Quantifiable and Inexpensive Cell-Free Fluorescent Method to Confirm the Ability of Novel Compounds to Chelate Iron

Abstract

Cancer cells require large amounts of iron to maintain their proliferation. Iron metabolism is considered a hallmark of cancer, making iron a valid target for anti-cancer approaches. The development of novel compounds and the identification of leads for further modification requires that proof of mechanism assays be carried out. There are many assays to evaluate the impact on proliferation; however, the ability to chelate iron is an important and sometimes overlooked end-point measure due to the high costs of equipment and the challenge to quickly and reproducibly quantify the strength of chelation. Here, we describe a quantifiable and inexpensive cell-free fluorescent method to confirm the ability of novel compounds to chelate iron. Our assay relies on the commercially available inexpensive fluorescent dye Calcein, whose fluorescence can be quantified on most fluorescent microtiter plate readers. Calcein is a weak iron chelator, and its fluorescence is quenched when it binds Fe^2+/3+; fluorescence is restored when a novel chelator outcompetes Calcein for bound Fe^2+/3+. The removal of fluorescent quenching and the resulting increase in fluorescence allows the chelation ability of a novel putative chelator to be determined. Therefore, we offer an inexpensive, high-throughput assay that allows the rapid screening of novel candidate chelator compounds.

Introduction

Phenotypic changes to cells that relate to the development of cancer through a common set of altered biological capabilities are now commonly referred to as the hallmarks of cancer. Amongst them are changes resulting from the reprogramming of energy metabolism, which are widespread in cancer cell biology¹. Such metabolic reprogramming includes an increased requirement for iron to support rapid proliferation and tumor growth². This thirst for iron leads to dysregulated iron metabolism, which in and of itself is considered a hallmark of cancer^3,4, with dysregulation occurring at all stages⁵. The hallmarks of metastasis, more recently proposed by Welch and Hurst, include a role for iron⁶ since iron can induce oxidative stress, and this can, in turn, mediate changes to the genome, epigenome, and proteome, enhancing the possibility of metastasis⁷. A link between iron levels and an increased occurrence of cancer has been demonstrated through epidemiological studies⁸.

Since cancer cells require large amounts of iron, they are susceptible to iron deficiency and, therefore, iron chelation. We have recently published a review article highlighting the potential for iron chelation in reversing several hallmarks of cancer through NDRG1 disrupting oncogenic signaling pathways⁹. However, the use of iron chelation as stand-alone cancer therapy has not yielded positive results in clinical trials due to their toxicity, short half-life, rapid metabolism, and emerging resistance mechanisms. Nevertheless, iron chelators have shown promise in in vitro and in vivo investigations, indicating that more work is needed to develop effective iron chelators for cancer therapy. Specific iron chelation is a validated strategy in anticancer drug discovery, but only a few classes have been reported to date¹⁰.

The identification and characterization of novel iron chelators requires the ability to measure their effect on several endpoints. Many of these (such as proliferation, apoptosis, reactive oxygen species formation) are routinely measured and have been outlined and reviewed in the literature as methods to evaluate the hallmarks of cancer¹¹. When evaluating a novel iron chelator, many groups routinely examine the effect on anti-proliferative and redox activities as well as effects on iron influx or efflux. In silico prediction techniques¹² further increase the growing pool of iron chelators that can be screened.

Screening of iron chelators requires the ability to measure their effect on iron levels as a means to effectively demonstrate proof of principle. Currently, the most common method to do so is flow cytometry¹³ which is costly, time consuming and poorly quantifiable. The choice of assay is often based on the availability of experimental equipment, speed, and cost of an assay. Therefore, the ability to chelate iron can be an overlooked end point measure due to the high costs of equipment and the challenge to quickly and reproducibly quantify the strength of chelation. Here, we describe a quantifiable and inexpensive cell free fluorescent method to confirm the ability of novel compounds to chelate iron.

Protocol

1. Stock solution preparation

1. When preparing Calcein stocks solutions, prevent photo degradation by keeping solutions in the dark. Prepare a 1 mM solution of Calcein in Dulbecco's PBS, without magnesium or calcium¹⁴. For 5 mL of 1 mM Calcein solution, add 3.1 mg of Calcein (622.5 g/M) to 5 mL of Dulbecco's PBS. Using the 1 mM sock of Calcein, prepare 1 mL aliquots of secondary stock solutions in microfuge tube as detailed in Table 1.
2. Prepare ferric ammonium sulphate hexahydrate stocks by dissolving 19 mg of FAS hexahydrate (392.1 g/M) in 5 mL of Dulbecco's PBS to obtain a 10 mM ferric ammonium sulphate (FAS) solution. Prepare secondary stocks of FAS by dilution. To prepare a stock of 2 mM FAS (fA), dilute 200 µL of the 10 mM FAS stock with PBS with 800 µL of Dulbecco's PBS. To prepare a stock of 100 µM FAS (fB) dilute 50 µL of stock fA with 950 µL of Dulbecco's PBS.
3. To prepare 10 mM of Deferiprone stock, dissolve 2.8 mg (139.15 g/M) in 2 mL of Dulbecco's PBS. Prepare secondary stocks of 2mM iron chelator by diluting 200 µL of the 10 mM stock with 800 µL of Dulbecco's PBS.
   NOTE: In this assay the iron chelator of choice should be dissolved in Dulbecco's PBS without magnesium or calcium. As a working exemplar, we used the iron chelator Deferiprone.

2. Assay validation by determination of the linear range of Calcein fluorescence

1. Prepare a range of Calcein stocks 1 mM to 1 µM as described in stock solution table (see Table 1). Keep stock solutions in the dark to prevent photochemical degradation.
2. Pipette 90 µL of phosphate buffered saline (PBS) to wells A to H and 1 to 12 of a 96 well plate using a single channel manual pipette or a multichannel manual pipette.
3. Add 10 µL of 1 mM Calcein stock into column 12, rows A to H, of the 96 well plate using a single channel manual pipette.
4. Add 10 µL of the pre-prepared 800 µM Calcein stock (Table 1) to a 96 well plate in column 11, rows A to H, using a single channel manual pipette.
5. Repeat the addition procedure with the stocks from Table 1 so that column 10 contains 10 µL of the 400 µM stock and, column 9 contains 10 µL of the 200 µM stock etc. Continue this stepwise procedure across the columns until column 2 contains 10 µL of the 1 µM stock.
   NOTE: The final concentration of Calcein in each well is 10-fold lower than the stock solutions due to the 1:10 dilution.
6. Add 10 µL of PBS to Column 1, row A to E.
7. Incubate at room temperature for 30 min and set the analyzer on a microplate reader to an excitation of 490 nm and detection of 530 nm (or suitable filter set for Calcien fluorescence).

3. Assay validation by Fe²⁺ ion quenching to reduce Calcein fluorescence

1. Prepare a Calcein (cB) stock of 10 µM and a FAS stock (fA) of 2 mM (see Table of Materials). Keep the Calcein stock in the dark to prevent photodegradation.
2. Using a single channel manual pipette or a multichannel manual pipette, add 50 µL of PBS to columns 2 to 10, rows A to E. Do not add PBS to column 11.
3. Add 100 µL of a FAS (fA) stock of 2 mM into a 96 well plate in column 11 (Rows A to E) using a single channel manual pipette.
4. Using a multichannel manual pipette remove 50 µL of fA from column 11 and pipette it into the PBS already in column 10, mix thoroughly by pipetting.
5. Continue this stepwise double dilution procedure across the columns 9-2, rows A to E of the 96 well plate, and mix by pipetting.
6. Remove 50 µL of solution by pipetting from column 2 and discard the solution to waste.
7. Add 40 µL of PBS into all wells with a solution in them, columns 2-12, rows A to E.
8. Add 90 µL of PBS to wells column 1, rows A-E. These wells act as a positive control of 1 µM Calcein and 0 µM FAS.
9. Add 10 µL of 10 µM Calcein stock (cB) into each well with a solution in it (columns 1 to 12, rows A to E). This results in a top concentration of 1000 µM FAS in column 11, with FAS concentration halving for columns 10, 9, 8 across to column 2. Each of the resultant wells from the columns contains 1 µM Calcein and a dilution of FAS.
10. Add 100 µL of PBS to row F of columns 1 to 5 for the PBS alone control. Add 100 µL of 2 mM of FAS to row F of columns 6 to 10 for FAS alone control. Incubate at room temperature for 10 min in the dark.
11. Analyze on a microplate reader set to excitation of 490 nm and detection of 530 nm (or suitable filter set for Calcein fluorescence).

4. Assay for quantifying the ability of iron chelators to outcompete the weak chelator Calcein for Fe²⁺ ions

1. Prepare a Calcein stock (cB) of 10 µM, FAS stock (fB) of 100 µM and the iron chelator stocks of 2 mM as outlined in Table 1. Keep Calcein stock in the dark to prevent photodegradation.
2. Using a single channel manual pipette or a multichannel manual pipette, add 50 µL of PBS to columns 2 to 10, rows A to E, of a 96 well plate. Do not add PBS to column 11.
3. Add 100 µL of the 2 mM stock (fA) of the iron chelators, into a 96 well plate in column 11, rows A to E using a single channel manual pipette.
4. Using a multichannel manual pipette remove 50 µL of fA from column 11 and pipette it into the PBS in column 10, rows A to E, mixing thoroughly by pipetting.
5. Continue this stepwise double dilution procedure across the remaining columns 9-2, rows A to E of the plate, mix each well by pipetting. Remove 50 µL from column 2 and discard. This ensures a double dilution of 2 mM of the iron chelator (Deferiprone) across the 96 well plate columns 11 to 2, rows A to E.
6. Pipette 30 µL of PBS into all wells with a solution in them (e.g., columns 2-11 rows A to E).
7. Using a multi-channel pipette, add 10 µL of 100 µM FAS stock (fB) into columns 2-12, rows A to E. This results in a concentration of 1 µM Calcein and 10 µM FAS in each well, and a double dilution of the chelators, with the concentration halving each column across the plate, columns 11-2, with the highest concentration of iron chelator at 1000 µM in column 11 wells A to E.
8. Using a multi-channel pipette add 10 µL of 10 µM Calcein stock (cB) into columns 2-12, rows A to E.
9. Pipette 80 µL of PBS and 10 µL of Calcein stock (cB) and 10 µL of FAS stock (fB) to column 1, rows A-E using a pipette. Mix samples by pipetting. This provides the control of 1 µM Calcein with 10 µM FAS and no chelator.
10. Pipette 100 µL of PBS into row F columns 1 to 5. This provides the PBS negative control.
11. Add 100 µL of 2 mM of iron chelator (Deferiprone) to row F, columns 6 to 10. This provides the chelator alone control.
12. Incubate at room temperature for 10 min in the dark. Analyze on a multimodal microplate reader set to excitation at 490 nm and detection at 530 nm (or suitable filter set for Calcien fluorescence).

5. Data analysis

1. Analysis for assay validation by determination of the linear range of Calcein fluorescence.
   1. Calculate the mean Relative Fluorescence Units (RFU) of rows A-H for each concentration of Calcein in the columns of the 96 well plate from step 1 and plot against the concentration. Plot the line graph of Calcein concentration on the x-axis against mean RFU on the y-axis and use a linear regression to provide line of best fit.
   2. Quantify statistical significance by comparing Calcein concentrations in each column (2-11) to the control of PBS in column 1 rows A to E. Perform an Analysis of variance (ANOVA) and apply post hoc analysis using Least Significant Difference (LSD) with a threshold of p <0.001 to determine statistical significance, (denoted by ** in Figure 1).
2. Analysis for assay validation by Fe²⁺ ion quenching to reduce Calcein fluorescence.
   1. Calculate the average Relative Fluorescence Units (RFU) of rows A to E for each concentration of FAS in columns 2-11. Plot mean RFU for each FAS concentration on the y-axis. On the x-axis plot concentrations of FAS using a Log2 scale as itbetter distributes the data. Plot a line of best fit using a linear regression.
   2. Quantify statistical significance by comparing each columns mean to mean RFU of the control of PBS, 1 µM Calcein, 0 µM FAS (column 1 row A to E). Perform an ANOVA with LSD post hoc analysis with a threshold of p <0.001 to determine statistical significance.
3. Analysis for assay: Quantifying the ability of iron chelators to outcompete the weak chelator Calcein for Fe²⁺ ions.
   1. Calculate the average Relative Fluorescence Units (RFU) of rows A to E for each concentration of chelator in columns 2-11. Plot mean RFU for each chelator concentration on the y axis. On the x-axis plot concentrations of chelator using a Log2 scale as it better distributes the data.
   2. Quantify statistical significance by comparing each columns mean to mean RFU of the control of PBS 1 µM Calcein and 10 µM FAS (column 1 row A to E). Perform an ANOVA with LSD post hoc analysis with a threshold of p <0.001 to determine statistical significance.

Representative Results

For the method shown in step 2, this first experiment (Figure 1) established the linear range of the microtiter fluorescence plate reader when detecting Calcein fluorescent emissions. Our representative results show a wide linear range of Calcein fluorescence from 0-100 µM. ANOVA with LSD post hoc analysis demonstrates that there is a statistically significant differences in mean RFU for all Calcein concentrations when compared to a control of 0 µM of Calcein. In further experiments, 1 µM Calcein was used (steps 3 and 4) as it produces statistically significant differences from the 0 µM control and falls within the range that it is commonly applied to cells in the form of Calcein AM in flow cytometry¹³.

For the method shown in step 3, we demonstrate that when iron in the form of Fe²⁺ is added in a range of 0-1000 µM (as FAS) to 1 µM Calcein, the Fe²⁺ ions result in a linear decrease in Calcein fluorescence (RFU). This allows the observation of Fe²⁺ based quenching of Calcein fluorescence. Calcein as a weak iron chelator binds iron and this reduces the fluorescent output. In Figure 2, concentrations of 8.9 µM FAS and above showed a significant decrease, p<0.001, in Calcein fluorescence when compared to 1 µM Calcein alone. The use of 8.9 µM of FAS corresponded with a 33% reduction in Calcein fluorescence (mean RFU). For further experimentation (step 4) 10 µM FAS was used as it fell within the range of fluorescent quenching of Calcein.

For the method shown in step 4, the outcompeting of Calcein for Fe²⁺ ions by an iron chelator can be observed in Figure 3. Iron in the form of Fe²⁺ decreases Calcein fluorescence from the 1 µM control and the iron chelator deferiprone increases this fluorescence up to a peak, releasing the iron quenching of Calcein and increasing fluorescence (mean RFU). Deferiprone (used as an exemplar) at concentrations ranging from 0.97-512 µM resulted in statistically significant increases in fluorescence as observed with ANOVA and LSD post hoc analysis (p<0.001). The biggest fold change was a 3-fold change in mean RFU at 62.5 µM of chelator when compared to the control of 1 µM calcein AM and 10 µM FAS. Above 512 µM of deferiprone there was no statistically significant difference between the chelator and the control of 1 µM of Calcein and 10 µM of FAS.

Figure 1: Calcein concentrations and fluorescence (RFU) determined using a fluorescent microtiter plate reader. The Calcein concentrations of 0-100 µM incubated for 30 min and fluorescence (RFU) determined is shown here. The data shows the mean of 8 replicates per experiment and three independent experiments, error bars show a 95% confidence interval in the mean. Statistical significance is denoted by ** where p<0.001 as determined using ANOVA and LSD post hoc analysis (calculated using graph pad prism). Mean Calcein RFU of each concentration is compared to the control of PBS and 0 µM Calcein. Please click here to view a larger version of this figure.

Figure 2: Increasing concentration of FAS results in a linear decrease in Calcein fluorescence. The concentration of FAS was increased from 0-1000 µM which resulted in a linear decrease in 1 µM Calcein fluorescence after 30 min incubation in PBS The data shown are the mean of 5 replicates per experiment and three independent experiments, error bars show a 95% confidence interval in the mean. Statistical significance is denoted by ** where p<0.001 as determined using ANOVA and LSD post hoc analysis (calculated using graph pad prism). Mean Calcein RFU of each concentration of FAS and 1 µM of Calcein is compared to the control of 1 µM Calcein 0 µM of FAS in PBS. Please click here to view a larger version of this figure.

Figure 3: Increasing concentrations of deferiprone. The concentration of deferiprone was increased from 0-1000 µM with incubation of 1 µM Calcein, 10 µM FAS in PBS. Fluorescence was measured on a multimodal microplate reader. The data shown are the mean of 5 replicates per experiment and three independent experiments, error bars show a 95% confidence interval in the mean. Statistical significance is denoted by ** where p<0.001 as determined using ANOVA and LSD posthoc analysis (calculated using graph pad prism). Mean RFU where compared to 1 µM Calcein in PBS with 10 µM FAS and 0 µM of Deferiprone. Please click here to view a larger version of this figure.

Concentration (μm)	Volume from (1 mM stock)	Volume of PBS	Total volume
800	800 µL	200 µL	1 mL
400	400 µL	600 µL	1 mL
200	200 µL	800 µL	1 mL
100	100 µL	900 µL	1 mL
50	50 µL	950 µL	1 mL
25	25 µL	975 µL	1 mL
10	10 µL	990 µL	1 mL
6	6 µL	994 µL	1 mL
2	2 µL	998 µL	1 mL
1	1 µL	999 µL	1 mL

Table 1: Calcein secondary stock preparation.

Discussion

The over reliance of cancers on iron to fuel their metabolism makes iron chelation a potential addition to therapeutic regimes⁴. However, there is a limited ability to quickly screen novel metal ion chelators for their ability to bind iron ions. The commonly used and widely available fluorescent probe Calcein is known to act as a weak iron chelator and binding by iron ions quenches Calcein fluorescence. Fluorescence can then be recovered by competing for binding to iron using a stronger iron chelator. The reduction of fluorescence observed as Calcein binds iron and the restoration of fluorescence when it is outcompeted by a stronger chelator provides a quick and simple method for screening novel compounds with iron chelating potential. While Calcein based techniques¹³ are already used to visualize the ability of chelators to bind iron in cells, these techniques require costly flow cytometry apparatus, are only semi-quantitative, require extensive staff training and the use of the more expensive cell permeant form of Calcein, Calcien-AM. Moreover, the technique is poorly reproducible and does not readily facilitate the determination of data significance by statistical analysis. Therefore, we have created a simple inexpensive in vitro method which can act as a first pass screen of a novel compounds iron chelating abilities. This cell free assay abrogates the requirement for costly cell culture methods as a first pass high throughput screen for iron chelation ability.

There is increasing interest in the screening for, or development of novel iron chelators that may be used to target cancers due to their reliance on iron to drive their metabolic adaptations². In this paper we describe an in vitro technique which can be used to confirm the ability of novel compounds to chelate iron by outcompeting the weak chelator Calcein. This allows simple, quantitative, and inexpensive screening of novel compounds for their iron chelation ability.

One of the critical steps in the protocol is to establish the linear range of the fluorescent microtiter plate reader (Figure 1). Calcein has a wide fluorescent excitation and emission spectra that centers around an excitation peak of 501 nm and an emission peak at 521 nm. However, the linear range of the Calcein may be dependent on the filters used to excite and detect Calcein fluorescence emissions, the linear range of the photomultiplier used in detection will also in part determine the linear range. Therefore, it is important to first establish the range of concentrations at which Calcein fluorescence is linear to ensure the assay is working within the linear range of calcein on the fluorescent plate reader. Given that we observed linearity between 0 and 100 mM Calcein, this shows that Calcein has a broad linear range. Furthermore, fluorescent microtiter plate readers commonly have filter sets that can measure Calcein fluorescence as it has similar spectra to some of the most commonly used green light emitting fluorophores. Therefore, this assay is widely applicable to a large range of users.

The use of FAS as an iron ion donor is preferred as it provides Fe²⁺ ions which is the main iron ion in the labile iron pool in cells. However similar results to those shown here (Figure 2) have been achieved with Ferric Ammonium Chloride (FAC) which provides Fe³⁺ (data not shown). Therefore, the assay is compatible with both redox states of iron ions broadening the applicability to chelators with preference for either of these redox states.

In this study, we present data for the iron chelator Deferiprone, however, we have also evaluated the ability of other iron chelators including Mimosine to outcompete Calcein for iron ion binding using this assay (Data not shown). One of the limitations of the assay is that above 512 mM of the chelator deferiprone there was no statistically significant increase in fluorescence when compared to a control of 1 µM calcein and 10 µM of FAS. It is unclear why higher concentrations of chelator do not result in a fluorescent increase although this has been observed with other chelators (data not shown). However, a lead therapeutic compound should act <100 µM so the assay works at concentrations ranges required for lead compound efficacy.

This assay therefore allows any potential iron chelator to be quickly screened for the ability to chelate iron as a proof of mechanism. This is an important first step before the novel chelators is screened in expensive and laborious cell based or in vivo assays. A limitation of this assay is that it is performed using aqueous solvent and therefore requires the chelator to be soluble in aqueous buffers. However, given that the highest concentration used in this study was 10 mM Deferiprone which is a small lipophilic iron chelator¹⁵, then chelators with limited aqueous solubility can still be tested.

To conclude the assay described here may prove valuable in the screening of novel iron chelators which is an ongoing focus of cancer therapeutic development.

Materials

Name	Company	Catalog Number	Comments
Ammonium iron(II) sulfate hexahydrate	Sigma-Aldrich	215406	other wise known as FAS
Calcein	Sigma-Aldrich	C0875
Deferiprone	Sigma-Aldrich	379409
Dulbecco′s Phosphate Buffered Saline	Sigma-Aldrich	D5652	magnesium and calcium free
Greiner CELLSTAR 96 well plates	Sigma-Aldrich	M0812	any optically transparent 96 well plate will work