Malten, a new synthetic molecule showing in vitro antiproliferative activity against tumour cells and induction of complex DNA structural alterations

Background: Hydroxypyrones represent several classes of molecules known for their high synthetic versatility. This family of molecules shows several interesting pharmaceutical activities and is considered as a promising source of new antineoplastic compounds. Methods: In the quest to identify new potential anticancer agents, a new maltol (3-hydroxy-2-methyl-4-pyrone)-derived molecule, named malten (N,N′-bis((3-hydroxy-4-pyron-2-yl)methyl)-N,N′-dimethylethylendiamine), has been synthesised and analysed at both biological and molecular levels for its antiproliferative activity in eight tumour cell lines. Results: Malten exposure led to a dose-dependent reduction in cell survival in all the neoplastic models studied. Sublethal concentrations of malten induce profound cell cycle changes, particularly affecting the S and/or G2-M phases, whereas exposure to lethal doses causes the induction of programmed cell death. The molecular response to malten was also investigated in JURKAT and U937 cells. It showed the modulation of genes having key roles in cell cycle progression and apoptosis. Finally, as part of the effort to clarify the action mechanism, we showed that malten is able to impair DNA electrophoretic mobility and drastically reduce both PCR amplificability and fragmentation susceptibility of DNA. Conclusion: Taken together, these results show that malten may exert its antiproliferative activity through the induction of complex DNA structural modifications. This evidence, together with the high synthetic versatility of maltol-derived compounds, makes malten an interesting molecular scaffold for the future design of new potential anticancer agents.

Despite increasing advances in modern treatments for tumours, the survival of many patients affected by cancer remains extremely poor. The high possibility of relapse, the absence of specific therapies and resistance to standard chemotherapy represent the main causes of failure in cancer treatment (Souhami and Tobias, 2005). There is an urgent necessity for research focused on the identification and description of new antitumour agents.
Hydroxypyrones and their close congeners, hydroxypyridinones, comprise several classes of molecules characterised by high synthetic versatility. Their affinity for a wide range of metal ions makes them ideal metal chelators in the formulation of therapeutic and diagnostic metallopharmaceuticals. These families of molecules are used, or have good potential, in clinical applications, such as in the improvement of metal ion balance in disorders of metal metabolism, the enhancement of tissue uptake and retention of insulin-enhancing agents, as well as in protection from heavy metal ion toxicity and attenuation of redox metal neurotoxicity (Thompson et al, 2006). Maltol (3-hydroxy-2-methyl-4-pyrone), one of the best known hydroxypyrones, is a natural compound extensively used for its flavour and antioxidant properties (Gralla et al, 1969;Bjeldanes and Chew, 1979) as an additive in food, beverage, tobacco, brewing and cosmetics. Although its high bioavailability and favourable toxicity profiles have been known for a long time, maltol was found to perform interesting antineoplastic activities against different cancer cellular models. Reactive oxygen species (ROS) generation and the consequent induction of DNA breaks have been hypothesised as a possible mechanism responsible for this activity (Hironishi et al, 1996;Yasumoto et al, 2004;Murakami et al, 2006a, b).
Furthermore, several maltol-derived compounds have been synthesised and exploited in the formulation of new potential metal-based antitumour drugs that can interfere with DNA (e.g., DNA binding, DNA intercalation and ROS generation), providing interesting results (Jakupec and Keppler, 2004;Barve et al, 2009;Kandioller et al, 2009). In the belief that these compounds have good potential as anticancer agents, we synthesised new maltolderived molecules and investigated their biological activity.
In this study, the antineoplastic potential of a selected molecule, malten (N,N 0 -bis((3-hydroxy-4-pyron-2-yl)methyl)-N,N 0 -dimethylethylendiamine), was evaluated for its ability to decrease the survival of eight different tumour cell lines. Subsequently, cell cycle modifications and the induction of programmed cell death were investigated at both biological and molecular levels to clarify the biological response to malten exposure. Finally, the malten action mechanism was investigated in both cellular and cell-free assays, directing attention to the possible cause of changes in DNA structure.

Chemicals and synthesis of malten
All chemicals and compounds used in this study maltol, N,N 0 -dimethylethylendiamine (DMEDA), 3-hydroxy-2-dimethylaminomethyl-4-pyrone (DMAMP), and cisplatin (CDDP) were of the highest quality, commercially available and purchased from Sigma-Aldrich (St Louis, MO, USA). The synthesis of malten (research paper in preparation) was carried out as follows: a protected maltol derivative was synthesised by silylation and allylic bromination of commercial maltol. After removal of the hydroxyl protective group and without purification, the maltol derivative was combined with N,N 0 -dimethylethylenediamine. Malten was then purified by precipitation as perchlorate salt, obtaining an overall yield of 50%.
Malten was dissolved at 25 mM in double-distilled water as stock solution, stored at À801C and subsequently diluted before use. Treatments were carried out at the concentrations reported in the figures and repeated every 24 h. Cellular viability was evaluated by Trypan blue dye exclusion assay 72 h after treatment started. The 50% inhibitory concentration (IC50) was calculated for each cell line as the drug concentration, which resulted in a 50% reduction of viable cells compared with untreated cells.

Cell cycle analysis and hypodiploid cells estimation
Cell cycle and hypodiploid cells were analysed using the propidium iodide staining procedure as previously reported (Amatori et al, 2009). Cells were fixed in ice-cold 70% ethanol and stained using a propidium iodide staining solution (50 mg ml À1 ). Cytofluorimetric acquisitions were carried out with a BD FACScalibur flow cytometer (BD Biosciences, San Jose, CA, USA) and sample analysis was carried out with FlowJo 8.6.3 software (Tree Star, Inc., Ashland, OR, USA). Cell cycle percentage values were calculated using a Watson pragmatic model and the significance of changes was evaluated by a Student t-test (Po0.05).

DNA laddering assay
About 2 Â 10 6 cells were harvested, washed once with ice-cold PBS and incubated in 2 ml of 1 Â TBS, 0.5% Tween20 and 1 mM EDTA for 30 min at 41C. Cells were centrifuged at 1200 r.p.m. for 5 min at 41C and incubated with 0.1% SDS solution for 30 min at room temperature, in a final volume of 0.5 ml. DNA was purified using a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany), following the manufacturer's instructions, separated by 2% agarose gel electrophoresis (AGE) and detected by ethidium bromide staining.
Primers were designed using Primer Express software (PE Applied Biosystem, Foster City, CA, USA) and PCR assays were applied in a Rotor-Gene 6000 robocycler (Corbett Life Science, Sydney, Australia). The expression of different transcripts was normalised with GAPDH expression levels, and calculated as fold induction with respect to untreated cells. Filtering of results was carried out as follows: genes were considered regulated when their change was greater than ±2.5-fold.

DNA electrophoretic mobility assay
A volume of 100 -500 ng of circular pLL3.7 plasmid DNA (resuspended in 1 Â TE) was incubated with the compounds tested in 20 ml of 10 mM Tris-HCl (pH 7.4) buffer (371C), at the reported concentrations and time points. After incubation, DNA was separated by 0.8% AGE and then stained by ethidium bromide. Ascorbic acid and CuSO 4 at equimolar concentrations (50 nM) were used as a control of plasmid linearisation induced by DNA breaks (Gaetke and Chow, 2003).
Alternatively, linear pLL3.7 plasmid DNA, previously incubated with malten or CDDP for 2 h at 371C under conditions already described, was partially fragmented through mild sonication (once Â 10 s at 20 W -Labsonic L sonicator -B. Braun Biotech International, Melsungen, Germany) and separated by AGE.
DNA electrophoretic mobility assay of genomic DNA was carried out as follows: U937 cells, treated with malten for 8 h at the concentrations reported, were harvested, washed with ice-cold 1 Â PBS, resuspended in a solution containing 1 Â PBS, 0.1% SDS, 10 mg/ml RNase A and incubated overnight at 41C. Genomic DNA was separated by AGE and detected using ethidium bromide staining.

Malten synthesis and antiproliferative effects against tumour cell lines
The new maltol-derived molecule malten, prepared as shown in Figure 1 (described in Material and Methods), was obtained after precipitation as a white solid perchlorate salt (L . 2HClO 4 ), lightand air stable, easy to handle and, more importantly for future applications, soluble in water.
Different tumour cell lines, comprised of cells derived from both solid and haematopoietic malignancies, were subjected to 72 h of exposure to malten in dose -response experiments to evaluate the biological effects. We found that treatments with malten at concentrations ranging between 5 and 50 mM result in a dosedependent reduction in cell survival ( Figure 2A). Subsequently, IC50 value was calculated for each cell line showing appreciable variations (from 4.54 ± 0.66 mM to 30.47 ± 6.75 mM) and indicating a certain cell-type specificity of malten activity ( Figure 2B). Interestingly, RH30 was the most sensitive cellular population suggesting the presence of peculiar molecular features in alveolar rhabdomyosarcoma that could render this cell line more susceptible to malten exposure.

Analysis of malten-mediated cell cycle perturbations and induction of programmed cell death
The biological response to malten was further detailed by analysing the effects on cell cycle progression. Malten treatment of 72 h, at concentrations of 5, 10 and 25 mM, induced cell cycle perturbations that could be monitored in all the cell lines considered in this study. The effects were dose dependent and are characterised by the accumulation of cells mainly in S and/or G2-M phases ( Figure 3; Table 1 and Supplementary Figure S1).  Interesting differences between the cell lines were also detected. In particular, JURKAT and RH30 cells showed a robust reduction of cells in G1 phase concomitantly with an evident G2-M cell cycle arrest ( Figure 3 and Table 1). Malten-mediated apoptosis was also evaluated, after 72 h of treatment, monitoring an increase of hypodiploid cells ( Figure 4A and Supplementary Figure S2), with differences observed between the cell lines tested. Activation of the apoptotic programme was subsequently confirmed by analysing the internucleosomal DNA fragmentation induced by malten exposure in U937 cells, at concentrations of 10, 25 and 50 mM ( Figure 4B). The biological effects mediated by malten treatments were also investigated at the molecular level by analysing the expression of a panel of genes known to regulate cellular proliferation, apoptosis and cell cycle progression.
JURKAT and U937 cells were chosen as representative model systems that recapitulate the different biological responses of the cell lines tested. The results obtained by comparing untreated cells with those treated with malten at concentrations of 8 and 10 mM (across the IC50 value) show changes in the mRNA level involving genes having key roles in the control of cell cycle and DNA damage response ( Figure 5A Figure S1 for the other cell lines tested).
The upregulation of BRCA1 and CUL3 at both doses of malten was only observed in JURKAT cells ( Figure 5A). Subsequently, genes showing the highest transcriptional modulations were further investigated at protein level. Western blot analysis confirmed the increased expression of p21, p16 and p15 in both cell lines, whereas no changes were monitored for PCNA and CDK6 proteins as they were already monitored at mRNA level ( Figure 5C).
P53 is known to be functionally activated prevalently by protein stabilisation, hence its accumulation was monitored. We observed a slight increase in JURKAT cells after malten treatments and, as expected (Decker et al, 2003), the complete absence of expression in U937 cells ( Figure 5C). The expression level of BRCA1 protein was not detected, probably because of the low sensitivity of the antibody tested.

Evaluation of DNA structural modifications induced by malten
As already reported, maltol can damage DNA through ROS production (Murakami et al, 2006a, b). We first tested the hypothesis that malten could cause DNA breaks by analysing its effect on circular plasmid DNA by AGE. Dose -response experiments showed that after 16 h of malten exposure, there was neither plasmid linearisation nor evident DNA degradation ( Figure 6A). However, we observed that malten is able to modify the electrophoretic mobility of plasmid DNA causing the complete trapping of DNA at the top of the gel at the higher concentrations tested. This phenomenon is strictly dependent on incubation, as shown by the normal pattern of migration of plasmid DNA at the higher malten dose (4 mM) when incubation was omitted ( Figure 6A, lane n.i.), indicating that the presence of malten per se does not disturb the electrophoretic migration of DNA. DNA complexes with impaired electrophoretic mobility were also found using shorter exposure times and a malten concentration of 4 mM ( Figure 6B).
We then investigated malten-induced effects on the genomic DNA of cancer cells. U937 cells were exposed to different doses of malten (from 0.1 to 1 mM) for a relatively short period of time (8 h) to maximise the effect of the compound and avoid any interference from the biological response (e.g., internucleosomal cleavage of DNA related to the apoptotic response). We found that U937 cells exposed to malten show a concentration-dependent reduction of electrophoretic mobility of their genomic DNA ( Figure 6C).
Having established that the induction of DNA modifications can also occur in the cellular compartment, we investigated whether this phenomenon was a peculiar feature of the malten chemical  structure. To this end, plasmid DNA was exposed to compounds such as maltol, DMEDA and DMAMP, which represent part of the malten molecule and CDDP. Intriguingly, we found that the electrophoretic interferences on plasmid DNA are peculiar to the malten molecular structure and that CDDP, as expected (Bellon et al, 1991), causes just a slight retardation in the mobility of the supercoiled plasmid form ( Figure 6D). Moreover, in accordance with cell-free studies, maltol, DMEDA and DMAMP failed to induce a biological response in any of the cellular models tested (data not shown).
Finally, we further investigated the possibility that malten could molecularly modify the DNA structure by assaying its interference on DNA amplification. Plasmid DNA was amplified after malten exposure, by Q-PCR, using three sets of primers sharing the same Relative expression (folds) reverse oligonucleotide and able to generate amplicons of different sizes (121, 301 and 622 bp). First, we revealed that incubation with malten decreases the efficiency of PCR amplification ( Figure 6E, upper panel). It is important that, despite doubling the amplicon size (121, 301 and 622 bp), malten induced an exponential loss of PCR efficiency (À7.13-fold, À99.7-fold, À6472-fold; Figure 6E upper panel-see Discussion). Taken together, these results strongly support the hypothesis that malten can induce DNA structural modifications at a high level of complexity.

DISCUSSION
In this paper, we provide the biological properties of a new synthetic maltol-derived molecule (malten) and the preliminary attempt to understand its molecular mechanism of action.
The sensitivity to malten exposure of different tumour cell lines (derived from haematopoietic and solid tumours) was assayed in terms of decreased cell survival, programmed cell death induction and cell cycle perturbation.
Malten was able to drastically reduce the cell survival of all cell lines tested, showing interesting differences in terms of responsiveness and suggesting the presence of peculiar molecular properties that may render some cellular models more susceptible. High doses of malten activated the apoptotic cellular response, whereas sublethal treatments induced the accumulation of cells in both S and/or G2-M phases in almost all cell lines.
However, JURKAT and RH30 cells, two of the most sensitive cell lines in terms of reduced cell survival, showed a peculiar accumulation of cells in G2-M phases associated with a strong decrease of G1 phase, indicating that a possible different ability to activate cell cycle checkpoints may influence the antiproliferative effect of malten.  Gene expression studies, conducted at both mRNA and protein levels, support the biological data collected. In particular, upregulation of the broad-range cyclin-dependent kinase inhibitor (CDKi) p21, found in both JURKAT and U937 cells, is consistent with the late S and G2-M cell accumulation that we monitored.
However, the lack of p53 functionality in both cell lines (Cheng and Haas, 1990;Decker et al, 2003) indicates that the expression of p21 is induced in a p53-independent manner.
In addition to p21, we also found, in both JURKAT and U937 cells, upregulation of the INK4 family CDK-inhibitors p15 and p16, which are known to exert their activity during G1 -S transition.
An increased expression of CUL3, a member of the cullin-RING ligase family with ubiquitinating ligase activity, was detected in JURKAT cells. Considering the crucial role of CUL3 protein in sustaining caspase-8 activity, and the robust increase of hypodiploidy observed in JURKAT cells exposed to malten, it is reasonable to hypothesise that CUL3 upregulation could contribute to the malten-induced apoptotic response (Békés and Salvesen, 2009;Lall, 2009).
The biological and molecular aspects monitored strongly indicate that malten exposure may activate a DNA damage response. The late S and G2-M block is usually associated with compounds known to form adducts with DNA and actually included in several therapeutic protocols (e.g., CDDP; Siddik, 2003).
In addition to p16 upregulation, which can also be induced as a consequence of DNA damage (Robles and Adami, 1998;Shapiro et al, 1998), malten treatments increased the expression of BRCA1, which is known to be involved in DNA-damage signalling through its multifunctional properties as controller of DNA replication, activator of G2-M checkpoint and as initiator of DNA repair processes (Huen et al, 2010).
Finally, the slight accumulation of p53 protein in JURKAT cells, although not functional, can be considered as a further phenomenon that supports the induction of a cellular response to DNA damage.
As a consequence of these observations, the possible DNA structural alterations induced by malten have been investigated by cell-free studies. We showed that malten exposure induces the formation of DNA molecules with strongly impaired electrophoretic mobility.
In addition, we showed that treatments of DNA with compounds that are part of the malten molecule did not show any DNA modification, indicating that changes in the electrophoretic properties of DNA are a peculiar feature of the malten molecule.
Interestingly, a similar interference with DNA electrophoretic mobility has been already described for some DNA alkylating compounds (Pereira et al, 1998;David-Cordonnier et al, 2002;Novakova et al, 2009) and has been associated with the formation of a DNA matrix of high molecular weight, which results from the interhelical crosslinking of multiple DNA molecules.
In linking cell-free assays with cellular response, it is extremely important to note that genomic DNA, extracted from maltentreated U937 cells, shows a dose-dependent reduction in its  electrophoretic mobility. This finding suggests that the DNAmodifying function of malten could also occur in a cellular context. Moreover, we accumulated further experimental observations suggesting that malten could trigger complex DNA structural modifications. The first indication is represented by the ability of malten to reduce the number of plasmid DNA available for amplification and thus decrease the efficiency of PCR assay ( Figure 6, upper panel). Surprisingly, when amplicon length was increased (doubling the amplicon length from 121 to 622 bp), the number of amplifiable sequences decreases exponentially after incubation with malten ( Figure 6E, upper panel).
In our opinion, these results are of great interest, especially considering the fact that the same assay shows a different behaviour if applied to plasmids exposed to CDDP. In fact, we reproduced a CDDP-dependent PCR inhibition similar to that observed by malten, considering the 121 bp amplicon (loss of amplificability of around seven-to eight-fold; Figure 6E, lower panel -see Supplementary Figure S3 for treatment of CDDP at the same concentration of malten). From this, we monitored a proportional decrease in PCR efficiency as the amplicon was progressively increased in length: CDDP modifies the single DNA molecules and is not able to generate intermolecular crosslinking (different DNA molecules covalently bound); thus, PCR inhibition is expected to be proportional to the increase in amplicon length. The observed exponential decreased amplificability of DNA induced by malten suggests the induction of DNA structural alterations at a higher level of complexity.
These observations, together with the evidence that maltenmodified plasmid DNA is resistant to fragmentation by both sonication ( Figure 6E) and nuclease digestion (see Supplementary Figure S4), whereas CDDP-treated DNA is fragmentable, strongly sustain the idea that malten could induce the formation of covalently bound DNA structures.
However, at this stage, other activities, such as the induction of ROS-mediated DNA damage already reported for higher doses of maltol with respect to those tested in our study (Murakami et al, 2006a, b), cannot be completely excluded as part of the malten mechanism of action.
Although more studies will be necessary to completely understand the activity of malten, and the precise mechanism of action, we believe that this compound can be considered as an interesting molecular scaffold, worthy of further investigation, with the aim of developing new potential anticancer agents.