Vitamin C and Doxycycline: A synthetic lethal combination therapy targeting metabolic flexibility in cancer stem cells (CSCs)

Abstract

Here, we developed a new synthetic lethal strategy for further optimizing the eradication of cancer stem cells (CSCs). Briefly, we show that chronic treatment with the FDA-approved antibiotic Doxycycline effectively reduces cellular respiration, by targeting mitochondrial protein translation. The expression of four mitochondrial DNA encoded proteins (MT-ND3, MT-CO2, MT-ATP6 and MT-ATP8) is suppressed, by up to 35-fold. This high selection pressure metabolically synchronizes the surviving cancer cell sub-population towards a predominantly glycolytic phenotype, resulting in metabolic inflexibility. We directly validated this Doxycycline-induced glycolytic phenotype, by using metabolic flux analysis and label-free unbiased proteomics. Next, we identified two natural products (Vitamin C and Berberine) and six clinically-approved drugs, for metabolically targeting the Doxycycline-resistant CSC population (Atovaquone, Irinotecan, Sorafenib, Niclosamide, Chloroquine, and Stiripentol). This new combination strategy allows for the more efficacious eradication of CSCs with Doxycycline, and provides a simple pragmatic solution to the possible development of Doxycycline-resistance in cancer cells. In summary, we propose the combined use of i) Doxycycline (Hit-1: targeting mitochondria) and ii) Vitamin C (Hit-2: targeting glycolysis), which represents a new synthetic-lethal metabolic strategy for eradicating CSCs. This type of metabolic Achilles' heel will allow us and others to more effectively "starve" the CSC population.

Keywords: cancer stem-like cells (CSCs); doxycycline; mitochondrial DNA (mt-DNA); mitochondrial biogenesis; vitamin C.

Figure 1. Generating MCF7 DoxyR cells

Doxycycline-resistant (DoxyR) MCF7 cells were generated by serially passaging MCF7 cells, in the presence of increasing step-wise concentrations of Doxycycline (12.5, 25 and 50 μM), over a period of 9 weeks. See the Materials and Methods section for further details. Unless stated otherwise, MCF7 cells resistant to 25 μM Doxycycline were utilized for experiments, such as unbiased proteomics analysis.

Figure 2. MCF7 DoxyR cells exhibit an increase in mitochondrial mass

A.-D. MCF7 cells were treated with DMSO or Doxycycline for acute (48 h) and chronic stimulation (3 weeks), as specified in Materials and Methods, and then mitochondrial mass was quantitated by FACS analysis using the probe MitoTracker Deep-Red (640-nm). Note that MCF7 cells chronically treated with 12.5 μM (A., fold change 1.33), 25 μM (B., fold change 1.68) and 50 μM (C., fold change 1.36) Doxycycline show a significant increase in mitochondrial mass compared to MCF7 cells treated with vehicle. Data shown are the mean ± SEM of at least 3 independent experiments performed in triplicate. (**) p < 0.01; (***) p < 0.001. D. Representative plots showing increased mitochondrial mass in MCF7 DoxyR cells as compared to MCF7 cells. E. Evaluation of the mitochondrial protein TOMM20 in MCF7 and MCF7 DoxyR cells by western blotting. Side panel shows densitometric analysis of the blots normalized to β-actin. Data shown are the mean ± SEM of 3 independent experiments. (**) p < 0.01.

Figure 3. Mitochondrial respiration is inhibited in MCF7 DoxyR cells

The metabolic profile of MCF7 DoxyR cells monolayers chronically treated with increasing concentrations of Doxycycline (12.5 μM ÷ 50 μM), as described in Materials and Methods, was assessed using the Seahorse XF-e96 analyzer. A. Representative tracing of metabolic flux. Dose-dependent significant reduction in basal respiration, proton leak, maximal respiration, ATP levels and spare respiratory capacity were observed B. Data shown are the mean ± SEM of 3 independent experiments performed in sextuplicate. (*) p < 0.05; (**) p < 0.01; (***) p < 0.001.

Figure 4. Glycolysis is increased in MCF7 DoxyR cells

The metabolic profile of MCF7 DoxyR cells monolayers chronically treated with increasing concentrations of Doxycycline (12.5 μM ÷ 50 μM), as described in Materials and Methods, was assessed using the Seahorse XF-e96 analyzer. A. Representative tracing of metabolic flux. B. Dose-dependent significant increase in glycolysis and decrease in glycolytic reserve as well as glycolytic reserve capacity were observed. Data shown are the mean ± SEM of 3 independent experiments performed in sextuplicate. (*) p < 0.05; (**) p < 0.01; (***) p < 0.001.

Figure 5. MCF7 DoxyR cells show increased CSC markers

48h after seeding, MCF7 and MCF7 DoxyR cells were processed for the evaluation of ALDEFLUOR activity, an independent marker of CSCs. Each sample was normalized using diethylaminobenzaldehyde (DEAB), a specific ALDH inhibitor, as negative control A. The tracing of representative samples is shown B. 48h after seeding, MCF7 and MCF7 DoxyR cells were re-plated on low-attachment plates, for anoikis assay for 10 hours. Expression of CSC markers (CD24 and CD44) was analysed by FACS C. Representative dot plot for the the CD44+/CD24low cell population is shown D. This represents an ∼10-fold increase in ALDH functional activity and a ∼3-fold induction of the CD44+/CD24low population. Data are the mean ± SEM of 3 independent experiments performed in triplicate. (***) p < 0.001.

Figure 6. Mammosphere formation is inhibited in MCF7 DoxyR cells: Targeting DoxyR cells with Atovaquone and Chloroquine

Evaluation of mammosphere formation in MCF7 and MCF7 DoxyR cells cultured in low attachment plates and treated with vehicle or the selective OXPHOS inhibitor Atovaquone (ATO) A. or Chloroquine B. (which has been shown to impair mitochondrial metabolism), for 5 days before counting. Note that sphere formation is inhibited in MCF7 DoxyR cells as compared to MCF7 cells. In addition, mitochondrial-targeting agents like atovaquone and Chloroquine were effective in reducing the number of spheres in both MCF7 and MCF7 DoxyR cells. Data shown are the mean ± SEM of 3 independent experiments performed in triplicate. (***) p < 0.001.

Figure 7. MCF7 DoxyR cells show a quiescent phenotype, with significantly reduced proliferation and cell migration, as well as suppression of ERK- and AKT-signaling

Evaluation of cell proliferation by EdU incorporation assay using FACS analysis in MCF7 and MCF7 DoxyR cells 48h after seeding A. Note the reduction in EdU positive population in MCF7 DoxyR cells as compared to MCF7 cells. The tracing of a representative sample is shown B. Data shown are the mean ± SEM of 4 independent experiments performed in triplicate. (***) p < 0.001. Evaluation of cell migration by wound healing assay in MCF7 and MCF7 DoxyR cells which were seeded in 6 well plate to create a confluent monolayer. 24h after seeding a wound was created, then cells were washed and incubated at 37°C for 24 h. Images were acquired at 0 h and 24 h using Incucyte Zoom (Essen Bioscience). Quantification of cell migration was performed using ImageJ software and was expressed as % of wound closure C. Note the low migratory capacity of MCF7 DoxyR cells as compared to MCF7 cells. Representative images showing scratch assay D. Bar scale 100 μm. Data shown are the mean ± SEM of 3 independent experiments performed in triplicate. (***) p < 0.001. Evaluation of ERK1/2 E. and AKT Ser 473 F. phosphorylation in MCF7 and MCF7 DoxyR cells by western blotting. Side panels show densitometric analysis of the blots normalized to ERK2 and AKT respectively. Data shown are the mean ± SEM of 3 independent experiments. (*) p < 0.05; (**) p < 0.01.

Figure 8. A two-hit synthetic lethal strategy for eradicating DoxyR CSCs

Here, we outline a new therapeutic strategy for targeting CSCs. Our experimental results indicate that DoxyR cells acquire a predominantly glycolytic phenotype, to escape the anti-mitochondrial effects of Doxycycline. As such, they should be extremely sensitive to additional metabolic stressors, allowing them to be eliminated completely. This immediately suggests a new synthetic lethal strategy for the metabolic eradication of CSCs, to avoid resistance to Doxycycline. Specifically, if we consider DoxyR as the first metabolic Hit in a two-Hit scheme, then DoxyR cells should be extremely susceptible to a second metabolic Hit. This second metabolic Hit could be achieved by using virtually any other “safe” metabolic inhibitors, targeting either glycolysis, OXPHOS or autophagy.

Figure 9. Metabolic inhibitors successfully employed for the eradication of DoxyR CSCs

Briefly, a list of small molecules that we successfully used in conjunction with Doxycycline is shown. These include 9 known inhibitors of OXPHOS, glycolysis and autophagy. Two natural products (Vitamin C and Berberine), six clinically-approved drugs (Atovaquone, Chloroquine, Irinotecan, Sorafenib, Niclosamide, and Stiripentol) and one experimental drug (2-DG), are all highlighted.

Figure 10. Glycolysis inhibitors reduce mammosphere formation in MCF7 DoxyR cells

Evaluation of mammosphere formation in MCF7 and MCF7 DoxyR cells cultured in low attachment plates and treated with Vehicle or increasing concentrations of the glycoysis inhibitor 2-deoxy-glucose (2 DG) (10 mM to 20 mM) for 5 days before counting A. Mammosphere formation is inhibited in MCF7 DoxyR cells cultured in low attachment plates and treated with increasing concentrations of the glycoysis inhibitor Ascorbic Acid (100 μM to 500 μM) for 5 days before counting B. Data shown are the mean ± SEM of 3 independent experiments performed in triplicate. (***) p < 0.001.

Figure 11. A panel of clinically-approved drugs inhibits mammosphere formation in MCF7 DoxyR cells

Evaluation of mammosphere formation in MCF7 DoxyR cells cultured in low attachment plates and treated with Vehicle or increasing concentrations of the LDH inhibitor Stiripentol (2 μM to 100 μM) A. or the OXPHOS inhibitors Irinotecan (500 nM to 80 μM) B., Sorafenib (500 nM to 40 μM) C., Berberine Chloride (500 nM to 10 μM) D. and Niclosamide E.-F. for 5 days before counting. Data shown are the mean ± SEM of 3 independent experiments performed in triplicate. (**) p < 0.01: (***) p < 0.001.

Figure 12. Vitamin C and Doxycycline: A synthetic lethal combination therapy for eradicating CSCs

Note that both OXPHOS and the glycolytic pathway jointly contribute to ATP production. Doxycycline inhibits mitochondrial biogenesis and OXPHOS, by acting via mitochondrial ribosomal proteins (MRPs); Vitamin C inhibits glycolytic metabolism by targeting and inhibiting the enzyme GAPDH. Therefore, their use together, as a sequential drug combination, will more severely target cell metabolism and energy production, thereby preventing or blocking the propagation of CSCs.