Enterobactin, an Iron Chelating Bacterial Siderophore, Arrests Cancer Cell Proliferation

Abstract

Iron is essential for many biological functions, including being a cofactor for enzymes involved in cell proliferation. In line, it has been shown that cancer cells can perturb their iron metabolism towards retaining an abundant iron supply for growth and survival. Accordingly, it has been suggested that iron deprivation through the use of iron chelators could attenuate cancer progression. While they have exhibited anti-tumor properties in vitro, the current therapeutic iron chelators are inadequate due to their low efficacy. Therefore, we investigated whether the bacterial catecholate-type siderophore, enterobactin (Ent), could be used as a potent anti-cancer agent given its strong iron chelation property. We demonstrated that iron-free Ent can exert cytotoxic effects specifically towards monocyte-related tumor cell lines (RAW264.7 and J774A.1), but not primary cells, i.e. bone marrow-derived macrophages (BMDMs), through two mechanisms. First, we observed that RAW264.7 and J774A.1 cells preserve a bountiful intracellular labile iron pool (LIP), whose homeostasis can be disrupted by Ent. This may be due, in part, to the lower levels of lipocalin 2 (Lcn2; an Ent-binding protein) in these cell lines, whereas the higher levels of Lcn2 in BMDMs could prevent Ent from hindering their LIP. Secondly, we observed that Ent could dose-dependently impede reactive oxygen species (ROS) generation in the mitochondria. Such disruption in LIP balance and mitochondrial function may in turn promote cancer cell apoptosis. Collectively, our study highlights Ent as an anti-cancer siderophore, which can be exploited as an unique agent for cancer therapy.

Graphical Abstract

1. Introduction

Cancer is a life-threatening disease that has caused an estimated 8.2 million deaths worldwide [1]. While the current therapeutics, including conventional chemotherapy, radiotherapy and advanced targeted therapies (e.g. monoclonal antibodies), inhibit cancer progression, these treatments can have off-target effects and can destroy normal cells [2–4]. This instigates a critical need to develop more specific, effective therapeutics against cancer cells. Compared to normal cells, cancer cells require a greater abundance of micronutrients due to their continuous and rapid proliferation. One of the most crucial nutrients is iron, as it is important for several diverse, cellular metabolic pathways, including oxygen transport, DNA synthesis and ATP generation [5–7]. To sequester greater amounts of iron, cancer cells perturb iron homeostasis by increasing the levels of transferrin receptor 1 (TfR1; iron importer) and attenuating expression of ferroportin (FPN; iron exporter) [8]. Therefore, targeting the altered iron homeostasis in neoplastic cells could provide a mechanism to eliminate cancer cell generation and progression. Recent studies have targeted TfR1 via antibody (Ab)-mediated degradation, which has shown positive response for cancer therapy [4]; however, TfR1 is expressed on nearly all cell types, which instigates caution for using such type of therapy.

In addition to Ab-mediated targeted therapy, it has been suggested that bacterial siderophores could be used for anti-cancer therapy. Siderophores are ferric iron selective chelators that have a greater affinity for iron than host proteins like transferrin. The utilization of siderophores to deprive iron from cancer cells could induce cell death, analogous to normal red blood cells (RBCs) with iron deficiency anemia [9]. Moreover, low iron levels could trigger other pathways involved in cell cycle arrest and apoptosis [10]. Therefore, natural and synthetic siderophores like Deferoxamine (DFO; alias desferrioxamine, Desferal; isolated from Streptomyces pilosus) and Deferasirox, respectively, have been investigated for potential anti-tumor properties. DFO [11, 12] and Deferasirox (as reviewed in [13]) have shown significant anti-cancer effects in several studies by limiting iron bioavailability to malignant cells; yet, their substantial side effects and low efficacy have hindered their translation to pre-clinical trials. Deferasirox, for example, is considered as one of the drugs on the list of ‘most frequent suspected drugs in reported patient deaths’ assessed by the Institute for Safe Medical Practices, 2009. Moreover, the slow activity of the hydrophilic DFO [14] coupled with its very short plasma half-life [15] may prove inefficient in depriving iron from cancer cells. Therefore, while siderophores have been studied for a number of years to be used in anti-cancer iron chelation therapy, there has been limited success thus far. Hence, there is an unmet need to find a more effective iron chelator that exhibits specificity towards neoplastic cells.

We investigated whether the catecholate siderophore, enterobactin (Ent; aka Enterochelin, expressed by gram-negative bacteria) could be more effective in killing cancer cells in virtue of its membrane permeability, along with its strong and faster iron chelation property compared to other iron chelators and host iron-binding proteins [16, 17]. Our results demonstrated that iron-free Ent (alias apo-Ent), but not iron-bound Ent (alias holo-Ent), promoted apoptosis in two monocyte-related tumor cell lines (RAW264.7 and J774A.1) and the mouse Insulinoma 6 (MIN6) cell line. Noteworthy, apo-Ent failed to show its cytotoxic effects on primary macrophages derived from bone marrow. Moreover, our results suggest that Ent can chelate intracellular iron and disrupt iron-dependent effector mechanisms in the cancer cells, but not in primary cells. Such disparity between cancer and primary cells could be, in part, due to higher levels of lipocalin 2 (Lcn2; alias Siderocalin, immune protein that sequesters Ent) in primary cells than in cancer cell lines. Additionally, we demonstrated that Ent can disrupt mitochondrial function by dampening reactive oxidative species (ROS) generation, which may lead, in part, to the elevation in apoptosis activity. Remarkably, DFO failed to exert cytotoxic effects on macrophage cell lines as well as primary macrophages. The disparity between the cytotoxic effects of DFO and Ent may be due to the fact that DFO is not cell-permeable due to its high hydrophilicity [18], whereas the hydrophobic nature of Ent facilitates its membrane permeability. This supports and suggests that Ent is a more effective iron chelator that can target cancer cells by limiting both extracellular and intracellular levels of iron. Taken together, our current study highlights the anti-cancer role of Ent and its superiority compared to DFO.

2. Materials and methods

2.1. Reagents

Iron-free enterobactin (Escherichia coli), deferoxamine mesylate (DFO), pyoverdine (Pseudomonas fluorescens), ferrichrome (Ustilago sphaerogena), ferric chloride, Histopaque®−1077 and 1119, RPMI, Triton X-100, PIPES, DMSO, mannitol, sucrose, EGTA, 2’,7’-Dichlorodihydrofluorescein diacetate (DCFH-DA), fatty acid-free bovine serum albumin (BSA) and Lipopolysaccharide (LPS) were purchased from Sigma (St Louis, MO). FITC Annexin-V apoptosis detection kit was purchased from BD Pharmingen (BD Biosciences, San Jose, CA) and Griess reagent kit from Molecular Probes (Life Technologies, Columbus, OH). SYBR® Green mix and qScript cDNA synthesis kit were procured from Quanta Biosciences (Beverly, MA). Duoset enzyme-linked immunosorbent assay (ELISA) kits for mouse lipocalin 2 (Lcn2), and interleukin (IL)-6 were obtained from R&D Systems (Minneapolis, MN). Chrome azurol S (CAS) was purchased from Acros Organics (Geel, Belgium).

2.2. Mice

C57BL/6J wild-type (WT) mice were obtained from the Jackson Laboratory (Bar Harbor, ME) and bred in the animal vivarium at The Pennsylvania State University and University of Toledo. All the animal experiments were approved by The Institutional Animal Care and Use Committee (IACUC) at The Pennsylvania State University and University of Toledo.

2.3. Cell culture

The mouse macrophage cell lines, RAW264.7 (Abelson murine leukemia virus-transformed, 5^th passage) and J774A.1 (4^th passage), were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS, 100 IU/ml penicillin, and 100 μg/ml streptomycin, and maintained at 37°C in a humidified incubator with 5% CO₂. Cells were seeded in 12-well plates (2.0×10⁶ cells/ml) or 24 well plates (1.0×10⁶ cells/ml) and treated with Ent (0–50 μM), LPS (100 ng/ml), FeCl₃ (25 μM) or Ent+FeCl₃ (1:1 ratio) for 0–24 h. Culture supernatants and lysates in radio immunoassay buffer II (RIPAII) were collected and stored at −80°C until analysis. Culture supernatants were analyzed for nitrite measurement and cytokine production.

The murine insulinoma (MIN6, 4^th passage) cells were maintained in DMEM containing 25 mmol/L glucose, supplemented with 15% heat-inactivated fetal bovine serum (FBS), 100 units/ml penicillin, 100 μg/ml streptomycin, 100 μg/ml L-glutamine, and 5 μL/L β-mercaptoethanol in a humidified incubator with 5% CO₂ at 37°C.

2.4. Bone marrow-derived macrophages (BMDMs) culture

Bone marrow cells from BL6 mice were used to generate BMDMs as described by Zhang et al. [19]. Briefly, the BMDMs were cultured in the presence of macrophage colony-stimulating factor (M-CSF; 100 U/ml, R&D Systems, Minneapolis, MN) at 37°C with 5% CO₂ [20]. On every other day, 60% of the media was replaced with fresh complete media containing M-CSF. After 7 days, non-adherent cells were removed, and adherent cells were re-plated on 6 or 24 well plates at a density of 1.0 ×10⁶ cells/ml.

2.5. Cell proliferation assay

RAW264.7, J774A.1 and BMDMs (5×10⁴ cells/well) were seeded in 96-well plates containing DMEM media (performed in octuplicates). After 24 h, Ent (25 μM) or DFO (25 μM) were added to respective wells in incomplete DMEM media in octuplicates and cells were monitored for 90 h using the IncuCyte® system (Essen Bioscience, Ann Arbor, MI) to capture phase contrast images every 2 h and cell viability was analyzed using the integrated confluence algorithm. After 90 h, % of cell viability and time kinetics growth curves were generated using the IncuCyte S3 plate map software. The values were the average of octuplicate wells for each treatment.

2.6. Lactate dehydrogenase assay

Lactate dehydrogenase (LDH) levels in the culture supernatant of RAW264.7, J774A.1 and BMDMs cells were measured using a kit from Randox (Crumlin, UK) according to the manufacturer’s instructions.

2.7. Determination of nitrite

The production of nitrite (product of inducible nitric oxide synthase, iNOS) in the culture supernatant of RAW264.7, J774A.1 and BMDMs cells were measured by using the Greiss reagent kit (Molecular probes, Life Technologies, Columbus, OH) according to the manufacturer’s instructions.

2.8. Cell viability and apoptosis assay

RAW264.7, J774A.1, MIN6 insulinoma cell and BMDMs (2.0×10⁶ cells/well) prepared in 1 ml incomplete DMEM media were plated in 12-well plates. Ent (0–50 μM) or Ent+FeCl₃ (1:1 ratio) were added to their respective wells and incubated for 0–24h at 37°C and 5% CO₂. Cell viability and apoptosis were measured using the FITC Annexin-V apoptosis detection kit (BD Biosciences, San Jose, CA) according to manufacturer’s instruction. Results were acquired via flow cytometry (Accuri C6, BD Biosciences, San Jose, CA) and analyzed using the BD Accuri C6 Software (Becton Dickinson, Canaan, CT). Results were presented as the percentage of late apoptotic (Annexin-V and propidium iodide both positive) cells.

2.9. Immunoblotting

RAW264.7, J774A.1 and BMDMs cells were lysed in RIPA buffer (Cell Signaling, Danvers, MA) containing a protease and phosphatase inhibitor cocktail (Roche, Indianapolis, IN). Cell lysates (40 μg protein per lane) were fractionated via SDS-PAGE (4–20%, Bio-Rad), transferred to PVDF membrane (Bio-Rad, Des Plaines, IL), and probed with anti-mouse antibodies to GAPDH (1:5000, Cell signaling, Danvers, MA) or cleaved caspase 3 (1:1000, Cell signaling, Danvers, MA). Li-COR secondary antibodies were used according to manufacturer’s protocol. The fluorescence of the membranes was scanned by LI-COR Odyssey CLX imaging system.

2.10. ELISA

Interleukin-6 (IL-6) and Lipocalin-2 (Lcn2) levels in the culture supernatant of RAW264.7, J774A.1 and BMDMs cells or Lcn2 in cell lysate were analyzed using Duoset ELISA kits from R&D Systems (Minneapolis, MN) according to manufacturer’s instructions.

2.11. Chrome Azurol S (CAS) assay

CAS agar plates and liquid reagent were prepared as previously described by Schwyn and Neilands [21, 22]. The principle of the assay is that CAS remains blue when complexed with iron, but changes to orange when iron chelators sequester iron. Equal concentrations of enterobactin (Ent) and deferoxamine (DFO) (1mM, 0–24 h) were incubated on CAS agar plate overnight and monitored for formation of orange halo over time. To assess iron chelation property, Ent and DFO (25 μM) were added to the CAS liquid reagent (100 μl), incubated for 20 min at room temperature, and the change in absorbance was measured at 630 nm at different time points. Percent iron chelation was calculated using 0–100 μM of pyrocatechol as positive control as described previously [14].

2.12. Quantification of intracellular labile iron pool (LIP)

RAW264.7, J774A.1, and BMDMs cells were incubated with 0.5 μM Calcein-AM for 15 min. The cells were washed 2× with PBS and treated with either Ent or DFO (25 μM) for 3–24 h. After washing, cells were analyzed by using Accuri c6 flow cytometer (BD Biosciences, San Jose, CA) and the mean fluorescence intensity (MFI) was determined using the Accuri c6 Software (BD Biosciences, San Jose, CA). The levels of intracellular labile iron (LIP, ΔF) were calculated by subtracting the difference in the MFI, before and after treatment, with the iron chelators as previously described [23, 24].

2.13. Isolation of liver mitochondria

Liver mitochondria were isolated by using differential centrifugation method as previously described [25]. In brief, liver was excised from C57BL/6 mice and homogenized using mitochondrial isolation buffer containing [mannitol (225 mM), sucrose (75 mM), EGTA (1 mM), fatty acid-free BSA (0.1%), and Tris-HCl (10 mM, pH 7.4)]. The homogenate was centrifuged at 1,000 g for 10 min, then the supernatant was transferred to a fresh tube and centrifuged at 12,000 g for 10 min. Mitochondrial pellet was washed twice with mitochondrial isolation buffer, centrifuged at 12,000 g, and re-suspended in isolation buffer without EGTA. All centrifugation steps were performed at 4°C. Protein concentration was quantified using the bicinchoninic acid method from Pierce with bovine serum albumin as standard.

2.14. Determination of reactive oxygen species (ROS) generation

ROS generation was examined in hepatic mitochondria isolated from the livers of BL6 mice by using the redox-sensitive fluorescent probe 2’,7’-dichlorofluorescein-diacetate (DCFH-DA) as described previously [26, 27]. Briefly, isolated mitochondria (0.2 mg protein) were incubated in the assay media consisting of KCl (137 mM), MgCl₂ (2.5 mM), K₂HPO₄ (2 mM), Tris-HCl (10 mM, pH 7.4), glutamate (5 mM), malate (5 mM) and DCFH-DA (5 μM) at 37°C for 10 min to allow DCFH-DA to cross the mitochondrial membrane. The solution was then centrifuged at 12,000 g for 10 min, and the supernatant was discarded. The pellets were resuspended in fresh assay media without DCFH-DA. The mitochondrial pellets were incubated with either Ent, DFO, ferrichrome, pyoverdine, FeCl₃ (0–10 μM) or Ent+FeCl₃ (1:1 ratio) and immediately oxidant generation was measured using Molecular device fluorescence microplate reader for 30 min (excitation 488 nm and emission 525 nm). The amount of ROS generation was expressed as DCF formed per minute per milligram of protein.

2.15. Statistical analysis

All in vitro experiments were performed in triplicates and the results expressed as mean ± SEM. Data presented were representative of three independent experiments. Unpaired, two-tailed t-test was used to assess statistical significance between two groups. One-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test (when comparing each group with the control) or Tukey’s multiple comparison tests (when comparing each group with every other group) were used for comparing multiple groups. All statistical analyses were performed with the GraphPad Prism 7.0 software (GraphPad Inc, La Jolla, CA). p< 0.05 was considered statistically significant and denoted as * p< 0.05 and ** p<0.01 and *** p<0.001.

3. Results:

3.1. Ent is cytotoxic to monocyte tumor cells

The natural iron chelator, DFO, is the first bacterial siderophore to be clinically-approved as a therapeutic agent for treating iron overload disorders [15]. Despite its affinity and specificity for iron, DFO is not cell-permeable due to its hydrophilicity [18], which limits its efficacy. Compared to DFO, the catecholate-type, Ent, displays not only an unmatched affinity for iron, but also superior membrane permeability due to its hydrophobic nature [18]. The disparity between the iron-chelating property of Ent and DFO were evident when assayed on a CAS plate over a period of 24 h. This assay is based on the principle that CAS remains blue when complexed with iron, but turns orange when the iron is chelated by siderophores. Ent (1 mM) formed a clear halo within 1 h of incubation, whereas DFO at the same concentration produced a halo that was hardly noticeable even after 24 h (Fig. 1A). To confirm this observation, we performed a more sensitive time-dependent CAS liquid assay [14] with 25 μM of Ent and DFO. The iron-chelating activity of Ent was apparent within 20 min and achieved 90% chelation by 3 h of incubation. In contrast, DFO reached 60% and 70% iron chelation after 12 and 24 h, respectively (Fig. 1B).

Fig. 1. Ent promotes cytotoxicity to proliferating cells but not in primary cells.

(A) Formation of orange halo by Ent or DFO (1 mM) on CAS agar plate over different time periods. (B) Line graph indicates the relative iron chelation activity of Ent and DFO (25 μM) detected via CAS liquid assay. RAW264.7, J774A.1 and BMDMs cells (2.0×10⁶ cells/ml) were pretreated with Ent (25 μM) and LPS (100 ng/ml) for 24 h. For RAW264.7, (C) Western blot analysis showing expression of iNOS, (D) Bar graph represents the nitrite production measured in the supernatants after 24 h using Griess reagent, (E) Lcn2 and (F) IL-6 in the culture supernatant were determined by ELISA. For J774A.1, (G) Western blot showed the expression of iNOS, (H) Bar graph denotes the nitrite production in the supernatants, (I) Lcn2 and (J) IL-6. For BMDMs, (K) Immunoblot of iNOS, (L) nitrite production from macrophage cells was measured in the culture after 24 h, (M) Lcn2 and (N) IL-6. (O-Q) Line graphs showing % cell confluence of RAW264.7, J774A.1 and BMDMs for 90 h with and without Ent or DFO (25 μM) monitored by IncuCyte® system. The values are the average of octuplicate wells for each treatment. (R) Live cell images (phase contrast, 10× magnification) of RAW264.7, J774A.1 and BMDMs after 12 h of Ent or DFO treatment were captured using IncuCyte S3 plate map software. In vitro assays were performed in triplicates and data represented as mean ± SEM. * p< 0.05, ** p<0.01 and *** p<0.001.

Previously, we have shown that Ent, but not DFO, could impede the antimicrobial functions of neutrophils [24] and macrophages [23]. To further demonstrate the immunomodulatory effects of Ent, we pretreated macrophage cell lines (i.e. RAW264.7 and J774A.1) and primary macrophages (i.e. BMDMs) with Ent following LPS stimulation. As anticipated, iron-free Ent inhibited iNOS expression, nitrite production, and secretion of pro-inflammatory proteins (Lcn2 and IL-6) in RAW264.7 and J774A.1 macrophage cell lines (Fig.1C–J). Intriguingly, these inhibitory effects of Ent were not observed in BMDMs. On the contrary, Ent instead augmented LPS-induced iNOS and nitrite production but did not exert effects in Lcn2 and IL-6 secretion in BMDMs (Fig.1K–N).

The discrepancy between macrophage cell lines and BMDMs in their response to Ent could be due to that cell lines are highly proliferative in nature, whereas BMDMs are terminally-differentiated primary cells. As such, RAW264.7 cells may have a higher iron requirement to fuel their growth, which made them more susceptible to be impacted by Ent. Hence, we next employed the IncuCyte® live-cell imaging system to investigate whether Ent could differentially alter cell proliferation of transformed macrophages and BMDMs over 90 h following treatment. Interestingly, RAW264.7 and J774A.1 cells treated with 25 μM of Ent were mostly round in appearance and the few adherent cells were sparsely distributed on the culture plate, when compared to their vehicle-treated counterparts that reached confluency at the 12 h time-point. The loss in cell number in Ent-treated RAW264.7 and J774A.1, relative to their respective vehicle-treated controls (Fig. 1O, P, R), suggests that Ent may be cytotoxic to these cell lines. This was not the case for the BMDMs, whose cell number did not differ irrespectively of receiving vehicle or Ent treatment (Fig. 1Q, R, Supplementary video 1–6). We noted that 25 μM of DFO also exerted cytotoxicity to J774A.1 cells (Fig. 1P, R), albeit the effect was delayed and less pronounced when compared to Ent. However, no cytotoxicity was observed in RAW264.7 and BMDMs treated with DFO (Fig. 1O, Q, R).

3.2. Aferric-Ent, but not ferric-Ent, induces apoptosis in uncontrolled proliferative cells

To substantiate that Ent was promoting cell death, we next assayed the cell-free culture supernatant for lactate dehydrogenase (LDH), whose release indicates cytosolic leakage and can also be used as a cell death index [28]. Intriguingly, Ent promoted LDH release from RAW264.7 cells in a dose and time-dependent fashion (Fig. 2A, B), suggesting that Ent could exert a cytotoxic effect on this cell line. To affirm these results, the Ent-treated RAW264.7 cells were stained with annexin-V and propidium iodide (PI), whose co-positivity would indicate that these cells have undergone late apoptosis. Indeed, the percentage of apoptotic cells treated with Ent increased by 3-fold relative to control by 24 h post-treatment (Fig. 2C–E). Of note, Ent treatment also upregulated the expression of cleaved caspase 3 in RAW264.7 cells (Fig. 2F). The heightened LDH release, annexin-V and PI positivity, and activation of caspase 3 apoptotic pathway following Ent treatment collectively confirm that Ent is cytotoxic to RAW264.7 cells. These effects of Ent were likely to depend on its iron-chelating property as the aforementioned markers of cell death were diminished in cells that were treated with Ent bound with an equimolar concentration of Fe⁺³ (Fig. 2D–G). Similar results were observed with J774A.1 cells (Fig. 2H–J).

Fig. 2. Iron-free Ent but not ferric-Ent, induces apoptosis in uncontrolled multiplying cells.

RAW264.7 (2.0 ×10⁶ cells/ml) were treated with Ent (0–50 μM) or Ent+FeCl₃ (25 μM, 1:1 ratio) for 24 h. Ent induced the release of lactate dehydrogenase (LDH) in the culture supernatant in a (A) dose and (B) time dependent manner. (C) Ent induced cellular apoptosis in a dose dependent manner as measured by flow cytometry using the Annexin-V/PI positivity. Bar graphs indicate the % late apoptosis (% Annexin-V+ PI) at 24 h treatment. (D) Representative dot plots show the percentage of early and late apoptosis in Ent (25 μM) or Ent+FeCl₃ (1:1 ratio) treated RAW264.7 cells. (E) % apoptosis by Ent in presence of with and without equimolar Fe³⁺ (F) Western blot analysis of cleaved caspase 3, and GAPDH (as loading control) at 24 h post treatment. (G) Release of LDH in culture supernatant by Ent treatment in presence of with and without equimolar Fe³⁺. J774A.1 and MIN6 insulinoma cells (2.0 ×10⁶ cells/ml) were treated with Ent (25 μM) or Ent+FeCl₃ (1:1 ratio) for 24 h. For J774A.1, (H) Graphs represent the release of lactate dehydrogenase (LDH) in the culture supernatant after 24 h. (I) Apoptosis was measured by flow cytometry using the Annexin-V/PI positivity. Bar graph indicate the % late apoptosis (% Annexin-V+ PI). (J) Western blot of cleaved caspase 3. For MIN6, (K) LDH released in the insulinoma cell culture supernatant. (L) % apoptosis was measured by flow cytometry using the Annexin-V/PI positivity. (M) Western blot of cleaved caspase 3 and GAPDH (as loading control). In vitro assays were performed in triplicates and data represented as mean ± SEM. * p< 0.05, ** p<0.01 and *** p<0.001.

To confirm that the effects of Ent is specific to cancer cells, we performed similar experiments using another cancer cell line, MIN6 insulinoma cells. As expected, Ent-treated MIN6 recapitulated pro-apoptotic features of Ent-treated RAW264.7 and J774A.1 cells (Fig. 2K–M). Conversely, we did not observe any apoptotic effects of Ent in BMDMs (Fig. 3A–D). None of the apoptotic effects were observed when similar experiments were performed with DFO (Fig. 3E–G).

Fig. 3. Ent fails to induce apoptosis in primary (BMDM) cells.

BMDMs (2.0 ×10⁶ cells/ml) were treated with Ent (25 μM) or Ent+FeCl₃ (1:1 ratio) for 24 h. (A) LDH released in the BMDMs culture supernatant after 24 h treatment. (B) Apoptosis was measured by flow cytometry using the Annexin-V/PI positivity. Graph represents the % late apoptosis (% Annexin-V+ PI) at 24 h treatment. (C) Dot plots display % early and late apoptosis in Ent (25 μM) or Ent+FeCl₃ (1:1 ratio) treated BMDMs. (D) Immunoblot showing cleaved caspase 3 and GAPDH. RAW264.7, J774A.1 and BMDMs (2.0×10⁶ cells/ml) were treated with DFO (25 μM) or DFO+FeCl₃ (1:1 ratio) for 24 h. Bar graphs represent LDH release in the culture supernatants (E) RAW264.7 (F) J774.1 and (G) BMDMs. In vitro assays were performed in triplicates and data represented as mean ± SEM. * p< 0.05 and *** p<0.001.

3.3. Proliferating cells display higher levels of labile iron but lower levels of Lipocalin 2

Labile iron pool (LIP; alias catalytic iron) denominates the redox-active form of iron that serves essential roles in facilitating cellular metabolism and the various redox-dependent biological processes. Unlike normal cells, many uncontrolled multiplying cell-types display an altered iron metabolism that favors increased mobilization of iron stores and intracellular LIP. Such notion led recent studies to investigate whether intracellular LIP may be a key determinant for the survival of these cells when subjected to iron-targeted therapy [29]. Considering that iron-free Ent was more potent than iron-bound Ent in inducing cancer cell apoptosis, we hypothesized that such effects of Ent may likely be mediated via disrupting the intracellular LIP. To measure intracellular LIP, we employed an assay using the cytochemical calcein-AM, a cell-permeable dye whose fluorescence is quenched by weak binding [30, 31]. In principle, the addition of Ent would chelate the calcein-bound iron, thus releasing the fluorescent calcein; in this way, the intracellular LIP was quantified at 3 and 24 h based on the change in mean fluorescent intensity (ΔMFI). At 3 h, the level of cytosolic LIP accessible to Ent were 4-fold and 2-fold higher in RAW264.7 and J774A.1 cells, respectively, than BMDMs (Fig. 4A–D). The pool of Ent-chelatable LIP remained higher both in RAW264.7 and J774A.1 cells (2-fold and 3-fold, respectively) than BMDMs at the 24 h time-point (Fig. 4E–H). When the cells were treated with DFO instead of Ent, the level of cytosolic LIP in BMDMs was still detected as being 2-fold and 1.5-fold lower than in RAW264.7 and J774A.1 cells, respectively, at 3 and 24 h (Fig. 4I, J). In line with the slower iron-binding kinetics of DFO compared to Ent [14], we noted that DFO chelated less intracellular LIP than Ent in all the cell-types analyzed herein.

Fig. 4. Proliferating cancer cells possesses more chelatable labile iron pool (LIP) and substantially reduced Lcn2 expression compared to primary cells.

RAW264.7, J774A.1, and BMDMs cells were incubated with 0.5 μM calcein-AM for 15 min. and then treated with Ent (25 μM) for 3 h. After washing, cytosolic LIP was quantitated by flow cytometry. (A) Bar graphs represented the cytosolic LIP (ΔF) in RAW264.7, J774A.1, and BMDMs cells after 3 h of Ent treatment. (B, C and D) histograms represent the mean fluorescence intensity (MFI) of control and Ent treated cells. RAW264.7, J774A.1, and BMDMs were incubated with 0.5 μM calcein-AM for 15 min. The cells were treated with Ent (25 μM) for 24 h and with DFO (25 μM) for 3h and 24 h. After washing, cytosolic LIP was quantitated by flow cytometry. (E) Bar graphs denote cytosolic LIP (ΔF) in RAW264.7, J774A.1, and BMDMs cells after 24 h of Ent treatment. (F- H) Histograms represent the mean fluorescence intensity (MFI) of control and Ent treated RAW264.7, J774A.1, and BMDMs. Bar graphs denote cytosolic LIP (ΔF) in RAW264.7, J774A.1, and BMDMs cells after (I) 3 h and (J) 24 h of DFO treatment. RAW264.7, J774A.1, and BMDMs cell lysates were prepared in RIPA buffer and normalized with protein concentration and (K) intracellular and (L) extracellular Lcn2 were determined by ELISA. Specifically, extracellular Lcn2 was measured in culture supernatant after 24 h. In vitro assays were performed in triplicates and are representative of three independent experiments. Results are expressed as mean ± SEM. * p< 0.05, *** p<0.001.

Lcn2 is a multifaceted innate immune protein, whose well-established function is to sequester bacterial siderophores like Ent and prevent bacteria from acquiring iron from their host. Hence, we asked whether there is any difference in the expression of Lcn2 between these proliferating cells and primary cells. To confirm this hypothesis, we measured the levels of Lcn2 in both cell lysate and culture supernatant from RAW264.7, J774A.1 and BMDMs. Indeed, the BMDMs maintained significantly higher intracellular levels of Lcn2, i.e. ~74-fold and ~35-fold more compared to RAW264.7 and J774A.1 cells, respectively (Fig. 4K). Likewise, BMDMs also secreted more Lcn2 than RAW264.7 and J774A.1 cells (Fig. 4L). Of note, the elevated levels of intracellular and extracellular Lcn2 in primary cells mirrored their reduced Ent-chelatable LIP; such correlation was notably reversed in the cancer cells. The robust expression of Lcn2 in primary cells could neutralize Ent, thus safeguarding their LIP; such features may also, in part, underlie the reduced susceptibility of primary cells to the cytotoxicity effects of Ent. Further studies are certainly warranted to ascertain the extent to which the interplay between Lcn2 and Ent impact cellular functions.

3.4. Ent impedes mitochondrial ROS generation

Mitochondria are the most important sites for cellular respiration and ROS production, which are generated as byproducts of partial reduction of molecular oxygen in the electron transport chain; yet, these organelles are highly vulnerable to oxidative damage and dysfunction due to elevated intracellular LIP [32]. To examine the effects of Ent on mitochondrial function, we measured the ROS generation in isolated mitochondria from the liver either in the absence or presence of increasing Ent concentration (0–10 μM). We observed that Ent dose-dependently reduced mitochondrial ROS production, effectively at 1 and 10 μM concentration (Fig. 5A).

Fig. 5. Effect of Ent on mitochondrial ROS generation.

ROS generation was measured in mitochondria isolated from the livers of BL6 mice using the redox-sensitive fluorescent probe 2’,7’-dichlorofluorescein-diacetate (DCFH-DA). Isolated mitochondria were incubated with DCFH-DA (5 μM) at 37°C for 10 min. The stained mitochondria (0.2 mg protein) were incubated with either Ent (10 μM) or FeCl₃ (10 μM) or Ent+FeCl₃ (1:1 ratio) and ROS generation was measured using fluorescence microplate reader for 30 min (excitation 488 nm and emission 525 nm). (A) Bar graphs showing the dose dependent effects of Ent on mitochondrial ROS generation. (B) The line graphs were represented as the DCF fluorescence over time. (C) Bar graphs represent the DCF formed/min/mg protein at the end point. (D) The DCFH stained mitochondria (0.2 mg protein) were incubated with either Ent (1–10 μM), FeCl₃ (1–10 μM) or Ent+FeCl₃ (1:1 ratio) and ROS generation was measured. (E) The mitochondria were incubated with either Ent, DFO, ferrichrome or pyoverdine (0–10 μM) and ROS generation was measured as DCF formed/min/mg protein. In vitro assays were performed in triplicates and are representative of three independent experiments. Results are expressed as mean ± SEM. * p< 0.05.

Next, to determine whether Ent need to be in its iron-free form to dampen mitochondrial ROS generation, we treated mitochondria with either iron-free Ent or Ent in combination with equimolar concentration of Fe³⁺ (1 μM and 10 μM). Mitochondria ROS generation was significantly attenuated in presence of iron-free Ent over time (0–30 min), whereas such inhibition was partially abrogated when Ent was bound to Fe³⁺. The end-point for ROS generation in mitochondria treated with Fe³⁺ alone was comparable with their basal levels (Fig. 5B–D). Interestingly, other water soluble siderophores such as ferrichrome (from U. sphaerogena), pyoverdine (from P. fluorescens) and DFO (from S. pilosus) failed to inhibit mitochondrial ROS production at similar concentrations (Fig. 5E).

4. Discussion

Iron is an important metal ion that is required by almost all cells and organisms for their growth, metabolism and survival. Though iron deficiency causes cellular impairment, excess iron is toxic due to its participation in redox reactions that result in the formation of free radicals [33]. Accordingly, iron homeostasis is tightly regulated throughout its absorption, transport and storage at both the cellular and whole body levels. Of note, disruption of iron homeostasis has been associated with various diseases, including cancer [34–36], whereby iron overload could augment cancer growth, proliferation and metastasis. This point is well-substantiated in clinical studies which document that cancer patients display elevated levels of iron in tissues and in systemic circulation [36, 37]. The notion that iron-deprivation could be a useful anti-cancer strategy has gained much traction in recent years, thus raising the prospect that iron-chelating agents could be employed as a part of treatment regimen for different types of cancer.

Siderophores are small chemical, low molecular weight iron chelators that are ubiquitously expressed by many bacteria to facilitate iron acquisition. DFO is one such siderophore that has been clinically used for treating iron overdose and acute iron poisoning. Though the anti-tumor properties of DFO have been documented more than a decade ago [38–40], its application in cancer therapy has not entered clinical trials until much recently. However, the efficiency of DFO is considerably limited due to its hydrophilic nature which hinders it from permeating the cell membranes of cancer cells [41]. DFO also suffers from having a short plasma half-life, along with requiring multiple doses and long-term treatment to be effective. Nevertheless, increasing the concentration of DFO seems to rectify some of its shortcomings against metastatic and non-metastatic breast cancer cells [36]. Recent studies also sought to conjugate DFO with cell-penetrating and mitochondria-penetrating peptides to enhance its permeability into the cytosol [42] and mitochondria [18], respectively. To further surmount the limitations of DFO, many studies have begun to supplant this chelator by studying other types of siderophores or iron-chelators for their anti-tumor activities [41, 43–47]. The present study exemplify this premise by demonstrating that Ent could be exploited as an alternative to DFO as a therapeutic against cancer cells.

Among the naturally-occurring iron chelators, Ent is perhaps the quintessential catecholate-type siderophore known to exhibit the strongest affinity towards ferric iron (pFe = 34.3) [48]. In this study, we compared the anti-cancer properties of Ent with DFO, particularly in regards to their differences in effective dose, internalization into cytosol and mitochondria, and capacity to disrupt LIP in cancer cells. We focused on monocyte-derived cancer cell lines (i.e. RAW264.7 and J774A.1) as a model for this study given their dependency on iron for growth and survival, though we envision that the study outcomes could be broadly applicable to other cancer cell-types as well. Herein, we showed that a low dose of Ent (~10–25 μM) was sufficient to eliminate cancer cells as assessed via live-cell imaging, LDH leakage, induction of cleaved caspase-3 and AnnexinV+PI positivity. Conversely, DFO failed to demonstrate any cytotoxic effects at these concentrations, thus indicating that Ent may be a superior alternative to DFO as an anti-cancer siderophore. Moreover, the hydrophobicity of Ent also allows it to permeate cell membranes and gain access to the intracellular LIP more easily than DFO. Indeed, using the standard calcein-AM-based LIP assay, we showed that Ent could chelate ~3-fold more cytosolic LIP than DFO.

One pivotal study by Qi and Han elegantly demonstrates that iron-bound Ent could transport its iron into mitochondria and increase mitochondrial LIP [49]. It is tempting to speculate that a different outcome may be observed with iron-free Ent, which can disrupt mitochondrial iron homeostasis by chelating its LIP instead. This study, in part, addresses such a possibility by showing that iron-free Ent could inhibit mitochondrial ROS production in a dose-dependent manner, whereas such effects were mitigated when Ent was iron-bound. It is important to note that other water-soluble iron chelators, such as ferrichrome, pyoverdine and DFO, failed to inhibit mitochondrial ROS production at similar concentrations and time points. Considering the importance of mitochondrial function for cancer cells [50], we envision that its disruption by Ent could be one of the underlying mechanisms by which Ent could be cytotoxic to cancer cells. Such mechanism of action involving mitochondria dysregulation has also been reported in other studies which observed induction of cancer cell death following treatment with therapeutics that target LIP and/or iron homeostasis [4, 51–53].

Notwithstanding the advantages of Ent over DFO as a potential anti-cancer agent, Ent is not without its own set of limitations. For instance, Ent is insoluble in water, which may impede its mode of administration. Another drawback of Ent is its susceptibility to be neutralized by the anti-siderophore protein, Lcn2 [54], that may be expressed by certain cancer-types [55]. Yet, in the context in which cancer cells express lower levels of Lcn2 than normal cells, such disparity in Lcn2 expression could be a key determinant which predisposes the former, but not the latter, to the cytotoxic effects of Ent. This may be the case for BMDMs whose expression of Lcn2 may underlie their reduced levels of chelatable LIP as well their resistance towards Ent-induced cell death. This point may explain why we did not observe induction of apoptosis in our previous study in which BMDMs was employed instead of using cell lines [23]. However, our study does not rule out the possibility that Lcn2 and Ent could also synergize with one another to promote cancer cell death. Though this notion seems contrary to intuition, Devireddy et al. discovered that the fate of cancer cells depends on whether Lcn2 transports either iron-free or iron-bound Ent [56]. In the case of iron-free Ent, the resulting Lcn2-Ent complex could dissociate once internalized, thus allowing Ent to deplete intracellular iron and trigger apoptosis [56, 57]. Further studies are certainly warranted to determine whether the antagonistic relationship between Lcn2 and Ent could be harnessed into a therapeutic partnership in anti-cancer therapy.

Taken together, our study provides evidence that iron-free Ent is cytotoxic to highly proliferative cells, potentially by disrupting LIP homeostasis and/or dampening mitochondrial function. Though the ability of siderophores to induce cell death may not be surprising, our study highlights the distinction in the pro-apoptotic efficacy between hydrophobic (Ent) and hydrophilic (ferrichrome, pyoverdine and DFO) siderophores, in which the former was observed to be notably superior. Further studies are certainly critical to evaluate the extent to which the selective anti-cancer activity of Ent could be employed against cancer cells in vivo.

Abbreviation

TfR1

Transferrin receptor 1

FPN

Ferroportin

Ab

Antibody

RBCs

red blood cells

DFO

Deferoxamine

Ent

Enterobactin

Lcn2

lipocalin 2

ROS

Reactive oxidative species

DCFH-DA

2’,7’-Dichlorodihydrofluorescein diacetate

LPS

Lipopolysaccharide

DMEM

Dulbecco’s modified Eagle’s medium

RIPAII

Radio immunoassay buffer II

BMDMs

Bone marrow-derived macrophages

LDH

Lactate dehydrogenase

LIP

Labile iron pool

ANOVA

One-way analysis of variance

ΔMFI

mean fluorescent intensity