Mitochondria as target to inhibit proliferation and induce apoptosis of cancer cells: the effects of doxycycline and gemcitabine

Abstract

Doxycycline has anti-tumour effects in a range of tumour systems. The aims of this study were to define the role mitochondria play in this process and examine the potential of doxycycline in combination with gemcitabine. We studied the adenocarcinoma cell line A549, its mitochondrial DNA-less derivative A549 ρ° and cultured fibroblasts. Treatment with doxycycline for 5 days resulted in a decrease of mitochondrial-encoded proteins, respiration and membrane potential, and an increase of reactive oxygen species in A549 cells and fibroblasts, but fibroblasts were less affected. Doxycycline slowed proliferation of A549 cells by 35%. Cellular ATP levels did not change. Doxycycline alone had no effect on apoptosis; however, in combination with gemcitabine given during the last 2 days of treatment, doxycycline increased caspase 9 and 3/7 activities, resulting in a further decrease of surviving A549 cells by 59% and of fibroblasts by 24% compared to gemcitabine treatment alone. A549 ρ° cells were not affected by doxycycline. Key effects of doxycycline observed in A549 cells, such as the decrease of mitochondrial-encoded proteins and surviving cells were also seen in the cancer cell lines COLO357 and HT29. Our results suggest that doxycycline suppresses cancer cell proliferation and primes cells for apoptosis by gemcitabine.

Figure 1

Doxycycline inhibits mitochondrial translation. (a) De novo mitochondrial protein synthesis of fibroblasts (Fibs), A549 and A549 ρ° cells treated with vehicle (−) or doxycycline (+). After 5 days of treatment, cultures were pulse-labelled with [³⁵S]-methionine in the presence of emetine. Protein samples were subjected to gel electrophoresis. The gel was stained with Coomassie brilliant blue followed by autoradiography to reveal the labelled mitochondrial translation products, indicated on the left. The migration of protein standards is indicated in the centre. (b) Mean MTCO2 + MTCO3 labelling signals in vehicle (Veh) and doxycycline (DC)-treated cells relative to the mean value of vehicle-treated A549 cells (n = 4). (c) Western blot images of samples from fibroblasts and A549 cells treated with doxycycline over a 5-day period and from untreated A549 ρ° cells. To facilitate quantification, serial dilutions of untreated cells, harvested at t = 0, were also applied. Blots were probed with antibodies against the indicated proteins. Migration of protein standards is indicated on the right. Full-size blots are presented in Supplementary Fig. S11. (d) Mean amounts of the indicated proteins in the treated cells over a 5-day period and of untreated A549 ρ° cells, relative to the amount in the cells at t = 0 (n = 3). Error bars indicate standard deviations. Asterisks denote statistically significant differences (p < 0.05).

Figure 2

Doxycycline decreases oxidative phosphorylation. (a) Brown histochemical staining for cytochrome-c oxidase activity in fibroblasts, A549 and A549 ρ° cells treated with vehicle or doxycycline for 5 days. Cell nuclei were counterstained purple with haematoxylin. Scale bar represents 10 μm (b,c) Mean cytochrome-c oxidase and citrate synthase activity in lysates of cells treated with vehicle (Veh) or doxycycline (DC) for 5 days (n ≥ 4). (d) Mean basal respiration, maximal respiration, spare respiratory capacity and respiration coupled to ATP production in cells treated with vehicle or doxycycline for 5 days (n ≥ 5). (e) Mean basal glycolysis, glycolytic capacity and glycolytic reserve in cells treated with vehicle or doxycycline for 5 days (n ≥ 5). Error bars indicate standard deviations. Asterisks denote statistically significant differences (*p < 0.05; **p < 0.01).

Figure 3

Doxycycline increases the inhibitory effect of gemcitabine on cell growth and the inducing effect of gemcitabine on caspase 9 and 3/7 activities. Fibroblast, A549, A549 ρ⁰, COLO357 and HT29 cells were treated with vehicle or doxycycline (DC) for 5 days. During the last 2 days, cultures were not co-treated or co-treated with gemcitabine (Gem). (a) Mean number of live cells after 5 days of treatment, relative to vehicle-treated cells (n ≥ 4). (b) Percentage of dead cells relative to total cell number after 5 days of treatment (n = 4). (c) Mean caspase 9 activity after 5 days of treatment, as determined in Caspase-Glo 9 luminescence activity assays (n = 4). (d) Mean caspase 8 activity after 5 days of treatment, as determined in Caspase-Glo 8 luminescence activity assays (n = 4). (e) Mean caspase 3/7 activity after 5 days of treatment, as determined in Caspase-Glo 3/7 luminescence activity assays (n = 4). Error bars indicate standard deviations. Asterisks denote statistically significant differences (*p < 0.05; **p < 0.01).

Figure 4

Doxycycline has no effect on cellular ATP levels synthesized through glycolysis or oxidative phosphorylation. Fibroblast, A549 and A549 ρ⁰ cultures were treated with vehicle (Veh) or doxycycline (DC) for 5 days. To determine the contribution of glycolysis and oxidative phosphorylation to the cellular ATP pool, cultures were subsequently left untreated or were treated with the oxidative phosphorylation inhibitor oligomycin A (Oligo) and/or the glycolysis inhibitor 2-deoxyglucose (2DG) directly prior to the ATP assays. Bars represent the mean cellular ATP levels (n = 4). Error bars indicate standard deviations. Asterisks denote statistically significant differences (p < 0.05).

Figure 5

Doxycycline causes mitochondrial fragmentation and decreases ΔΨ_m. Fibroblast, A549 and A549 ρ⁰ cultures were treated with vehicle (Veh) or doxycycline (DC) for 5 days, followed by staining with the ΔΨ_m-dependent dye TMRM. (a) Representative fluorescent micrographs of TMRM-stained cells. Scale bar: 10 μm. Inserts show mitochondrial morphology. (b) Mean TMRM fluorescence per mitochondrial mass per cell (n ≥ 36) in arbitrary fluorescent units (AFU). Error bars indicate standard deviations. Asterisks denote statistically significant differences (p < 0.005).

Figure 6

Doxycycline causes oxidative stress in A549 cells. (a) Western blot images of samples from fibroblasts and A549 cells treated with doxycycline over a 5-day time period and untreated A549 ρ⁰ cells. To facilitate quantification, serial dilutions of untreated cells, harvested at t = 0, were also applied. Blots were probed with antibodies against SOD2 and β-actin. Migration of protein standards is indicated on the right. Full-size blots are presented in Supplementary Fig. S12. (b) Mean amounts of SOD2 in the treated cells over a 5-day time period and untreated A549 ρ⁰ cells, relative to the amount in cells at t = 0 (n = 3). (c) Fibroblast, A549 and A549 ρ⁰ cultures were treated with vehicle (Veh) or doxycycline (DC) for 5 days, followed by staining with the ROS indicator dye DHE. Bars represent the mean DHE fluorescence increase in arbitrary fluorescent units (AFU) in the nucleus per minute (n ≥ 42). (d) Fibroblast, A549 and A549 ρ⁰ cultures were treated with vehicle or doxycycline for 5 days, followed by staining with the GSH indicator dye mBCI. Bars represent the mean mBCI fluorescence in the cell (n ≥ 33). Error bars indicate standard deviations. Asterisks denote statistically significant differences (***p < 0.005, ****p < 0.001).