Investigation of the effects and mechanisms of anticancer action of a Ru(II)-arene iminophosphorane compound in triple negative breast cancer cells

Abstract

Triple negative breast cancer (TNBC) is one of the breast cancers with poorer prognosis and survival rates. TNBC has a disproportionally high incidence and mortality in women of African descent. We report on the evaluation of Ru-IM (1), a water-soluble organometallic ruthenium compound, in TNBC cell lines derived from patients of European (MDA-MB-231) and African (HCC-1806) ancestry (including IC₅₀ values, cellular and organelle uptake, cell death pathways, cell cycle, effects on migration, invasion, and angiogenesis, a preliminary proteomic analysis, and an NCI 60 cell-line panel screen). 1 was previously found highly efficacious in MDA-MB-231 cells and xenografts, with little systemic toxicity and preferential accumulation in the tumor. We observe a similar profile for this compound in the two cell lines studied, which includes high cytotoxicity, apoptotic behavior and potential antimetastatic and antiangiogenic properties. Cytokine M-CSF, involved in the PI3/AKT pathway, shows protein expression inhibition with exposure to 1. We also demonstrate a p53 independent mechanism of action.

Graphical Abstract

First preclinical attempt to assess health disparities in triple negative breast cancer treatments by using an organometallic compound (Ru-IM). Its evaluation in TNBC cell lines derived from patients of European (MDA-MB-231) and African (HCC-1806) descent affords a similar profile, displaying high cytotoxicity, apoptotic behavior and potential antimetastatic and antiangiogenic properties, plausibly involving the PI3/AKT pathway.

Introduction

Triple negative breast cancers (TNBCs) are some of the breast cancers with poorer prognosis, risk of recurrence after 5 years, and survival rates. Triple-negative breast cancer tumors test negative for three receptors (estrogen, progesterone, and excess human epidermal receptor HER-2). TNBCs are transcriptionally heterogenous (there are four main molecular sub-types) and this molecular heterogeneity is thought to be one of the key drivers of morbidity.^[1] Conventional hormone and HER-2 targeted therapies commonly used in other breast cancer types, have not been very successful in the treatment of TNBCs. The most common treatments currently employed are based on chemotherapeutic agents used in neoadjuvant settings (administered prior to other interventions such as surgery or radiation).^[2] Combination therapy (chemotherapeutics in conjunction with monoclonal antibodies, check-point inhibitors, growth factor receptor inhibitors or kinase inhibitors) are being explored in clinical trials, with 112 active trials currently being evaluated.^[3] The most common treatment for moderate-to-high risk TNBC is sequential, dose-dense anthracycline-taxane (ACT) combination.^[4] Immunotherapy treatments have also been explored and have delivered positive results for some patients, but it has been suggested that these agents should be used in combination therapy for more aggressive or metastatic cancers.^[5] A new type of targeted therapy has been recently developed (antibody drug conjugate sacituzumab govitecan, Trodelvy, which targets the protein Trop-2) which also includes a cytotoxic payload.^[6] The many clinical trials already completed or ongoing attest to the urgent need to find efficient and affordable treatments for TNBCs.

There is ample interest in the development of chemotherapeutics with low systemic toxicity and improved efficacy and pharmacological profile for TNBC to be used alone or in combination therapy. In this context, small molecules for the treatment of TNBCs have been reviewed and include some examples of organometallic compounds.^[7] We have recently reviewed the activity and potential of metal-based compounds as chemotherapeutics for TNBC including recent clinical trials (up to October of 2020).^[8] In this review we report not only on platinum compounds (that as cross-linking DNA agents are meant to be efficacious for TNBCs with BCRA1/2 mutations^[2,9]) but on a number of other metals (mostly ruthenium, gold and palladium^[8]). We focused on examples of compounds that have been efficacious in vivo and their mechanisms studied (at least in a preliminary way). Some other relevant examples have been recently reported.^[10,11]

Ruthenium compounds have emerged as potential alternatives to classical platinum-based chemotherapeutics for cancer.^[12] Recent successful clinical trials with two ruthenium compounds for treating non-muscle invasive bladder cancer (Ru(II)-based photosensitizer TLD1433: [Ru(bpy)(IP-TT)]²⁺, IP-TT = 2-(2′,2″:5′′,2‴- terthiophene)imidazo[4,5-f][1,10]phenanthroline)^[13,14] and for the treatment of solid tumors in combination therapy (Ru(III) antimetastatic agent BOLD-100: Na[trans-RuCl₄(Ind)₂], Ind = indazole)^[15,16] underscore the progress in this field. Ruthenium compounds have shown high efficacy in pre-clinical models in breast cancer^[17] including TNBCs mice models^.[8,18–22]

Our lab described the synthesis, characterization and biological evaluation on selected cancer cell lines of a series of cationic organometallic ruthenium (II) complexes with varying iminophosphorane (IM) ligands like water-soluble compound [(η⁶-p-cymene)Ru(k-N,O-Ph₃P=N-CO-2-N-C₅H₄)]Cl 1 (Ru-IM in Scheme 1).^[20,21] Ru-IM (1) was found to be effective against several cisplatin-resistant cell lines (including TNBC cell line MDA-MB-231) but less toxic to healthy human renal proximal tubular (RPTC) cell lines. Initial mechanistic studies in leukemia Jurkat and A549 lung cancer cells showed the induction of canonical or caspase-dependent apoptosis and an independence of p53.^[20,21] Ru-IM (1) interacts weakly with DNA (interaction of electrostatic nature). In vivo efficacy studies using NOD.CB17-Prkdc SCID/J mice bearing xenografted MDA-MB-231 tumors, showed an impressive tumor reduction (shrinkage) of 56% after 28 days of treatment (14 doses of 5 mg/kg of 1 every other day), low systemic toxicity and preferential accumulation in the tumor. This high efficacy in combination with its water-solubility and stability, makes Ru-IM (1) the most promising ruthenium-based preclinical candidate for TNBC described so far. Ru-IM (1) was quickly absorbed into blood plasma according to pharmacokinetic studies.^[20,21] The elimination half-life for 1 was 12.67 h, a value similar to those reported for well-known Ru derivatives (RAPTA-C, NAMI-A, and KP1019) in other type of cancer tumors.^[23,24] We demonstrated that the coordination of a bioactive ligand to a Ru(II) center was not strictly necessary in order to have high efficacy in vitro and in vivo or an improved pharmacological profile.

Scheme 1.

Structure of compound Ru-IM (1) evaluated in this study. Structure of compounds NAMI-A, RAPTA-C which are also evaluated for comparative purposes.

In terms of the higher incidence of TNBCs, it is known that it affects younger women, Hispanic women and disproportionately, women of African ancestry.^[25–27] While the higher mortality rate for TNBC in women of color (in the US and other countries) may be due to socioeconomic status, socioeconomic factors alone cannot fully explain the disparities observed.^[28–30] More recently, it has been found that disparities in TNBC may be attributed to racial differences in biological factors along with many other non-biological factors. In addition, these non-biological factors may potentially interact with risk factors of biological nature directly linked to TNBCs in women of African ancestry promoting a more aggressive type of disease and decreasing survival rates. ^[31] To the best of our knowledge, there is no systematic or purposeful study in evaluations of metal-based compounds in TNBCs cells or tumors derived from patients from different ethnicities. While a systematic study would involve a large number of cell lines with known genetic background and ancestry and it is worth undertaking, it is also relevant to start including cell lines derived from patients other than those of European descent in evaluation and mechanistic studies to acquire more data. The evaluation of compounds on different TNBC types and sub-types is very relevant as well. ^[32,33]

Here we report on the evaluation and behavior of compound in the commonly used TNBC cell line MDA-MB-231 (European descent) as well as on a second TNBC cell line (HCC-1806) of African ancestry. Studies include cell viability, cellular and organelle uptake, cell death pathways, cell cycle, effects on migration, invasion, and angiogenesis as well as a preliminary proteomic analysis via a proteome profiler human protease array kit, and validation of protein modification by western blot. We make comparisons with ruthenium compounds NAMI-A^[34–37] and RAPTA-C^[38–40] (Scheme 1). NAMI-A ^[34] (scheme 1) is a very well-known anionic Ru(III) antimetastasis inhibitor which was the first ruthenium anticancer compound to be studied on human beings, in two clinical trials.^[34–37] RAPTA-C, a Ru(II) compound containing p-cymene, exhibits anti-metastatic, anti-angiogenic, and anti-tumoral activities through protein and histone–deoxyribonucleic acid alterations.^[40]

We also report here on results on the evaluation of compound 1 on the NCI 60 cell-line panel which includes three other TNBC cell lines.

Results and Discussion

Cellular viability and cellular and organelle uptake. Results on the NCI 60 cell line panel five-dose screen

The cell lines described in this study are collected and described in Table 1. Some were directly studied by us (Table 2) or as part of the NCI 60-cell line panel (see supporting information). Table 1 includes details on the age and ethnicity of the patient (all female), intrinsic or TNBC sub-type,^[41] basal sub-type,^[41] histology, and mutated genes (including those relevant to TNBC).

Table 1.

Classification of the TNBC cell lines evaluated in this study and in the NCI 60 cell line panel five-dose screen.

Cell lines Derived from patient (ethnicity/ age) a	TNBC Subtype/Basal Subtype b	Histology a	Genes Altered in TNBC b	Total Mutant Genes c
MDA-MB-231 White 51 yo	MSL[d] / Basal B	Metastatic, pleural effusion adenocarcinoma	BRAF, CDKN2A, KRAS, NF2, TP53	364
BT549 White 72 yo	M[e] / Basal B	Primary ductal carcinoma	PTEN, RB1, TP53	210
Hs578T White 74 yo	MSL[d] / Basal B	Primary Mammary gland, breast carcinoma	CDKN2A, HRAS, PIK3R1, TP53	150
MDA-MB-468 Black 51 yo	BL1[f] / Basal A	Metastatic, pleural effusion adenocarcinoma	PTEN, RB1, SMAD4, TP53	288
HCC-1806 Black 60 yo	BL1[f] / Basal A	Primary, acantholytic squamous cell carcinoma	CDKN2A, KDM6A, STK11, TP53	6

^[a]Source: American Type Culture Collection (ATCC) datasheets.

^[b]Source: ref. 41.

^[c]Source: mutations taken from COSMIC database (www.sanger.ac.uk/genetics/CGP/cosmic/).

^[d]MSL = Mesenchymal stem like

^[e]ML = Mesenchymal like

^[f]BL1 = Basal like.

Table 2.

IC₅₀ values (μM) for TNBC cell lines MDA-MB-231, MDA-MB-468 and HCC1806, and non-malignant breast cells MCF10A and lung fibroblasts IMR-90 treated with compound 1 (Ru-IM). Ruthenium compounds. NAMI-A and RAPTA-C (Structures in Scheme 1) and cisplatin were used for comparative purposes. ^[a]

Cell lines [a]	Ru-IM (1)[b]	NAMI-A [c]	RAPTA-C	cisplatin
MDA-MB-231	3.7 ± 0.3	>300	>300	131.8 ± 18
MDA-MB-468	6.98 ± 2.74	>300	>300	44.19 ± 2.1
HCC-1806	9.29 ± 1.2	>300	>300	163.1 ± 5.86
MCF10a	2.66 ± 0.43	>300	>300	32.89 ± 6.68
IMR90	53.67 ± 3.3	>300	>300	148.9 ± 13.1

^[a]Compounds Ru-IM (1) and NAMI-A were dissolved in H₂0 while RAPTA-C was dissolved in DMSO, and cisplatin was dissolved in 0.9% NaCl solution All compounds were further diluted in cell culture media before addition to cells for a 24 h incubation period. The IC₅₀ values are reported with the standard deviation of the sample mean (triplicates).

We evaluated the cytotoxicity of Ru-IM (1) in three TNBC cell lines derived from patients of European (MDA-MB-231) and African (MDA-MB-468 and HCC-1806) ancestry. For comparative purposes, the cytotoxic profile of ruthenium compounds NAMI-A, RAPTA-C described above, and FDA-approved cisplatin was also determined. In this assay TNBC (MDA-MB-231, MDA-MB-468 and HCC1806) cell lines and non-malignant breast cells (MCF10A) and lung fibroblasts (IMR-90) were incubated with the above-described compounds for 24 hours. The compounds were assayed by monitoring their ability to inhibit cell growth using the PrestoBlue™ Cell Viability assay (see Experimental Section). The results are summarized in Table 2.

The cytotoxic profile of Ru-IM (1) in TNBC cell lines show values in the low micromolar range, with European-derived MDA-MB-231 (3.7 ± 0.3 μM) displaying slightly higher cytotoxicity than African-derived MDA-MB-468 (6.96 ± 2.74 μM), and HCC-1806 (9.29 ± 1.2 μM) lines. As reported, ^[20,21] these values are lower than those found for cisplatin (at 24 h) in Table 2. The cytotoxicity of Ru-IM (1) in MDA-MB-231 had been previously reported by us (MTT assay, 24 h) obtaining similar values (IC₅₀ = 2.61 ± 1.2 μM). ^[20,21] While an increase cytotoxicity is seen in normal breast tissue line MCF10a, the value from lung fibroblast IMR90 suggests selectivity. Furthermore, compound Ru-IM (1) displayed high toxicity in HEK-273 cell lines (2.8 ± 0.2 μM) but when evaluated in patient derived renal proximal tubular cell lines the value was higher (13.84 ± 1.46 μM). Similar increased cytotoxicity values have also been seen by other groups in regards to the MCF10a cell line.^[42] Importantly, Ru-IM was found to have very little systemic toxicity in vivo. ^[20,21] It is known that sometimes the toxicity in vitro does not correlate with what happens in mice xenograft models. The values found for NAMI-A and RAPTA-C were as expected as these compounds are known to have very high IC₅₀ values in MDA-MB-231 and they do not display cytotoxic properties. ^[43,44]

Ruthenium compounds are known for their variation in intracellular interactions, showing localization to nuclear,^[45] lysosomal,^[46] and mitochondrial^[47] compartments, amongst others. To understand the potential target of Ru-IM (1) ruthenium distribution between cytosolic and mitochondrial fractions after 2 and 24 h treatments with 2 μM of compound were analyzed with inductively coupled plasma optical emission spectrometry (ICP-OES) in TNBC MDA-MB-231 and non-malignant breast MCF10a lines (Figure 1).

Figure 1.

Intracellular distribution of ruthenium accumulation in cytosolic and mitochondrial fractions in MDA-MB-231 and MCF10a lines after 2 h and 24 h exposure to 2 μM of Ru-IM (1). The concentration of Ru in the cells was examined with inductively coupled plasma optical emission spectrometry (ICP-OES) analysis. Ru uptake (nmol) in mitochondrial and cytosolic fractions are expressed as a proportion of total ruthenium levels. Values represent the mean of three independent experiments. Error bars indicate standard deviation.

Total Ru accumulation (nmol) within the cell was ~75 % (Figure 1). Ruthenium distribution was calculated as a percent of total Ru dose (nmol) in cellular compartments as compared to initial Ru added and cellular fractionation was completed with the use of a mitochondria/cytosol fractionation kit (BioVision Inc., Milpitas, CA, USA). Ru accumulation remained stable at both time points in MDA-MB-231 (35% cytosol, 38% mitochondria) (Fig. 1 A) with a slight increase seen in MCF10a (35% to 38% cytosol and 40% to 39% in mitochondria in 2 h and 24 h respectively) (Fig. 1 B) suggesting that Ru-IM (1) shows cellular accumulation within 2 h of exposure in intracellular cytosolic fluid and the mitochondria. Given that initial analysis showed a caspase-dependent mechanism,^[20] it is not surprising to see mitochondrial accumulation.

Ru-IM (1) was tested against the National Cancer Institute 60-cell line panel at concentrations ranging from 10 nM (1×10–8 M) to 100 μM (1×10–4 M) with incubation times of 48 h and the cell viability evaluated by the sulforhodamine B assay. Subsequently, the GI₅₀ values in each cell line were obtained (See experimental and Tables in Supplementary Information for data on GI₅₀, TGI and LC₅₀). Figure 2 summarizes the range of GI₅₀ values of Ru-IM in each cancer category. It can be observed that Ru-IM (1) is active against a variety of cell lines and cancer types in the low micromolar range, roughly between 1 μM to 4 μM. The only exception is noticed in the case of renal cancer where activity is found in the range 2.5 μM to 13.2 μM. For most of the categories, all the values fall between the relevant points, the first quartile and the third quartile. In the case of colon cancer and melanoma, there is an outlier which fall above the maximum for each case. For the category of interest to us, the breast cancer category, the values are agreeable in the range of 2 μM and 2.70 μM. This category includes four triple negative breast cancer cell lines, MDA-MB-231, BT459, Hs578T, and MDA-MB-468 (Table 1).

Figure 2.

Box and whiskers plot representative of antiproliferative activity (GI₅₀ values) of ruthenium complex Ru-IM (1) in the NCI-60 cell line screen (NSC D-781091, Experiment ID 1408NS31). CNS: Central Nervous System. NSCL: Non-Small Cell Lung.

We selected for further studies the cell lines MDA-MB-231 (European ancestry) for which we had in vivo data and for which most preclinical studies with metal-based drugs have been conducted, and HCC-1806 (African ancestry) due to the higher difference in terms of IC₅₀ values obtained in our lab (Table 2). There are marked differences in these cell lines, which include differing TNBC subtypes (Basal B mesenchymal stem-like for MDA-MB-231 and Basal A basal-like for HCC-1806), histology (pleural effusion adenocarcinoma for MDA-MB-231, acantholytic squamous cell carcinoma for HCC-1806), and mutant genes (364 known mutants in MDA-MB-231, 6 mutants in HCC-1806) We thought it was worthy to further study the behavior of Ru-IM (1), NAMI-A, and RAPTA-C in these two TNBC cell lines, as their different molecular types, sub-types and backgrounds would allow us to view possible differences and similarities when treated with ruthenium derivatives.

Cell death type and cell cycle arrest

Following the evaluation of the cytotoxicity of the compounds, we proceeded to evaluate the mechanism of cell death of Ru-IM (1). Compounds NAMI-A and RAPTA-C were used for comparative purposes. Previous testing with the use of phosphatidylserine (PS) exposure as a measure of apoptosis showed an induction of 80% apoptosis in Jurkat (T-lymphocyte) cells with Ru-IM (1) treatment, but this was not tested on any TNBC cell lines.^[12] For this assay, TNBC MDA-MB-231 and HCC-1806 were incubated with the indicated compound at the IC₅₀ concentration for 24 h and then stained with FITC-Annexin V to measure cell death (Figure 3 depicts the percentages and Figure S1 in supporting information depicts the extent of viable cells undergoing apoptosis as determined by the percentage of the total cells that were positive for Annexin V with no propidium iodide stain, four quadrants).

Figure 3.

Cell death assay for Ru-IM (1) and other ruthenium compounds NAMI-A and RAPTA-C. Cells were incubated with compounds at IC₅₀ concentration over 24 h, stained with Annexin V/PI and analysed through flow cytometry. IC₅₀ values employed for NAMI-A and RAPTA-C were 900 μM and 750 μM respectively, for both cell lines. IC₅₀ values for Ru-IM (1) are provided in table 2 for both cell lines. The data reported in the graph and standard deviation of the sample mean result from at least two independent trials. Paired t-test analysis showed no significant differences (p < 0.05).

We observed that all compounds for Ru-IM (1), NAMI-A, and RAPTA-C, 46.05%, 35.35% and 30.8% of MDA-MB-231 cells die by apoptosis (Fig. 3 A). Similarly, in HCC-1806 cells treatment of Ru-IM (1), NAMI-A, and RAPTA-C results in 44.05%, 25.05%, and 27.25% apoptosis respectively (Fig. 3 B). It should be noted that these experiments are performed at IC₅₀ concentration values, with Ru-IM (1) having a concentration ca.240- or 200-fold lower than that of NAMI-A and RAPTA-C, respectively (Fig. 3). The induction of apoptosis is typical with treatment of RAPTA-C (activation by JNK pathway)^[48] and NAMI-A (activation through inhibition of MAPK/ERK pathway)^[49] and has also been reported for other ruthenium complexes.^[50] No major signs of necrosis were observed with any of the ruthenium compounds (<1%).

Next, we evaluated the effects of these three compounds on cell cycle arrest (Figure 4 and Figure S2 in supporting information). Following 24h treatment, cells were stained with propidium iodide to measure DNA content. In MDA-MB-231 cells this resulted in a slight increase of G2/M arrest across all treatments, with 24% of Ru-IM (1) and NAMI-A and 23% of RAPTA-C cells arrested, compared to 16% of non-treated cells (Fig. 4A). This trend was less noticeable in the HCC-1806 line, where values remained similar in treated and non-treated cells (Fig. 4B). There is also a slight increase of cells in the sub-G0 phase with Ru-IM (1) treatment across both cell lines, although the difference is negligible (<1%). Other ruthenium compounds have also shown this cell cycle variability across different breast cancer cell lines ^[51]. The induction of G2/M arrest is typical of RAPTA-C,^[50] NAMI-A,^[52] treatment and has also been observed in other ruthenium-based compounds.^[45,53] This arrest has also been associated with enhanced apoptosis which corroborates the data obtained in in the cell death assay.

Figure 4.

Cell cycle analysis and quantification for Ru-IM (1) and other ruthenium compounds NAMI-A and RAPTA-C in MDA-MB-231 (A) and HCC-1806 (B) cells. Cells were incubated over 24 h with an IC₂₀ concentration of respective compound and then stained with propidium iodide/RNase A solution. IC₂₀ values employed for Ru-IM (1), NAMI-A and RAPTA-C were 2.5 μM, 550 μM and 250 μM, for both cell lines. Cells were then analysed through flow cytometry. The data reported in the graph and standard deviation of the sample mean result from at least two independent trials. Paired t-test analysis showed no significant differences (p < 0.05). The tables collect the cell % numerically.

Inhibition of migration and invasion

Two hallmarks of metastasis are the increased local cell migration and later the distal invasion seen in advanced tumors.^[54,55] We evaluated the antimigration and antiinvasive properties of Ru-IM (1) and those of NAMI-A and RAPTA-C.

The effect of (IC₂₀) the compounds on migration in cell lines MDA-MB-231 and HCC-1806 was determined using a wound-healing 2D scratch assay on a collagen-coated plate (Fig. 5). Usually, IC₂₀ amounts are chosen in these assays as at those concentrations around 80% of cells are alive and the effect measured is not due to cell death.^[56] Migration is quantified by measuring the space in the wound gap that is occupied by cells 24 h after treatment.

Figure 5.

Cell migration assay for Ru-IM (1) and other ruthenium compounds NAMI-A and RAPTA-C. Inhibition of migration (2D wound‐healing scratch assay) was determined for each compound at its IC₂₀ value. IC₂₀ values employed for Ru-IM (1), NAMI-A and RAPTA-C were 2.5 μM, 550 μM and 250 μM, for both cell lines. A) Images show the width of the scratch at 0 and 24 h after it was inflicted in MDA-MB-231 and HCC-1806 lines. B) Quantification of cell migration at 24 h relative to the 0 h timepoint in MDA-MB-231 and HCC-1806 lines. The data reported in the graph and standard deviation of the sample mean result from at least two independent trials averaging blind quantitation from three fields of view per trial. Paired t-test analysis showed no significant differences (p < 0.05).

From this data, we infer that Ru-IM (1) has antimigratory properties in vitro that are even improved with respect to NAMI-A and RAPTA-C with well-known antimetastatic properties in vitro and in vivo.^[35.40]

Since these experiments are performed at IC₂₀ concentration values, with Ru-IM (1) having a concentration much lower than that of NAMI-A and RAPTA-C (220- or 100-fold lower, respectively), the antimigratory properties of this compound are very significant. In the HCC-1806 cell line, we observed a similar trend with the compounds reducing migration by 68.9% (Ru-IM, 1), 60.9% (NAMI-A) and 42.7% (RAPTA-C) (Fig. 5 B). The only noticeable change is a decrease in antimigratory properties in this cell line with respect to MDA-MB-231 for RAPTA-C (ca. 17.6%).

In addition, all the compounds studied interfere with invasion as evaluated in a 3D Transwell assay fitted with Geltrex® matrix an extracellular matrix analogue (Fig. 6 A). They follow a similar trend as that of migration with invasion reductions even more significant of 85.2% (Ru-IM, 1), 75.2% (NAMI-A), and 66.9% (RAPTA-C) in the MDA-MB-231 cell line (Fig. 5 B). In the HCC-1806 cell line (Fig. 6 B) there is a decrease of anti-invasive properties with invasion reductions found of 70.7% (Ru-IM, 1), 57.6% (NAMI-A) and 41.2% (RAPTA-C). This implies a reduction of the antiinvasive properties of 14.5% (Ru-IM, 1), 17.6% (NAMI-A) and 25.7% (RAPTA-C) in the HCC-1806 cell line with respect to the MDA-MB-231 cell line. Again, these experiments are performed at IC₂₀ concentration values, with Ru-IM (1) having a concentration much lower than that of NAMI-A and RAPTA-C (Fig. 3). Antimigratory and anti-invasive properties are typical of other ruthenium compounds in TNBC.^{[8,17,45,46,57]}

Figure 6.

Cell invasion assay for Ru-IM (1) and ruthenium compounds NAMI-A and RAPTA-C. Inhibition of invasion (3D Transwell assay) was determined for each compound at its IC₂₀ value. IC₂₀ values employed for Ru-IM (1), NAMI-A and RAPTA-C were 2.5 μM, 550 μM and 250 μM, for both cell lines. A) Panels show representative phase-contrast images images of cells 24h following fixation with paraformaldehyde and crystal violet stain. B) Quantification of cell invasion at 24 h as stained cells in proportion to non-stained cells in MDA-MB-231 and HCC-1806 lines. The data reported in the graph and standard deviation of the sample mean result from at least two independent trials averaging blind quantitation from two fields of view per trial. Paired t-tests (*p < 0.05) assuming Gaussian distribution shows

Inhibition of angiogenesis

Tumors need oxygen and nutrients to grow and therefore the development of blood vessels (neovascularization) is key in tumor malignancy.

Angiogenesis is the process by which tumors cells (or surrounding cells) generate these blood vessels (by production of growth factors) are formed by tumor. In order to examine potential disturbing effects on angiogenesis by compound 1, NAMI-A, and RAPTA-C, we chose to examine the effect of their IC₁₀ concentrations in an in vitro model to analyze the formation of tube-like structures by HUVEC cells.^[56] In this assay, the endothelial tube formation of Human umbilical vein endothelial cells (HUVECs) on an ECM-like matrix is determined as a function of length of the uninterrupted tubes (TL) and number of branching point or nodes in the tubes (TN) (Fig. 7). The number of tubes and nodes is counted using Image-J with the Angiogenesis plugin. Inhibition of tube formation is corelated to length of tube (TL), the lower length, the greater the inhibition.

Figure 7.

Tube formation assay for Ru-IM (1) and ruthenium compounds NAMI-A and RAPTA-C. Inhibition of angiogenesis was measured through endothelial cell reorganization into 3D vessel structures with human umbilical endothelial cells (HUVEC) at IC₁₀ concentrations of each compound in these cells (Ru-IM 1: 1.2 μM, NAMI-A and RAPTA-C: 150 μM). (A) representative phase-contrast images captured 24 h after the compounds were added. (B) Quantification of tube formation. Length of the uninterrupted tubes (TL) and number of branching points or nodes (TN) were used to evaluate tube formation. The data reported in the graph and standard deviation of the sample mean result from at least two independent trials averaging blind quantitation from three fields of view per trial. Paired t-test analysis showed no significant differences (p < 0.05).

The antiangiogenic properties of a compound are correlated to the number of nodes (TN) with the lower number associated to the higher anti-angiogenic properties. NAMI-A and RAPTA-C are known to display antiangiogenic properties,^[58,59] and this trend has been viewed in other ruthenium complexes as well.^[45] Ru-IM (1) and has a disruption of tube length of 34%, while RAPTA-C has 27% and NAMI-A has 43%. The disruption in tube length (TL) for Ru-IM is of 29% while NAMI-A and RAPTA-C both provoke disruptions of 22%. While all compounds display similar antiangiogenic properties, the concentrations employed are very different, with Ru-IM (1) needing a concentration 125-fold lower than that of NAMI-A and RAPTA-C.

Proteomic Analysis

As already reported, Ru-IM (1) interacts weakly with DNA (electrostatic interaction) and it does not inhibit Cathepsin-B as opposed to other antimetastatic Ru agents.^[20,21] In order to have a preliminary insight into potential targets for Ru-IM (1) in TNBC cells, we performed a proteomic analysis (IC₂₀ concentration) in MDA-MB-231 and HCC-1806 cells with the use of the Proteome Profiler Human XL Oncology Array (R&D Systems, Minneapolis, MN, USA). This assay allows for the detection of the differences in 84 cancer-related proteins between treated and untreated samples.^[60] Although not at significant levels, many of the same targets were expressed in both cell lines which include those related to the immune response (Cathepsin B, Cathepsin B, Cathepsin S, IL-8), anti-inflammatory targets (HMOX-1, IL-8, progranulin) and markers involved in the apoptotic pathway (BCL-xL, Survivin, Galectin-3, FOXO1) (Figure 8). In terms of significant protein expression level change following treatment of Ru-IM (1), the macrophage colony-stimulating factor (M-CSF) was expressed 1.5-fold and 2-fold less in MDA-MB-231 (Fig. 8 A) and HCC1806 (Fig. 8 B) cells respectively. This is a cytokine that regulates the growth, proliferation, and differentiation of macrophages, monocytes, monoblasts, and other hematopoietic stem cells. The M-CSF signalling pathway is activated by receptor c-Fms and results in a signalling cascade which begins with the phosphorylation of tyrosine residue Y559 in the cytoplasmic tail of c-Fms, and ends with the activation of the phosphatidylinositol 3-kinase (P13K)-protein kinase B (AKT) pathway.^[61] This pathway is frequently altered in TNBC, with mutations and alterations seen in roughly 25% of TNBCs.^[62] Down regulation of this pathway has resulted in the inhibition of cancerous growth for other ruthenium complexes^[46,63] although it has not been tested extensively in TNBC. To validate results, M-CSF protein expression levels following Ru-IM (1) treatment were evaluated using western blot analysis (Supplementary Figure 3). It was demonstrated that M-CSF protein levels decreased over 24h in both cell lines, confirming the results seen in the proteome profiler assay. The inhibition of M-CSF by Ru-IM (1) is promising and further investigation into the downstream targets of PI3K/AKT may implicate this pathway into the mechanism of action.

Figure 8.

Proteomic analysis of 84 cancer targets Proteome Profiler Human XL Oncology Array (R&D Systems, Minneapolis, MN, USA) in TNBC cell lines MDA-MB-231 and HCC-1806 untreated (blue) and treated (red) with Ru-IM (1). Treated cells were dosed with an IC₂₀ concentration of Ru-IM (1) (2.5 μM) over 24 h before cell lysates were collected. Red asterisk represents protein marker M-CSF and blue asterisk represents protein marker p-53. The data reported in the graph and standard deviation of the sample mean result from two independent trials.

While previous experiments revealed a p53 independent mechanism of action, this was completed in A549 lung cancer cells.^[20] The biomarker p53 elicited a very slight increase in MDA-MB-231 cells in the proteomic analysis. The expression of p53 was further tested in MDA-MB-231 and HCC-1806 TNBC lines by western blot analysis (Supplementary Figure 4). Given that HCC-1806 is negative for the expression of p53 due to a truncation mutation ^[64] we did not observe any activity in this line with or without Ru-IM (1) treatment as anticipated. In addition, MDA-MB-231 showed no changes through 6h and 24h treatment with 1, confirming the initial results showing a p53 independent mechanism of action.

Conclusion

In conclusion, we have explored the cellular activity of previously reported Ru-IM (1) in five different TNBC cell lines (including some evaluated in the NCI 60 cell line panel screen). The studies confirm that Ru-IM (1) has high cytotoxicity (below 10 μM) for all the cell lines in 24/48 h cell viability assays. As described before, the effects of this compound are quite fast (8 h). Like NAMI-A and RAPTA-C, Ru-IM (1) is apoptotic and displays slight increase of G2/M arrest. Ru-IM (1) has a good cellular uptake (75%) with a 50:50 accumulation in the cytosol and mitochondria in the TNBC cell line MDA-MB-231, and in a non-malignant cell line (MCF10a). This compound displays relevant antimigratory, anti-invasive and antiangiogenic properties in two TNBC cell lines that are better or comparable to those of NAMI-A and RAPTA-C (at much lower concentrations). The differences of behavior found for the two cell lines studied derived from patients of European (MDA-MB-231) and African (HCC1806) ancestry are not considerable, but in all cases the Ru compounds studied performed better in the European (MDA-MB-231) cell line, with Ru-IM (1) compound being the compound displaying less differences. Although NAMI-A and RAPTA-C show limited cytotoxic profiles, the use of the same dosing concentration across all experiments allows comparison of biological activity in vitro. Overall, further understanding of the biological basis of responsive to treatment that considers racial differences is needed for the development of any TNBC drug. From the proteomic analysis (and confirmed by western blot analysis) we have identified the inhibition of the macrophage colony-stimulating factor (M-CSF) which is involved upstream in the in the PI3/AKT pathway. We have also have confirmed that the cytotoxic effects of Ru-IM (1) are non-dependent on the p53 pathway.

The results presented here along with those obtained previously on its high efficacy in MDA-MB-231 xenografts (58% tumor reduction) with little systemic toxicity and preferential accumulation in the tumor, makes Ru-IM (1) one of the most promising metal-based drugs for TNBC reported to date. Moreover, Ru-IM (1) is a highly water-soluble compound, which is very relevant for potential oral availability. Ru-IM (1) is easy to prepare, and it is relatively cheap (when compared to cisplatin, carboplatin and other drugs currently used in chemotherapeutic TNBC treatments). The use of bioconjugation with targeting vectors (e.g. anti-Trop 2) or nanoparticle and liposomal encapsulation techniques, may be explored in the future to increase cancer cell selectivity and overall efficacy of Ru-IM (1). Future detailed mechanistic investigations on cells and tumors derived from patients from different backgrounds, are underway, and will be key to advance this compound beyond the preclinical stage.

Experimental Section

Metal-based compounds used

Cisplatin was purchased from Strem and used without further purification. Ruthenium compounds [(η⁶-p-cymene)Ru(k-N,O-Ph₃P=N-CO-2-N-C₅H₄)]Cl Ru-IM (1),^[20,21] imidazolium[RuCl₄(imidazole)(DSMO)] NAMI-A^[23] and [Ru(η⁶-p-cymene)Cl₂(1,3,5-​triaza-7-phosphaadamantane]) RAPTA-C^[38,39] were prepared as described previously.

Cell lines

Human breast adenocarcinoma line MDA-MB-231 and human lung fibroblast line IMR90 were obtained from the American Type Culture Collection (ATCC; Manassas, Virginia, USA) and cultured with Dulbecco’s modified Eagle’s medium (DMEM; Fisher Scientific) containing 10 % fetal bovine serum, certified, heat inactivated, US origin (FBS; Fisher Scientific), 1 % minimum essential media (MEM) nonessential amino acids (NEAA; Fisher Scientific), and 1 % penicillin–streptomycin (PenStrep; Fisher Scientific). Human breast primary acantholytic squamous cell carcinoma HCC-1806 line was cultured using Roswell Park Memorial Institute (RPMI‐1640; Fisher Scientific, Hampton, NH, USA) and supplemented with 10 % FBS, 1 % MEM‐NEAA, and 1 % PenStrep. Human breast adenocarcinoma line MDA-MB-468 was obtained from ATCC and cultured in Leibovitz’s L-15 medium (L-15; Fisher Scientific) containing supplements 10 % FBS, 1 % MEM‐NEAA, and 1 % PenStrep. Non-malignant breast epithelial line MCF10a cells was purchased from ATCC and cultured with mammary epithelial cell growth medium (MEGM; Lonza, Basel, Switzerland) supplemented with cholera toxin and insulin. Human umbilical vein endothelial cells (HUVEC) were obtained from ATCC and cultured in vascular cell basal medium (ATCC). All cells were cultured at 37 °C under 5 % CO2 and 95 % air in a humidified incubator.

Cell viability analysis

The cytotoxic profile (IC₅₀) of the compounds were determined by assessing the viability of MDA-MB-231, MDA-MB-468, HCC1806, IMR90, and MCF10a lines. Cells were seeded at a concentration of 5.6×103 – 6× 103 cells per well in 100 μL of appropriate complete media into tissue‐culture‐grade 96‐well flat‐bottom microplates (CELLTREAT, Pepperell, MA, USA) and grown for 24 h at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator. Ru-IM (1) and NAMI-A were dissolved in H₂O while RAPTA-C was dissolved in DMSO, and cisplatin was dissolved in 0.9% NaCl solution. Dilutions ranging from 0.5–100 μM for Ru-IM (1) and Cisplatin and 100–1000 μM for RAPTA-C and NAMI-A were added, directly followed by a 24 h incubation period. PrestoBlue was used to quantitively measure variations in the viability of treated cells. 11 μL per well of 10× PrestoBlue (Invitrogen, Carlsbad, CA, USA) labeling mixture was added to the cells at a final concentration of 1× and incubated for 1 h at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator. The optical fluorescence of each well in a 96‐well plate was quantified using a BioTek Synergy Multi‐mode microplate reader (BioTek Instruments, Inc., Winooski, VT, USA) set at 560/590 nm. The percentage of surviving cells was calculated from the ratio of absorbance of treated to untreated cells. The IC₅₀ value was calculated as the concentration required to decrease cell proliferation by 50 % and is presented as the mean ± SEM of at least two independent experiments, each with triplicate measurements.

Cellular and organelle uptake

To determine the ruthenium metal uptake in MDA-MB-231 and MCF10a lines, cells were seeded onto a 6 well plate (Corning) at a density of a density of 25×103. Treatment with IC₂₀ concentration of Ru-IM (1) over 2 h and 24 h was then completed. Following incubation, cellular fragmentation of mitochondrial and cytosolic components was completed with the use of a mitochondria/cytosol fractionation kit (BioVision Inc., Milpitas, CA, USA). Samples were then digested over 72 h in a 0.7 mL 70% nitric acid and 1.5 mL 35% hydrogen peroxide mixture at 60 °C. samples were analyzed with a PerkinElmer Optima 7300 DV spectrometer and was calibrated prior to use. Signals at a wavelength of 240.272 nm were monitored. Biodistribution values are presented as the percent of the accumulated dose and were calculated by including appropriate standards. All data presented are expressed as mean ± SD. All samples were completed in triplicate.

NCI60 cell five-dose screen

Ru-IM (1) complex was submitted to the National Cancer Institute for screening on its panel of 60 cancer cell lines (NSC D-781091, Experiment ID 1408NS31). The protocols used by the NCI have been described previously.^[65,66] Cells of each cell line were seeded in 96-well plates at densities ranging from 5,000±40,000 cells per well depending on the doubling time of individual cell lines. Briefly, cells were exposed to Ru-IM (1) at 0, 0.1, 1.0, 10 and, 100 μM and incubated for 48 h at 37 °C in a humidified atmosphere containing 5% CO₂. The cells were then fixed and stained with sulforhodamine B (SRB) solution to determine their viability, and subsequently calculate the GI₅₀, TGI and LC₅₀.

Cell death assay

For the assessment of cell death, cells were cultured in 100‐mm tissue culture dishes with appropriate media type and allowed to reach ≈75 % confluency. The cells were treated with Ru-IM (1), NAMI-A, and RAPTA-C at their respective IC₅₀ concentrations and incubated for 24 h at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator. Then, cells were collected through trypsinization (Fisher Scientific) to later count with a hemocytometer and prepare 1×10⁶ cells per sample. After the samples were prepared, the cells in the samples were centrifuged, supernatant was aspirated from the pellet, and washed gently one time with PBS. The cells were centrifuged again, PBS was aspirated, and eBioscience Annexin V‐FITC Apop Kit (Invitrogen, Carlsbad, CA) was used to label cells as follows: Cells were resuspended in 100 μL 1× binding buffer with 5 μL Annexin V dye and 5 μL propidium iodide. The cells were incubated at room temperature for 15 min. After the incubation period, cells were resuspended in 400 μL 1× binding buffer. The dye’s fluorescence intensity was detected via flow cytometry using a BD C6 Accuri flow cytometer; 10×10⁵ events per sample were recorded. The flow cytometer was calibrated prior to each use.

Cell cycle profile

To determine the cell cycle profile of MDA-MB-231, HCC-1806 and MDA-MB-468 lines, cells were cultured in appropriate media containing IC₂₀ concentrations of Ru-IM (1), NAMI-A, and RAPTA-C for 24 h at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator. Cells were then collected and counted to 25×10⁴ cells per sample. Cells were fixed in 70% ethanol and incubated overnight at 4 °C. Cells were then washed with PBS three times and stained with propidium iodide solution (100 μg/ml RNase A (Sigma-Aldrich), 50 μg/ml Propidium Iodide (Sigma-Aldrich) in PBS) overnight at 4 °C. The dye’s fluorescence intensity was detected via flow cytometry using a BD C6 Accuri flow cytometer; 10×10⁵ events per sample were recorded. The flow cytometer was calibrated prior to each use.

Cell migration analysis

Cells were seeded onto a fibronectin‐coated six‐well plate (Corning Incorporated, Durham, NC, USA) and grown to a monolayer of ≈90 % confluency. Afterward, the monolayer was scratched using a 200 μL tip. The complete medium and cells detached due to the scratch were aspirated and replaced with serum‐free medium. The antimigratory profiles of Ru-IM (1), NAMI-A, and RAPTA-C were assessed with the IC₂₀ of each compound. Cells were incubated at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator over 24 h. At 0 and 24 h after the scratch, cells were photographed using a Leica MC120 HD mounted on a Leica DMi1 microscope at 5× magnification. The area invaded was measured in three randomly selected segments from each photograph, then averaged. Data was collected from at least three independent experiments.

Cell invasion analysis

A transwell assay was used to view cell invasion properties. A 24-well plate with invasion chamber inserts (Corning) were coated with Geltrex®, Reduced Growth Factor Basement Membrane Matrix (Invitrogen) over two hours at 37 °C to solidify, after which 2.5 × 103 cells containing serum-free media with an IC₂₀ concentration of Ru-IM (1), NAMI-A, and RAPTA-C were seeded to the top of the solidified layer. Complete media was added as a chemoattractant to the bottom of the 24 well plate. After a 24 h incubation at 37 °C under 5 % CO₂ and 95 % air in a humidified incubator, inserts were fixed with 4% formaldehyde and stained with crystal violet (Sigma). Cells were photographed from the bottom of the insert using a Leica MC120 HD mounted on a Leica DMi1 microscope at 10× magnification. Data analysis was completed blindly with ImageJ (Fiji) software with cell counter plugin. Data was collected from at least two independent experiments.

Angiogenesis analysis

The endothelial tube formation of human umbilical vein endothelial cells (HUVECs) was assessed for potential anti-angiogenesis properties. Briefly, 96-well plates were coated with Geltrex®, Reduced Growth Factor Basement Membrane Matrix (Invitrogen) and incubated at 37 °C for 30 min to allow gelation to occur. HUVECs were added to the top of the gel at a density of 6 × 103 cells/well in the presence of Ru-IM (1), NAMI-A, and RAPTA-C at their IC₁₀ concentrations. The diluting agent (0.1% DMSO for RAPTA-C and 0.1% water for Ru-IM and NAMI-A) served as positive control. Cells were incubated at 37 °C with 5% CO₂ for 20 h and pictures were captured with a Leica MC120 HD mounted on a Leica DMi1 microscope at 5x magnification. Quantification of tube formation was assisted by ImageJ (Fiji) angiogenesis analyzer plug-in. Tube formation quantified by number of branching points (tube nodes, TN) and total length skeleton (tube length, TL). The data were obtained from the average of three wells per treatment condition from two independent experiments.

Proteomic analysis

To establish a proteomic profile of MDA-MB-231 and HCC-1806 lines, cells were dosed with an IC₂₀ concentration of Ru-IM (1) for 24 h at 37 °C under 5 % CO2 and 95 % air in a humidified incubator. Cellular lysates were collected with RIPA buffer and sonication, after which total protein content was quantified with the use of a Bradford assay. Untreated cells were used as a control. Detection of the differences in 84 cancer-related proteins between treated and untreated samples occurred with the use of the Proteome Profiler Human XL Oncology Array (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s protocol.^[47] Briefly, 200 μg protein concentration/sample was added to each membrane and incubated with a blocking array overnight at 4°C. Membranes were washed and incubated with the detection antibody cocktail. Following a second wash step, membranes were exposed to chemireagent and images were acquired with a Fujifilm LAS 3000 Imager. Dot intensity was quantified with ImageJ (Fiji) with values normalized to internal reference control.

Western blot analysis

For the assessment of protein expression through western blot analysis, cells were cultured in 100‐mm tissue culture dishes with appropriate media type and allowed to reach ≈75 % confluency. Cells were then dosed with IC₂₀ concentrations of Ru-IM (1). Following 6h and 24h incubations with serum-free media, cell lysates were collected with the use of RIPA buffer and protein concentration quantification was completed with a Bradford assay. Western blot was carried out by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and electro-transferred onto a polyvinylidene difluoride (PDVF) membrane (150 mA, 1h). Membranes were dried completely and then blocked with Intercept Protein-Free Blocking Buffer (LICOR Biosciences, Lincoln, NE, USA) for 1h on a shaking platform. Membranes were then incubated overnight on a shaking platform at 4°C 1:200 with primary antibody P53-DO1 (Santa Clara Biotechnology, Dallas, TX, USA) or 1:1000 with M-CSF (R&D Systems), and with control primary antibody total H4 or alpha tubulin. Membranes were then washed five times with TBST and incubated with anti-rabbit and anti-mouse antibodies (LICOR Biosciences) for 1h at room temperature. Following five TBST and one TBS wash, signals were visualized with 600 and 700 channels with an Odyssey FC imaging system (LICOR Biosciences).

Statistical analysis

Data were recorded as mean ± SD. Comparisons between different groups were undertaken using a two-tailed paired t-test assuming Gaussian distribution. The limit of statistical significance was P < 0.05 for all experiments. Statistical analysis was performed using the Graphpad/Prism software.