1. Introduction
Combination therapy has long been studied in the treatment of cancer, including ovarian cancer [
1]. It is a logical approach, focusing on increasing the response and tolerability to treatment, while also decreasing resistance [
2]. Unfortunately, it can be difficult to assess whether a particular combination will behave in a synergistic, additive, or antagonist fashion when used on a particular cancer patient. The only known way of determining their effectiveness is to identify specific measures, such as response rate, survival, or time to progression, and assess if or whether the new combination is able to achieve a significant improvement [
2]. Combination therapy is essentially cooperative: each agent involved should have non-overlapping toxicities, different mechanisms of action with minimal cross-resistance, and individually proven in treatment by itself [
2].
Several cancers, including ovarian cancer, have demonstrated resistance or reduced sensitivity to cisplatin treatment, leading to decreased time to disease progression, increased likelihood of relapse, and reduced efficacy upon re-treatment during relapse [
3,
4]; cisplatin itself causes significant health problems, such as nephrotoxicity [
5]. Ideally, cisplatin in combination with a drug that diminishes its negative effects while enhancing the therapeutic effects would decrease resistance or relapse while mitigating its negative effects. For example, cisplatin (or another platinum agent) in combination with taxanes (e.g., paclitaxel) is now regarded as standard chemotherapy in ovarian cancer, where the taxane enhances the tumour’s radiosensitivity. Cisplatin and the adenovirus OBP-301 have also been shown to work synergistically [
1].
Cisplatin shares cytotoxic synergy with bee venom [
1]. Bee venom makes sense as a complement to cisplatin: it has protective effects in many areas of the body such as the blood and nerves [
6]; it is able to inhibit cell growth in tumours [
7]; and has even been examined and used in complementary treatments that are necessitated by the effects of chemotherapy, such as allodynia [
8] and neuropathy [
9]. Cisplatin in combination with bee venom has been successfully used against human cervical and laryngeal carcinoma cells, including their drug-resistant sublines [
10], and human glioblastoma [
11]. Several studies have found that phospholipase A
2, another component of bee venom, not only mitigates the negative impact of cisplatin on kidneys [
12], but also boosts regulatory T cell (Treg) numbers in the spleen and enhances Treg traffic to the kidneys following cisplatin exposure. This is important, as Tregs play a significant role in many mechanisms, including inflammation and autoimmunity suppression [
4,
12,
13]. A significant bonus in this dynamic is that bee venom has no negative impact on the anti-cancer properties of the cisplatin [
4], meaning it in no way diminishes the effect of the cisplatin treatment. This is critical; a combinatory agent which diminishes any positive effects of the primary treatment undermines its effectiveness overall. The effects that cisplatin and bee venom have together on ovarian cancer cells emerge from the synergistic relationship between the two agents [
14]. Alizadehnohi et al. report that separately and together, the two agents induce apoptosis in human ovarian cancer cells; bee venom appears to enhance the cytotoxic impact of cisplatin [
14]. It also appears that melittin provokes responses in cisplatin-sensitive cells which lead to decreasing levels of amino acids, which in turn affect the energy metabolism of the tumour [
15].
In order to understand these effects and the metabolic changes that the cisplatin-melittin combination has on ovarian cancer cells, metabolomic investigations can be undertaken. Metabolomics in the context of oncology usefully focuses on diagnosis and prognosis, as well as on evaluating the effectiveness of therapy [
16,
17]. For instance, one study employed nuclear magnetic resonance (
1H-NMR) spectroscopy and was able to accurately separate serum metabolite profiles of three groups of patients: namely women with ovarian cancer, normal premenopausal women, and women with a benign ovarian disease [
18]. In addition, LC-MS in combination with Biolog Microarray assays revealed that treatment of cisplatin-resistant and sensitive ovarian cancer cells with melittin distinctively altered their lipid profiles and their ability to metabolise certain carbon energy sources, [
15]. NMR spectroscopy has also been used to produce and examine metabolic profiles in other cancers such as hepatocellular carcinoma [
19].
In essence, the effects of the cytotoxic mechanisms of both cisplatin and melittin, together and separately, cause changes in cells which can be identified and measured through metabolomic analysis. The current study aimed to examine the metabolic effects of melittin in combination with cisplatin on A2780 and A2780CR human ovarian cancer cell lines using a LC-MS based metabolomics approach employing a ZIC-pHILIC column. Multivariate data analysis was performed based on PCA and OPLS-DA models constructed using the SIMCA-P software.
3. Discussion
Cisplatin is one of the most effective anticancer drugs currently used for treating many types of cancer, however, it comes with serious side effects. Combination therapy has been used in cancer treatment in order to increase therapeutic response and tolerability, and to decrease resistance [
2].
The present study aimed to determine whether or not melittin, a cytotoxic peptide from bee venom, possesses a synergistic inhibitory effect in combination with cisplatin on A2780 and A2780CR cells. In addition, the study was intended to determine the metabolomics effects of the melittin + cisplatin combination treatment on the two cell lines which would corroborate any observed synergistic cytotoxic effects.
Cell viability assays using the Alamar
® Blue method confirmed the synergistic cytotoxic effects of melittin in combination with cisplatin at certain concentrations, although at other concentrations, antagonistic effects were observed. CI analysis of the cytotoxicity data showed that, on A2780 cells, the combinations had synergistic effects at 2 μg/mL of melittin plus either 5 or 6 μg/mL of cisplatin, respectively. In contrast, at melittin + cisplatin concentrations of 3 + 2 and 4 + 2 μg/mL, respectively, antagonistic effects were observed. Synergistic effects were observed in A2780CR cells when the melittin was combined with cisplatin at 2 + 10 and 5 + 10 μg/mL, while with a fixed concentration of melittin (2 μg/mL) and variable concentrations of cisplatin (20 and 30 μg/mL) in the combination, antagonist effects were observed. The CI analysis used the median effect equation of Chou and the combination index equation of Chou and Talalay to quantify synergism or antagonism at different concentrations, and to select the best pair of drugs to combine for potentially maximal antitumor efficacy [
22,
25]. This method of analysis has been useful in identifying effective combinations of anticancer drugs [
26,
27].
Recent studies have reported that components of bee venom may exert an anti-tumour effect on human ovarian cancer and that the venom has the potential for enhancing the cytotoxic effect of the antitumor agent cisplatin [
14]. Different melittin + cisplatin mechanisms could interact to either reduce or increase anticancer efficacy, thus producing three possible effects: (1) Additive, when the combined effect is equal to the sum of individual effects; (2) Antagonistic, when the effect of one or both compounds is less than when they are applied together than when individually applied; (3) synergism, when the effect of combined substances is greater than the sum of the individual effects [
24]. Our findings show that these effects can occur depending on the concentrations of melittin and cisplatin in the combination.
With respect to OCCs, several previous studies have analysed the metabolic responses of OCCs to various compounds [
28,
29,
30]. Additionally, there have been some previous metabolomic studies on the comparison between the effects of cisplatin on squamous cancer cell lines sensitive and resistant to cisplatin [
31], and the effects of docetaxel on ovarian cancer stem cells [
30]. Our previous study employed a metabolomics approach to assess the effects of melittin monotherapy on OCCs that revealed significant changes in amino acid and carbohydrate metabolism [
15]. In addition, clear differences were previously observed in the metabolomes of the untreated cells [
15]. Although our study demonstrated profound metabolic changes in the cells after melittin monotherapy, there has been no metabolomics study to date that has comparatively profiled the metabolite composition of OCCs treated with a combination of melittin and cisplatin. A previous study suggested that combination therapy is more effective than monotherapy on cancer cells such as hepatocellular carcinoma [
19]. In the current untargeted metabolomics study, metabolic profiles of A2780 and A2780CR cells treated with melittin + cisplatin combinations were assessed using a LC-MS based metabolomics approach, with OPLS-DA models displaying good separation between the experimental groups, high-quality goodness of fit (R
2), and high-quality goodness of prediction (Q
2).
The metabolomics analysis demonstrated distinct metabolic profiles for the treated A2780 and A2780CR cells, although the treatments were adjusted in accordance with cell sensitivities. Specifically, the concentrations of melittin and cisplatin used were chosen based on cytotoxicity assays and CI values for synergy to allow detection of a combination effect rather than to achieve a maximal anticancer effect. The most altered metabolites in A2780 and A2780CR cells could be categorised under amino acid, energy, carbohydrate, and nucleotide metabolism. Most of the altered metabolites participate in more than one pathway in significant ways, and the change in that one metabolite could have a resonating effect for other pathways.
There was a very clear effect of the combination treatment on the purine and pyrimidine pathways. This was very different from the metabolic shifts observed for melittin and cisplatin alone [
15,
24], suggesting the combination has quite a different effect on cell metabolism. There was a very large increase in the levels of the adenine metabolite, hypoxanthine, and guanine in both cell lines. This possibly indicates that the combination of melittin with cisplatin is promoting greater adduct formation between cisplatin and DNA in comparison with the treatment with cisplatin alone, where there was no strong evidence for effects on the levels of DNA bases [
24]. Adducts formed with cisplatin are mainly intra-strand crosslinks joining two guanine residues and to a lesser extent intra-strand links between guanine and adenine [
32]. DNA is repaired by excision of damaged bases and this would correlate with largely increased levels of guanine and hypoxanthine, although this presumes that the cisplatin adduct somehow breaks down during the excision. The levels of the purine metabolites adenosine monophosphate (AMP), adenosine diphosphate (ADP), and guanosine triphosphate (GTP) were decreased in both cell lines by the combination treatment. In a previous study, it was found that FK866, a small molecule inhibitor of nicotinamide phosphoribosyltransferase (NAMPT), caused significant metabolic changes in purine metabolism in ovarian cancer and colorectal cancer cells [
30]. Moreover, Zhou et al. described a study in hepatocellular carcinoma (HepG2 cells) that showed that high-dose treatment with sorafenib, an oral multikinase inhibitor, affects purine metabolism with significant decreases in GTP levels [
33]. Despite the profound dose-dependent metabolic changes in HepG2 cells induced by sorafenib monotherapy, Zhou et al. showed that everolimus, another anticancer agent, in combination with first-line sorafenib therapy results in more pronounced metabolic changes to pyruvate, amino acid, methane, glyoxylate, and dicarboxylate, and glycolysis or gluconeogenesis in hepatocellular carcinoma cells [
33]. Other than purine metabolite changes, consistent variations were observed for pyrimidine metabolism. The levels of pyrimidine metabolites such as orotate, dihydrothymine, dihydrouracil, and uridine triphosphate (UTP) were reduced in both cell lines after exposure to the melittin + cisplatin combinations. The reason for the decrease in pyrimidine metabolites is not clear. Normally DNA damage might be associated with increased levels of dihydrothymine which is produced by excision of damaged thymine residues from DNA.
There were many altered metabolites belonging to several pathways for amino acid metabolism. Most of the metabolites grouped under the arginine and proline pathways were reduced in sensitive cells after the combination treatments; while the arginine metabolite was non-significantly altered in resistant cells. Similarly, our previous study examined the effect of melittin on A2780 and A2780CR cells which showed that the level of arginine was downregulated in cisplatin sensitive cells compared with resistant cells [
15]. A number of studies have reported that arginine deficiency enhances apoptosis in different cell lines including human lymphoblastic cell lines [
34], mesothelioma cells [
35], and melanoma cell lines [
36]. Some human cancers, such as melanoma and hepatocellular carcinoma [
37], do not express arginosuccinase synthase and therefore are unable to synthesise arginine from citrulline [
38]. A recent study observed that ovarian carcinoma SKOV3 cells under arginine deprivation showed increased sensitivity to treatment with paclitaxel, a chemotherapy drug used to treat cancers, at low doses. In this context, it is to be noted that paclitaxel is a disruptor of the cytoskeleton and negatively impacts on the autophagosome-lysosome fusion step [
39]. A previous study suggested that combinational treatment based on arginine deprivation and an autophagy inhibitor (for example chloroquine, a known nontoxic antimalarial drug) can potentially be applied as a second line treatment for a subset of ovarian carcinomas deficient in argininosuccinate synthetase [
39]. It was also observed that the development of chemoresistance to platinum compounds in ovarian carcinomas leads to collateral appearance of arginine auxotrophy due to the downregulation of argininosuccinate synthetase [
40]. The exact mechanism whereby deficiency arginine biosynthesis confers resistance remains unclear.
There was a strong effect of the combination treatment on cellular cysteine and glutathione metabolism; S-glutathionyl-L-cysteine, 3-sulfino-L-alanine, glutathione, L-cysteinylglycine, and L-cystathionine were all lower in cisplatin sensitive cells compared with resistant cells after being treated with the combinations. In a previous study, it was found that the level of glutathione was higher in resistant cells (A2780-CP20) than in sensitive cells (A2780) [
41]. This finding resembles the current results in which the level of glutathione was higher in resistant cells than sensitive cells following combination treatment. Two of the main reasons for platinum resistance in OCCs are the p53 mutation and drug-induced increases in intracellular glutathione concentration. A study by Mohell et al. showed that methylene quinuclidinone (MQ), in addition to binding to cysteine residues in p53, also binds to glutathione, decreasing intracellular glutathione levels in OCCs [
42]. Therefore, it is possible that the combination promotes greater binding of cisplatin to glutathione thus depleting its levels which occurs to a greater extent in the sensitive cells. Mohell et al. also observed that combination effects of APR-246 (which is a prodrug that is converted to the active compound MQ) with doxorubicin were the cause of a DNA damage response, including activation of the p53 pathway leading to apoptosis [
42]. Moreover, recent metabolomics based studies in OCCs have demonstrated that gossypol decreases cellular levels of GSH and induces apoptosis through oxidative stress [
30]. In our previous study it was observed that the levels of GSH were no different between resistant and sensitive cells. However, in the current study GSH is depleted by the combination treatment to a much greater extent in the sensitive cells. The depletion appears to be related to the ability of the cells to synthesise GSH. Although there is no difference between the cysteine levels in the two cell lines, there are marked differences in key intermediates which can be used to synthesise both cysteine and GSH including glycine, serine, cystathionine, and glutathione cysteine. There are also lower levels of S-adenosylmethionine in the treated sensitive cells which is a source of homocysteine which is also a precursor of cysteine.
Increased serine biosynthesis is one of many metabolic changes that have been reported in cancer cells [
43,
44], and serine is a central node for the biosynthesis of many molecules such as glycine and cysteine [
45]. High levels of serine in cancer cells have been linked to increased rates of cell proliferation [
46]. The level of serine was increased in resistant cells following the combination treatment. In contrast, treatment of the sensitive cells with the combination resulted in a decrease of serine within the cells and a further lowering of the nonessential amino acid glycine. Glycine is incorporated directly into purine nucleotide bases and into GSH. The conversion of serine to glycine, catalysed by serine hydroxymethyltransferase (SHMT), donates a one-carbon unit to tetrahydrofolate to produce 5, 10-methylenetetrahydrofolate (CH
2-THF). CH
2-THF is used in thymidine synthesis and is a precursor of other folate species that contribute to purine synthesis [
46]. The difference in OCCs could be reflected at the cellular level in terms of differences in the metabolite profiles. Serine is required for a number of biosynthetic and signalling pathways, including the production of phospholipids such as sphingolipids and phosphatidylserine [
46]. Previous studies have shown that serine biosynthesis appears to be part of an adaptive response to oxidative stress [
47]. The tumour suppressor p53 is emerging as an important regulator of cellular metabolism. P53 is a key player in the cellular response to stress in the form of numerous challenges, including DNA damage, hypoxia, and oncogene activation [
48]. Cells lacking p53 fail to respond to serine starvation due to oxidative stress, which leads to reduced viability and severely impaired proliferation [
49].
The level of ATP was found to be reduced in both cell lines after the combination treatment. ATP was found to be more reduced in both cell lines when treated with the combinations compared with the results observed previously with melittin monotherapy [
15]. It is known that glycolysis provides ATP and energy in most cell types, but cancer cells extensively use glycolysis to sustain anabolism, which is necessary for tumour growth [
50]. We found that the combinations inhibited glycolysis in both cell lines as indicated by lower levels of fructose bisphosphate and phosphopyruvate. In addition, several TCA cycle intermediates were lowered. Cell death can be executed by different mechanisms, including apoptosis, autophagy, necrosis, or combinations of these processes. Although different cell death mechanisms are unique in their molecular signalling cascades, one molecule is involved in the processes that mediates all types of cell death; ATP. During late-stage apoptosis, ATP levels sharply drop, mostly because of the loss of mitochondrial function and consumption by ATP-dependent proteases. In autophagy, a rescue process of self-degradation to compensate for energy paucity occurs, that also features ATP insufficiency prior to cell death [
50,
51]. During necrosis, depletion of ATP precedes mitochondrial permeability changes [
52]. The fact that ATP deprivation occurs in all types of cell death suggests that energy metabolism may play a critical role in the survival of cancer cells under stress. Thus, it could be possible that the A2780 cells may be undergoing late-stage apoptosis cell death in response to the combination treatment whereas the A2780CR may be undergoing early-stage apoptosis cell death. In our study, we found that combination treatment probably inhibited glycolysis in A2780 cells by depletion of NAD
+. Moreover, the level of NAD
+ was found to be decreased in A2780CR cells after combination treatment. It appears that the combinations had more impact on the oxidative phosphorylation pathway in both cell lines in comparison with melittin as a single treatment [
15]. The inhibition of NAMPT leads to suppression of tumour cell growth and induction of apoptosis due to NAD
+ depletion [
53]. NAMPT represents a promising therapeutic target for the development of potential novel cancer drugs [
54,
55]. In most cancer cells, poly (ADP-ribose) polymerase is activated due to DNA damage and cell death induced by oxidative stress [
56,
57]. Therefore, NAMPT inhibition leads to attenuation of glycolysis, resulting in further alteration of the carbohydrate metabolism in the cells [
53].