h of ADCC. (D) NK cells were pretreated with autophagy modulators and JIMT1 cell confluence was determined by image analysis at 0 h and 3 h of ADCC. All plots represent at least three independent experiments; values are given as mean ± SD (*** p < 0.001) (scale bar is 200 µm).

Figure 2

Bafilomycin A1 pretreatment of JIMT1 cells prevents apoptotic cell death and decreases membrane permeability in ADCC. (A) Annexin V—SYTOX Green staining of JIMT1 cells after 3 h ADCC. JIMT1 cells were stained with CellTracker Blue and pretreated with 500 nM bafilomycin A1 or vehicle for 20 h. After 3 h of ADCC, cells were stained with Annexin V—Alexa 647 (red) and SYTOX Green. (B) Image analysis and evaluation of Annexin V binding and SYTOX Green staining in JIMT1 cells. Alexa 647 and SYTOX Green intensity values are normalized to JIMT1 cell region area. All plots represent three independent experiments; values are given as mean ± SD (*** p < 0.001) (scale bar is 200 µm).

Figure 3

Effect of bafilomycin A1 on caspase activation and the expression of BCL-2 family genes in JIMT1 cells. (A) Evaluation of caspase 3/7 activity by live cell imaging in JIMT1 cells during ADCC. JIMT1 cells in 96-well HCS plates were stained with CellTracker Blue and pretreated with 500 nM bafilomycin A1 or vehicle for 20 h. Apoptotic cells were labelled with CellEvent Caspase-3/7 Green and images were taken after 3 h ADCC. The intensity of the fluorescent caspase substrate was evaluated with image analysis. (B) JIMT1 cells were treated with 500 nM bafilomycin A1 or vehicle for 24 h. mRNA expression levels of BCL-2 family genes in JIMT1 cells were determined by RT-qPCR. Expression levels were normalized to cyclophilin A and GAPDH and presented as fold change compared to control. All plots represent three independent experiments; values are given as mean ± SD (*** p < 0.001; ns = not significant). (C) DNA strand breaks were detected in ADCC cultures with indirect TUNEL method. JIMT1 cells were treated with 500 nM bafilomycin A1 or vehicle for 20 h before NK cells were added in the presence of trastuzumab for 3 h (at 2:1 effector–target ratio). TUNEL-positive area ratio was determined by image analysis. Plots represent three independent experiments; values are given as mean ± SD (*** p < 0.001) (scale bar is 100 µm).

Figure 4

Bafilomycin A1 decreases cell surface HER2 expression and trastuzumab binding in JIMT1 cells. Trastuzumab (Tr) binding of JIMT1 and NK cells was analyzed by flow cytometry. (A) JIMT1 cells were treated with 500 nM bafilomycin A1 or vehicle for 24 h (upper) or for the indicated times (lower) and labelled with Alexa 488-conjugated trastuzumab. (B) NK cells were treated with 500 nM bafilomycin A1 or vehicle for 24 h and labelled with Alexa 488-conjugated trastuzumab. Alexa 488 intensity median values are presented as fold change compared to control. All plots represent three independent experiments; values are given as mean ± SD (** p < 0.01; *** p < 0.001; ns = not significant).

Figure 5

Bafilomycin A1 modulates the expression of genes involved in ADCC. JIMT1 and NK cells were treated with 500 nM bafilomycin A1 or vehicle for 24 h. mRNA levels of genes involved in NK cell function were determined by RT-qPCR. Expression levels were normalized to cyclophilin A and GAPDH and presented as fold change compared to control. All plots represent three independent experiments; values are given as mean ± SD. (* p < 0.05, *** p < 0.001, ns = not significant).

Figure 6

Bafilomycin A1 reduces HER2 surface expression by blocking HER2 intracellular trafficking. (A) Confocal microscopic images of HER2 staining in JIMT1 cells. JIMT1 cells in 96-well HCS plates were treated with 500 nM bafilomycin A1 or vehicle for 24 h. Cells were fixed and stained with FITC-labelled anti-HER2 antibody, and DAPI. Confocal images were taken with the Opera Phenix HCA equipment. (B) Image analysis and evaluation of HER2 localization in JIMT1 cells. FITC intensity values were analyzed in the membrane region and cytoplasmic region. The number of HER2+ cytoplasmic vesicles indicates altered intracellular trafficking. (C) Image analysis of nuclear morphology of JIMT1 cells. All plots represent three independent experiments; values are given as mean ± SD (*** p < 0.001; ns = not significant) (scale bar is 50 µm).

Figure 7

Bafilomycin A1 induces HER-2 internalization. (A) Confocal microscopic images were obtained to visualize HER2 live staining in JIMT1 cells. JIMT1 cells seeded in 96-well HCS plates were stained with DRAQ5 Fluorescent Probe. After washing, FITC-labelled anti-HER2 antibody was added to the cells, either with 500 nM bafilomycin A1 or vehicle and incubated for 24 h. Confocal images were taken using the Opera Phenix HCA equipment immediately after treatment, as well as 3 h and 24 h post-treatment. (B) Image analysis was used to evaluate HER2 localization in JIMT1 cells. FITC-stained spots were analyzed in the cytoplasmic region of cells. HER2-containing vesicles were observable after 3 h in bafilomycin A1-treated cells. Notably, 24 h bafilomycin A1 treatment dramatically increased the number of HER2-FITC-containing vesicles, indicating that HER2 was internalized. All plots represent three independent experiments; values are given as mean ± SD (** p < 0.01, *** p < 0.001) (scale bar is 50 µm).