Neurymenolide A, a Novel Mitotic Spindle Poison from the New Caledonian Rhodophyta Phacelocarpus neurymenioides

The marine α-pyrone macrolide neurymenolide A was previously isolated from the Fijian red macroalga, Neurymenia fraxinifolia, and characterized as an antibacterial agent against antibiotic-resistant strains that also exhibited moderate cytotoxicity in vitro against cancer cell lines. This compound was also shown to exhibit allelopathic effects on Scleractinian corals. However, to date no mechanism of action has been described in the literature. The present study showed, for the first time, the isolation of neurymenolide A from the New Caledonian Rhodophyta, Phacelocarpus neurymenioides. We confirmed the compound’s moderate cytotoxicity in vitro against several human cell lines, including solid and hematological malignancies. Furthermore, we combined fluorescence microscopy and flow cytometry to demonstrate that treatment of U-2 OS osteosarcoma human cells with neurymenolide A could block cell division in prometaphase by inhibiting the correct formation of the mitotic spindle, which induced a mitotic catastrophe that led to necrosis and apoptosis. Absolute configuration of the stereogenic center C-17 of neurymenolide A was deduced by comparison of the experimental and theoretical circular dichroism spectra. Since the total synthesis of this compound has already been described, our findings open new avenues in cancer treatment for this class of marine molecules, including a new source for the natural product.


Introduction
Cell division and programmed cell death constitute fundamental processes in the maintenance of tissue homeostasis. Indeed, dysfunctions in these mechanisms can lead to genetic instability, including transition in U-2 OS human osteosarcoma cells, which then results in mitotic catastrophe leading to necrosis and apoptosis. Moreover, our determination of the absolute configuration of the stereogenic carbon provides a helpful and necessary step toward elucidating the structure-activity relationship between neurymenolide A and its pharmacological action, particularly once its molecular targets have been identified precisely.

Neurymenolide A Has Mild to Moderate Impact on the Growth of Cancerous Human Cell Lines
Neurymenolide A was screened for its effect on cell viability against a set of human cell lines selected to represent various cancer types: MCF-7 breast adenocarcinoma, HT-29 colorectal adenocarcinoma, Hep G2 hepatocellular carcinoma, IMR-32 neuroblastoma, U-2 OS osteosarcoma, AsPC-1 pancreatic adenocarcinoma, PANC-1 pancreatic epithelioid carcinoma, THP-1 acute monocytic leukemia, A3 acute T cell leukemia, K-562 chronic myelogenous leukemia, adriamycin-resistant K-562/ADR chronic myelogenous leukemia (Adriamycin ® , Pfizer, New York, NY, USA) and HEK-293 embryonic kidney non-cancerous cell line.
Neurymenolide A decreased the viability of cancer cell lines in a 24-h assay, with half-maximal inhibitory concentrations (IC 50 ) ranging from 73.3 to 446.5 µM ( Table 1). The efficacy of cancer chemotherapy is critically dependent upon tumor cell selectivity. For example, a selectivity index (SI) value less than two indicates inadequate selective toxicity toward cancer cells [19]. As Table 1 shows, neurymenolide A is 3.3 times more potent towards A3 acute T cell leukemia than the normal human embryonic kidney cell line, HEK-293, in vitro. This compound also showed favorable selectivity toward THP-1, U-2 OS, and MCF-7 cell lines with SI values of 2.5, 2.4 and 2.3, respectively (Table 1).  U-2 OS, which was the adherent cell line most sensitive to neurymenolide A among those we assayed, was subsequently used to investigate the mechanism(s) by which this compound triggers cell death.

Neurymenolide A Induces Necrosis and Apoptosis in U-2 OS Human Osteosarcoma Cells
The apoptosis of osteosarcoma cells induced by neurymenolide A was demonstrated by flow cytometry analysis, performed following staining with Annexin V and Propidium-Iodide (PI).
Investigations revealed that 102.8 µM neurymenolide A induced a significant increase in the percentage of cells in both necrotic (55.3%) and late apoptotic (28.3%) cell subpopulations ( Figure 1). obtained are shown in brackets. ** Selectivity index (SI): IC50 of pure compound on normal cell line divided by the IC50 of pure compound on cancer cell line. U-2 OS, which was the adherent cell line most sensitive to neurymenolide A among those we assayed, was subsequently used to investigate the mechanism(s) by which this compound triggers cell death.

Neurymenolide A Induces Necrosis and Apoptosis in U-2 OS Human Osteosarcoma Cells
The apoptosis of osteosarcoma cells induced by neurymenolide A was demonstrated by flow cytometry analysis, performed following staining with Annexin V and Propidium-Iodide (PI).
Investigations revealed that 102.8 µM neurymenolide A induced a significant increase in the percentage of cells in both necrotic (55.3%) and late apoptotic (28.3%) cell subpopulations ( Figure 1).

Neurymenolide A Induces Mitotic Spindle Destabilization at Prometaphase in U-2 OS Human Osteosarcoma Cells
As a first approach to study the mechanism(s) of action of neurymenolide A on cancer cell lines, we prepared time-lapse movies of living U-2 OS human osteosarcoma cells stably expressing H2B-mRFP (Histone2B-monomeric Red Fluorescent Protein). These time-lapse studies revealed that a 24 h treatment with 102.8 µM of neurymenolide A induced abnormal localization and dynamics of U-2 OS H2B-mRFP during cell division, whose genetic material looked irregularly star-shaped within the cell. To further explore cell division abnormalities, U-2 OS H2B-mRFP cells were fixed and stained for pericentrin (red), α-tubulin (green) and DAPI (blue) after a 24 h incubation period in the absence or presence of neurymenolide A. The mitotic index (MI), that is, the ratio of the number of cells in a population undergoing mitosis to the total number of cells, was also measured. These results are shown in Figures 2 and 3.

Neurymenolide A Induces Mitotic Spindle Destabilization at Prometaphase in U-2 OS Human Osteosarcoma Cells
As a first approach to study the mechanism(s) of action of neurymenolide A on cancer cell lines, we prepared time-lapse movies of living U-2 OS human osteosarcoma cells stably expressing H2B-mRFP (Histone2B-monomeric Red Fluorescent Protein). These time-lapse studies revealed that a 24 h treatment with 102.8 µM of neurymenolide A induced abnormal localization and dynamics of U-2 OS H2B-mRFP during cell division, whose genetic material looked irregularly star-shaped within the cell. To further explore cell division abnormalities, U-2 OS H2B-mRFP cells were fixed and stained for pericentrin (red), α-tubulin (green) and DAPI (blue) after a 24 h incubation period in the absence or presence of neurymenolide A. The mitotic index (MI), that is, the ratio of the number of cells in a population undergoing mitosis to the total number of cells, was also measured. These results are shown in Figures 2 and 3.
MI values were 7.5% for treated cells and 2.5% for vehicle (DMSO), which suggests that the marine natural product induced a general decrease in the speed of mitosis, or perhaps arrested or slowed a specific phase of mitosis ( Figure 2A). To distinguish these possibilities, we performed time-lapse to show that the number of cells in early mitosis dramatically increased over time (78.9%) after treatment with neurymenolide A, as cells entered mitosis, but failed to proceed through the later stages ( Figure 2B). The images of the cells blocked in an aberrant, prophase-like stage are similar to C-mitosis, i.e., cells treated with colchicine (see for example Sirri et al., 2000, [20]).     Data gathered from image processing confirmed the star-shaped, C-mitosis-like chromosome distribution at the prometaphase transition ( Figure 3A). Quantification revealed that 28.0% of cells in early mitosis showed a misalignment of chromosomes in prometaphase with disorganized spindles (p < 0.01) ( Figure 3B); videomicroscopy demonstrated that these cells underwent mitotic arrest and apoptosis, including the formation of vesicles of cellular debris (see Supplementary Materials Section, Videos S1 and S2).

Neurymenolide A Induces a Delay of Microtubule Repolymerization in U-2 OS Human Osteosarcoma Cells
In order to gain mechanistic information on how neurymenolide A destabilizes the mitotic spindle, we pre-treated U-2 OS cells for 24 h in the absence or presence of our compound. We then performed an in cellulo microtubule repolymerization assay ( Figure 4). Microtubules in pre-treated U-2 OS cells were depolymerized by cold treatment and then re-warmed to allow microtubules to repolymerize, still in the absence or presence of neurymenolide A. Nocodazole (Sigma-Aldrich, St. Louis, MO, USA), a known antagonist of microtubule polymerization, was used as a control. MI values were 7.5% for treated cells and 2.5% for vehicle (DMSO), which suggests that the marine natural product induced a general decrease in the speed of mitosis, or perhaps arrested or slowed a specific phase of mitosis ( Figure 2A). To distinguish these possibilities, we performed time-lapse to show that the number of cells in early mitosis dramatically increased over time (78.9%) after treatment with neurymenolide A, as cells entered mitosis, but failed to proceed through the later stages ( Figure 2B). The images of the cells blocked in an aberrant, prophase-like stage are similar to C-mitosis, i.e., cells treated with colchicine (see for example Sirri et al., 2000, [20]).
Data gathered from image processing confirmed the star-shaped, C-mitosis-like chromosome distribution at the prometaphase transition ( Figure 3A). Quantification revealed that 28.0% of cells in early mitosis showed a misalignment of chromosomes in prometaphase with disorganized spindles (p < 0.01) ( Figure 3B); videomicroscopy demonstrated that these cells underwent mitotic arrest and apoptosis, including the formation of vesicles of cellular debris (see Supplementary Materials Section, Videos S1 and S2).

Neurymenolide A Induces a Delay of Microtubule Repolymerization in U-2 OS Human Osteosarcoma Cells
In order to gain mechanistic information on how neurymenolide A destabilizes the mitotic spindle, we pre-treated U-2 OS cells for 24 h in the absence or presence of our compound. We then performed an in cellulo microtubule repolymerization assay ( Figure 4). Microtubules in pre-treated U-2 OS cells were depolymerized by cold treatment and then re-warmed to allow microtubules to repolymerize, still in the absence or presence of neurymenolide A. Nocodazole (Sigma-Aldrich, St. Louis, MO, USA), a known antagonist of microtubule polymerization, was used as a control. . Microtubule repolymerization assay. Microtubule regrowth was monitored in U-2 OS cells in which microtubules had been cold-depolymerized (i.e., 1 h on ice). Repolymerization is shown at intervals of 0-120 s after shifting the temperature from 0° to 37 °C. Glass coverslips containing U-2 OS cells were fixed in methanol at −20 °C for 10 min, followed by immunofluorescence to visualize pericentrin (red) and α-tubulin (green), and staining with DAPI (blue), as described in the Experimental Section. In merged images, red and green combine to make yellow. Scale bar = 10 µm. Figure 4. Microtubule repolymerization assay. Microtubule regrowth was monitored in U-2 OS cells in which microtubules had been cold-depolymerized (i.e., 1 h on ice). Repolymerization is shown at intervals of 0-120 s after shifting the temperature from 0 • to 37 • C. Glass coverslips containing U-2 OS cells were fixed in methanol at −20 • C for 10 min, followed by immunofluorescence to visualize pericentrin (red) and α-tubulin (green), and staining with DAPI (blue), as described in the Experimental Section. In merged images, red and green combine to make yellow. Scale bar = 10 µm.
The results shown in Figure 4 demonstrate both a delay in the re-polymerization of the microtubules in neurymenolide A-treated cells, compared to the DMSO control, and the inability to reorganize a spindle ( Figure 4).

Neurymenolide A Has R Absolute Configuration at Position C-17
Neurymenolide A is a polyunsaturated α-pyrone derivative isolated for the first time by Stout and his collaborators as two quickly interchanging atropisomers [14] ( Figure 5, see also Supplementary Materials Section, Figure S1). Studies were carried out to determine the absolute configuration of its chiral center, including unsuccessful attempts to utilize crystal X-ray diffraction [16,17]. Despite significant efforts to date, the absolute configuration of C-17 of neurymenolide A had not been reported at the time of this study.
Neurymenolide A is a polyunsaturated α-pyrone derivative isolated for the first time by Stout and his collaborators as two quickly interchanging atropisomers [14] ( Figure 5, see also Supplementary Materials Section, Figure S1). Studies were carried out to determine the absolute configuration of its chiral center, including unsuccessful attempts to utilize crystal X-ray diffraction [16,17]. Despite significant efforts to date, the absolute configuration of C-17 of neurymenolide A had not been reported at the time of this study.  The chiral properties of neurymenolide A include several chromophores such as a pyrone and two double bonds ( Figure 5). Notably, these chromophores are located near the stereogenic carbon C-17. Circular dichroism studies, which consist of comparing calculated ECD (electronic circular dichroism) spectra and experimental circular dichroism spectra, have been shown to be a powerful approach to determine the absolute configuration of chiral compounds that possess relevant

Neurymenolide A Has R Absolute Configuration at Position C-17
Neurymenolide A is a polyunsaturated α-pyrone derivative isolated for the first time by Stout and his collaborators as two quickly interchanging atropisomers [14] ( Figure 5, see also Supplementary Materials Section, Figure S1). Studies were carried out to determine the absolute configuration of its chiral center, including unsuccessful attempts to utilize crystal X-ray diffraction [16,17]. Despite significant efforts to date, the absolute configuration of C-17 of neurymenolide A had not been reported at the time of this study.  The chiral properties of neurymenolide A include several chromophores such as a pyrone and two double bonds ( Figure 5). Notably, these chromophores are located near the stereogenic carbon C-17. Circular dichroism studies, which consist of comparing calculated ECD (electronic circular dichroism) spectra and experimental circular dichroism spectra, have been shown to be a powerful approach to determine the absolute configuration of chiral compounds that possess relevant The chiral properties of neurymenolide A include several chromophores such as a pyrone and two double bonds ( Figure 5). Notably, these chromophores are located near the stereogenic carbon C-17. Circular dichroism studies, which consist of comparing calculated ECD (electronic circular dichroism) spectra and experimental circular dichroism spectra, have been shown to be a powerful approach to determine the absolute configuration of chiral compounds that possess relevant chromophores near the stereogenic centers [21]. Accordingly, we exploited this method to determine the absolute configuration of the C-17 of neurymenolide A.
An experimental CD spectrum of neurymenolide A revealed a weak positive Cotton effect at about 310 nm and a strong negative Cotton effect at about 205 nm (Figure 7). chromophores near the stereogenic centers [21]. Accordingly, we exploited this method to determine the absolute configuration of the C-17 of neurymenolide A.
An experimental CD spectrum of neurymenolide A revealed a weak positive Cotton effect at about 310 nm and a strong negative Cotton effect at about 205 nm (Figure 7).  Table S1. For each atropisomer, the resulting conformers of lower-energy were optimized using the density functional theory (DFT) method. All structures are presented in Figure 8 Table S1. For each atropisomer, the resulting conformers of lower-energy were optimized using the density functional theory (DFT) method. All structures are presented in Figure 8 (see Supplementary Materials Section, Tables S2-S25, for more information). chromophores near the stereogenic centers [21]. Accordingly, we exploited this method to determine the absolute configuration of the C-17 of neurymenolide A. An experimental CD spectrum of neurymenolide A revealed a weak positive Cotton effect at about 310 nm and a strong negative Cotton effect at about 205 nm (Figure 7).  Table S1. For each atropisomer, the resulting conformers of lower-energy were optimized using the density functional theory (DFT) method. All structures are presented in Figure 8    The ECD theoretical spectra were then calculated for each conformer by TDDFT (time-dependent density functional theory). The ECD total spectra, weighted-average Boltzmann associated to each conformer, were then simulated (Figure 9).  Taken together, the experimental and theoretical ECD spectra (calculated at B3LYP/6-31+G(d,p) level) allowed us to assign the absolute configuration of neurymenolide A as 17R.

Discussion
Although neurymenolide A was previously investigated for its antiproliferative activity, the previous experiments did not address its putative mechanism(s) of action. Accordingly, in this  Taken together, the experimental and theoretical ECD spectra (calculated at B3LYP/6-31+G(d,p) level) allowed us to assign the absolute configuration of neurymenolide A as 17R.

Discussion
Although neurymenolide A was previously investigated for its antiproliferative activity, the previous experiments did not address its putative mechanism(s) of action. Accordingly, in this Taken together, the experimental and theoretical ECD spectra (calculated at B3LYP/6-31+G(d,p) level) allowed us to assign the absolute configuration of neurymenolide A as 17R.

Discussion
Although neurymenolide A was previously investigated for its antiproliferative activity, the previous experiments did not address its putative mechanism(s) of action. Accordingly, in this paper, we paid particular attention to the mechanism(s) by which this marine natural product modulated the viability of cancer-derived cell lines in vitro. After confirming the cell growth-inhibiting properties of hundred-micromolar levels of this compound against several human cancer cells, we determined that neurymenolide A could act as a novel mitotic spindle poison. Indeed, our work demonstrates that neurymenolide A significantly delays the in vivo polymerization of tubulin to form microtubules and bi-polar mitotic spindles at the prophase-metaphase transition. Moreover, our results are consistent with the notion that mitotic perturbations induced by neurymenolide A trigger mitotic catastrophe that then results in cell death. Future studies, particularly in vitro experiments, will be required in order to answer basic questions concerning its mechanism(s) of action in deranging the mitotic spindle apparatus. For example, does neurymenolide A bind directly to tubulin subunits, or, more likely, to components of the microtubule organizing center or mitotic motors such as KiF11 (Eg5)? Does it have other bioactivities that are deleterious to cells? Further work will be needed to address these questions.
Notably, in our experiments, neurymenolide A concentrations that prevented mitotic progression or killed cells were much greater (IC 50 > 100 µM) than the concentrations commonly used for therapeutic drugs (generally 1-100 nM). This suggests a number of possibilities: the affinity of neurymenolide A for its target(s) is quite low; when bound to its target protein(s) its inhibition of the activity of targets is only partial; its hydrophobicity limits the amount of active neurymenolide A the cells are exposed to in aqueous culture medium; or (perhaps the most likely) the atropisomers differ in their potency and a racemic mixture therefore contains both highly active and less active forms of the drug. In addition, 102.8 µM and 46.5 µM are the IC 50 observed for U-2 OS cells treated with neurymenolide A for 24 h and 72 h, respectively; a longer incubation with the drug yielded cell death/mitosis prevention at lower neurymenolide A concentrations. In agreement with our results, Stout et al., 2009 [14] treated cells with neurymenolide A for 72 h and reported the killing of cell lines with IC 50 ranging from 3.9-28 µM. We utilized the shorter incubation time in our study as it optimized our ability to explore the mechanistic contributors to cell killing.
Drug-induced mitotic catastrophe represents a promising new approach in cancer prevention and treatment [3]. In addition to microtubules, a number of potential mitotic targets have been exploited, such as motor proteins or mitotic kinases [22,23]. Mitotic spindle kinases that are potential targets include monopolar spindle 1 (Mps1), polo-like kinases (e.g., Plk1) and aurora kinases (e.g., Aurora B) since all play critical roles during mitosis [24][25][26]. Thus far we have performed a molecular screen on a panel of kinase proteins that included HsAurora-B. However, neurymenolide A showed no inhibitory activity towards Aurora B (data not shown).
Neurymenolide A is an equimolar mixture of two atropisomers and, thus, there are two possible forms of the drug. More work will be required to define how quickly each atropisomer can racemize and under what conditions this occurs. If the stereochemistry of both atropisomer is Class III (very stable; see review by Glunz, 2018 for more information, [27]), then separation of each of the two atropisomers will be possible and we can test purified versions of each enantiomer to determine if each works with an identical mechanism and potency. On the other hand, if each racemizes with a very short half-time t 1 2 , (Class I) or an intermediate t 1 the structure will aid in forthcoming structure-activity relationship studies, particularly in vitro studies of neurymenolide A and tubulin. Taken together, the data described here reveal a new putative mitotic spindle poison, neurymenolide A, isolated for the first time from a New Caledonian species of macroalgae. This work has also opened interesting avenues with regards to the mechanism(s) by which this natural product may affect the mitotic spindle and potentially the polymerization of tubulin, a very suitable target for the development of cancer treatments. The key step will now be to lower the effective concentration of the molecule by designing new neurymenolide derivatives and/or by testing synergetic effects with other anticancer drugs [29], in order to render neurymenolide A amenable for use in clinical trials [30]. Neurymenolide A may thus be considered as a novel prototype that could eventually be useful in designing new anticancer drug therapies involving the induction of mitotic catastrophe.

Biological Material
The Rhodophyta Phacelocarpus neurymenioides (Florideophyceae, Gigartinales, Phacelocarpaceae) was collected from Nouméa, in the South lagoon at 70 m deep by SCUBA diving in October 2014. The specimen was identified by Dr. C.E. Payri (IRD Nouméa).
The 1 H NMR (500 MHz) and 13 C NMR (125 MHz) ( Table 2) data are in agreement with published data (see Supplementary Materials Section, Figures S2 and S3, for more information) [16].

Cell Treatments
Neurymenolide A was dissolved in DMSO and stored at −20 • C. The cells were treated for 24 or 72 h with neurymenolide A, vehicle (DMSO), doxorubicin (Adriamycin ® , Pfizer, New York, NY, USA) or nocodazole.

Flow Cytometry Analysis
The induction of apoptosis by neurymenolide A was assessed by flow cytometry using the FlowCellect Annexin Red kit (Merck Millipore, Billerica, MA, USA) and PI/RNase staining buffer (BD Pharmingen TM , BD Biosciences, San Diego, CA, USA). U-2 OS cells (5 × 10 5 cells/well) were distributed in 6-well plates and incubated in the absence or presence of neurymenolide A (102.8 µM) in a complete growth medium for 24 h. Cells were then washed and stained with Annexin V and propidium iodide, following the manufacturer's instructions. Data acquisition and analysis were performed by a BD FACsAria TM flow cytometer (BD Biosciences, San Jose, CA, USA) using BD FACSDiva TM software v. 6.1.3 and the PI − /Annexin V − , PI − /Annexin V + , PI + /Annexin V + and PI + /Annexin V − populations, corresponding to viable (Q3), early apoptotic (Q4), late apoptotic (Q2) and necrotic (Q1) cells, respectively.

Immunofluorescence Staining
U-2 OS H2B-mRFP cells (5 × 10 5 cells/Petri dishes) were incubated in the absence or presence of neurymenolide A (102.8 µM) for 24 h at 37 • C and 5% CO 2 . The cells were fixed with freshly prepared 4% paraformaldehyde for 10 min at room temperature (RT). Following three 5 min washes in PBS, the cells were mildly permeabilized with PBS containing 0.15% Triton for two minutes at RT. The cells were washed thrice in PBS for five minutes each and then blocked with 1% bovine serum albumin at RT for one hour. Incubation with primary antibodies was performed (1:1000 dilution of rabbit polyclonal pericentrin antibody, Novus NB.100-61071 1 mg/mL; 1:1000 dilution of mouse monoclonal α-tubulin antibody, Sigma-Aldrich T5168 100 U/mL), washed intensively and incubated with a 1:1000 dilution of the appropriate secondary antibodies labeled with Alexa 488 and Alexa 594 dyes (Invitrogen). After staining, coverslips were mounted in Mowiol with 4 ,6-diamidino-2-phenylindole DAPI (Mounting Medium for fluorescence with DAPI, H1200, Vector Laboratoire, lot: ZA111). Data acquisition and image processing were performed by a Microscope Zeiss (camera cool SNAP HQ2, 100X objective) using MetaMorph ® (MetaMorph Inc., Nashville, TN, USA) and ImageJ (National Institutes of Health, Bethesda, MD, USA) software.

Time-Lapse Movies
U-2 OS H2B-mRFP cells were treated at 37 • C and 5% CO 2 in the absence or presence of neurymenolide A (102.8 µM) for eight hours. The cells were then observed under a fluorescence microscope for 16 h at 37 • C. Data acquisition, including image capture (one image every 15 min) and image processing were performed on a Zeiss Microscope as described above.

Microtubule Regrowth Assay
U-2 OS cells were incubated at 37 • C and 5% CO 2 for 24 h in the absence or presence of neurymenolide A (102.8 µM). As a positive control, the cells were also treated with nocodazole (1 µM for 12 h), which is known to interfere with the polymerization of microtubules for 12 h; as a negative control cells were incubated with the vehicle alone (0.08% DMSO). Following drug treatment, coverslips were transferred into pre-cooled medium on ice for one hour. Coverslips were removed with forceps and immediately immersed into a pre-warmed medium at 37 • C for defined time intervals (0, 30, 60 and 120 s). Regrowth was then stopped by methanol fixation (10 min at −20 • C). Thereafter, incubation with primary pericentrin and α-tubulin antibodies was performed, coverslips were then washed intensively and incubated with appropriate secondary antibodies, and processed and imaged as described above in Section 4.8.

Circular Dichroism Data
Circular dichroism data were obtained using a J-810 spectropolarimeter (Jasco). TDDFT calculations of electronic circular dichroism (ECD) were performed with Gaussian 09 software (Gaussian Inc., Wallingford, CT, USA). The analysis of the distribution of the conformers and the optimized geometries were obtained using the semi empiric force field AM1 found in Spartan 08 software (Wavefunction Inc. Irvine, CA, USA). For each derivative, the structures were screened and filtered in order to avoid duplication. Then, each conformer was optimized in regards to the geometry using the DFT hybrid method B3LYP with the calculation bases 6-31+G(d,p) (B3LYP/6-31+G(d,p)), leading to thermochemical parameters at 298 K and 1 atm. The effects of solvation have been modeled using a continuous polarization model (PCM). Based on the TDDFT calculations carried out on the structure of each optimized conformer, the excitation energies (in nm) and the rotational force, R, broken down into velocity dipole (R vel ) and length dipole (R len ), were simulated using an ECD curve with the subsequent Gaussian function: where σ is the width of the band at 1/e peak-height, Ei and Ri are the excitation energy and rotator strength for transition i, respectively. Values of σ = 0.20 eV and R vel were used. SpecDis 1.60 software was employed to generate the spectra in Gaussian shape. The ECD spectra Boltzmann populations-weighted were obtained from the structures which were optimized at B3LYP/6-31+G(d,p) level.

Statistical Analysis
All data are expressed as mean ± standard deviation (SD). A Chi-squared test was used to compare the actual difference between the two populations. p < 0.01 was considered to be statistically significant and p < 0.001 highly significant.

Conclusions
In summary, the data here show that the α-pyrone macrolide neurymenolide A, isolated from the New Caledonian Rhodophyta P. neurymenioides, is able to induce a mitotic catastrophe, as well as cell death by both necrosis and apoptosis in U-2 OS human osteosarcoma cells.
To the best of our knowledge, our work constitutes a first attempt to address the mechanism of action by which neurymenolide A inhibits the division of cancer cells. Our results suggest that this marine natural product may act on spindle organization by delaying microtubule polymerization during the prometaphase transition. Further studies are underway to definitively identify neurymenolide A's targets and to explore structure-activity relationships.
In summary, our findings could possibly lead to strategies to exploit mitotic catastrophe for the treatment of cancer, especially for malignancies that are currently not amenable to treatment.