In vitro cytotoxicity and anticancer effects of citral nanostructured lipid carrier on MDA MBA-231 human breast cancer cells

Very recently, we postulated that the incorporation of citral into nanostructured lipid carrier (NLC-Citral) improves solubility and delivery of the citral without toxic effects in vivo. Thus, the objective of this study is to evaluate anti-cancer effects of NLC-Citral in MDA MB-231 cells in vitro through the Annexin V, cell cycle, JC-1 and fluorometric assays. Additionally, this study is aimed to effects of NLC-Citral in reducing the tumor weight and size in 4T1 induced murine breast cancer model. Results showed that NLC-Citral induced apoptosis and G2/M arrest in MDA MB-231 cells. Furthermore, a prominent anti-metastatic ability of NLC-Citral was demonstrated in vitro using scratch, migration and invasion assays. A significant reduction of migrated and invaded cells was observed in the NLC-Citral treated MDA MB-231 cells. To further evaluate the apoptotic and anti-metastatic mechanism of NLC-Citral at the molecular level, microarray-based gene expression and proteomic profiling were conducted. Based on the result obtained, NLC-Citral was found to regulate several important signaling pathways related to cancer development such as apoptosis, cell cycle, and metastasis signaling pathways. Additionally, gene expression analysis was validated through the targeted RNA sequencing and real-time polymerase chain reaction. In conclusion, the NLC-Citral inhibited the proliferation of breast cancer cells in vitro, majorly through the induction of apoptosis, anti-metastasis, anti-angiogenesis potentials, and reducing the tumor weight and size without altering the therapeutic effects of citral.

Scientific RepoRts | (2019) 9:1614 | https://doi.org/10.1038/s41598-018-38214-x route for the delivery of drugs. However, most of the therapeutic drugs have extremely low levels of oral bioavailability, water solubility, and sustainability problems 3 . In statistic, almost 70% of new drugs formulated are showing poor water solubility, which becomes the limiting factor in the absorption drug after oral admission 4 . Many approaches have been adopted to increase the drug solubility, sustainability, bioavailability and gastrointestinal permeability 5 . Nano-carrier has gained tremendous attention in the development of new pharmaceutical carrier and delivery system. One of the strategies to thwart this problem is to encapsulate citral into the biodegradable and biocompatible nanoparticle. Nanostructured lipid carrier (NLC) is one of the most recent lipid-based nanoparticles that enhanced the water solubility of the hydrophobic drug, increase the drug loading capacity and stability to accommodate more drug and avoid expulsion. Thus, drug delivery system particularly NLC is engaged to minimize the side effects while maintaining the effectiveness of the drug through prolonged sustainability 6,7 . The physicochemical properties of NLC were confirmed to be suitable as a potential delivery system of citral and nontoxic towards the healthy cell. Currently, there is no study that has been investigated on the NLC-Citral and its therapeutic properties although citral is known to have anticancer properties. To the best of our knowledge, this is the first report on the anti-cancer effects of NLC-Citral on human breast cancer cell line MDA MB-231 and murine breast cancer model.

NLC-Citral inhibited the proliferation of MDA MB-231 cells.
Anti-proliferative effects of NLC-Citral on human breast cancer cell line (MDA MB-231) and the non-transformed mammary epithelial cell line (MCF-10A) were assessed by MTT assay. Cells were treated with 2-fold serial dilutions of NLC-Blank, NLC-Citral, and citral for up to 72 hours. The IC 50 of NLC-Citral (18.2 ± 0.71 µg/mL) on MDA MB-231 was substantially lower than citral alone (20.5 ± 1.41 µg/mL) (Fig. 1B) at 48 hours treatment. In addition, it can be perceived that the IC 50 of each treatment on MCF-10A was relatively not detected even at the highest concentration tested (Fig. 1B). Meanwhile, the cytotoxicity of NLC-Blank was concurrently determined using MTT assay. There was no IC 50 detected for NLC-Blank on both cell lines at 48 hours post-treatment.
Based on the result from MTT assay, 3 doses of NLC-Citral (IC 25 , IC 50 , and IC 75 ) ( Table 1) on MDA-MB-231 were selected and used for the following assays.  75. In addition, after 48 hours of treatment with NLC-Citral, this study revealed that the percentage of Sub G0/ G1 cells also escalated from 1.21 ± 1.07% in the IC 25

NLC-Citral inhibited the invasiveness of MDA MB-231 cells.
Since MDA MB-231 cell line is a highly invasive human breast cancer cells, the wound healing, migration and invasion assays were conducted to study the effectiveness of NLC-Citral as an anti-metastatic agent for human breast cancer in vitro. Following treatment with 12.5 µg/mL of NLC-Citral for 24 hours, the substantial decrease in the percentage of wound closure has been displayed. The percentage of wound closure in NLC-Citral treated cells was 51.1 ± 1.22% while it was 95.9 ± 2.05% in the NLC-Blank treated cells (Fig. 6). There was also a significant decrease in the percentage of wound closure in NLC-Citral treated cells as compared to citral treated cells (66.4 ± 2.22%). Anti-migration effects of NLC-Citral and citral were further examined through the transwell in vitro migration assay. Figure 7 indicates that the number of cells migrated in NLC-Citral was significantly decreased by 13-fold from NLC-Blank. Nonetheless, the citral has declined the number of migrated cells substantially from NLC-Blank by 4-fold. On the other hand, the invasiveness of MDA MB-231 cells was tested under treatment of NLC-Citral and citral alone through a matrigel. This assay was conducted to further study the effectiveness of NLC-Citral in controlling the MDA MB-231 cell's invasion properties. From Fig. 8, it was clearly shown that the number of invaded cells has decreased significantly in NLC-Citral and citral by 15-fold and 9-fold respectively. Hence, it can be concluded that the NLC-Citral has quenched the migration and invasion abilities of MDA MB-231 cells in vitro.
An anti-angiogenic potential of NLC-Citral was investigated using ex-vivo mouse aorta ring assay. As depicted in Fig. 9, the number of micro-vessels outgrowth from the thoracic aorta was declined in numbers in an NLC-Citral treated ring as compared to the citral and NLC-Blank. The sprouted vessels formed in the NLC-Citral treated group was reduced by 12-fold as compared to the NLC-Blank group. In contrary, citral treated group showed only 4-foldreduction to NLC-Blank. This implies that the NLC-Citral possessed better anti-angiogenesis potential than citral alone. Microarray-based gene expression profiling. About 1100 genes were up-regulated and 1190 were down-regulated from NLC-Citral over control. On the other hand, 1999 genes were up-regulated and 1855 genes were down-regulated in citral versus control, subsequently. Pathway analysis further revealed the molecular processes that associated with the inhibition of cancer cell development of NLC-Citral. In particular, apoptosis, cell cycle mechanism, and metastasis-related pathways were closely examined. These result demonstrated that NLC-Citral and citral were regulated the changes in the expression level of genes in several signaling pathways that are crucial in cancer-associated activities in MDA MB-231 cells when compared to the control group such as the apoptosis, cell cycle mechanism, and metastasis signaling pathways. In brief, it can be observed that Bax gene was highly regulated by 5.48-fold in NLC-Citral while PTEN has increased by 8.53 in citral treated cells. In contrast, in cell cycle pathway CDKN1B was highly regulated (6.87-fold) in citral and PLK-1 has down-regulated to −3.32 fold in NLC-Citral while not significant in citral treated cells. Additionally, the result showed that GJA-1 gene is the most significantly increased gene by 18.32-fold in NLC-Citral with PXDN (−7.53) as the most down-regulated genes in the metastasis-related pathway.
Cluster analysis provides the better understanding of the degree of association between samples. Based on the heat map displayed in Fig. 10, the differential gene association in NLC-Blank group is closer to citral than NLC-Citral treated group considering the branches formulated in between the group. This showed that the level of gene expression in NLC-Blank is closer to citral than NLC-Citral.
To validate the microarray data, TREX NGS was performed. Seven up-regulated and 5 down-regulated genes were validated. InTruSeq, CDKN11B gene is the most up-regulated genes (13.4-fold) in NLC-Citral treated cells. On the contrary, FZD8 is the most down-regulated gene by 5.3-fold (Table 2). Additionally, 5 genes were selected for additional analysis by qPCR technique. The selected genes were representative of microarray and Truseq data.    The obtained data from protein profiler was validated using ELISA test in which 2 up-regulated and 2 down-regulated proteins were validated. In ELISA, Cytochrome C protein is the most up-regulated proteins (2.8-fold) in NLC-Citral treated cells. On the contrary, Bcl-2 is the most down-regulated protein by 2.5-fold (Fig. 13).

NLC-Citral reduced the weight and size of the 4T1 tumor growth in mice.
In this study, the weight of the NLC-Citral treated tumor has significantly (p < 0.05) reduced to 0.78 ± 0.04 g as compared to the NLC-Blank (2.0 ± 0.13 g) and pure citral (1.31 ± 0.09 g). At the same time, the size of the NLC-Citral treated tumor also significantly (p < 0.05) reduced to 0.69 cm 3 as compared to the NLC-Blank (1.76 cm 3 ) and pure citral (1.31 cm 3 ) (Fig. 15).

Discussion
From the IC 50 value of MTT assay, we realized that NLC-Citral inhibited the proliferation of MDA MB-231 better than citral. Nonetheless, there was an insignificant difference in the inhibitory effect of the cancer cells in between the NLC-Citral and citral alone in MDA MB-231 in both time points. Previous studies reported that the incorporation of drug into nanoparticle did not increase the efficiency of the drug in the inhibition of cancer cells [8][9][10] . In addition, there was no adverse effect on proliferation of normal human breast epithelial cell line MCF-10A at the same dosage and time used for treatments on MDA MB-231 cells. This suggests that the formulation is not toxic to normal cells. Meanwhile, a study suggested that the encapsulation of potential therapeutic agents in lipid nanoparticle generally reduced its toxicity towards normal cells 11 . In vitro cytotoxicity test of NLC-Thymoquinone revealed that NLC is safe in normal liver cells (WRL-68) 12 . Furthermore, relatively low toxicity effects from NLC-Blank reported after being tested for 72 hours suggest that the cause of the cell death was clearly not from the NLC as a carrier itself. In addition, it can be observed from the preliminary MTT assay that NLC-Blank demonstrated no cytotoxic effects on MDA MB-231 cells. This suggests that the strong inhibitory effect of NLC-Citral is majorly from the citral. Rationally, the goal of a drug delivery system is to enhance the bioavailability of the drugs towards targeted diseased cells, promoting the required response while minimizing its side-effects 13 . Considering the slow release effects of the citral loaded in NLC to act upon target cells, it is possible that NLC-Citral to be more effective than pure citral in terms of prolonged anticancer activity.
In the present study, it was found that NLC-Citral significantly inhibited the proliferation of breast cancer cell lines via apoptosis. Based on Annexin V assay, there was a shift in the pattern of the externalization of phosphatidylserine from viable to early apoptosis and to late apoptosis as the concentration of NLC-Citral increased. It was reported that citral has induced the apoptosis of MCF-7 breast cancer cells by increasing the percentage of early apoptotic cells 14 . Simultaneously, NLC-Citral was found to trigger the depolarization of mitochondrial membrane potential in MDA MB-231 cells using JC-1 analysis. The depolarization of mitochondrial membrane potential could be linked with the induction of apoptosis in breast cancer cells 15 . Based on the observation, the depolarization of membrane potential in NLC-Citral treated cells was significantly (p < 0.05) higher than in citral treated cells which reflects apoptotic cells.
Microarray expression profiling and protein assay analysis was performed to gain insight into the molecular activities and pathways related to apoptosis effects of NLC-Citral and citral in MDA MB-231 cells and finally, the results were validated using ELISA technique. Numbers of study have been done in the discovery of biomarkers involved in breast cancer in the identification of genes related to metastasis of breast cancer to other organs 16,17 . Therefore, in this present study, the regulation of differential gene expression that associated with apoptosis, cell    19 . On the other hand, Bcl-2L11 is a well-known gene to mediate cell apoptosis when it is induced by several apoptotic stimuli such as cytokines, radiation and cytotoxic peptide 20 . Recently, it was reported that the association of Bcl-2L11 gene with miR-24 holds potential targets for gene therapy of gastric cancer cells 21 . Interestingly, both of these genes were highly regulated in the NLC-Citral treated cells as compared to the pure citral. This suggests that the regulation of pro-apoptotic genes among the Bcl-2 family was induced better by the NLC-Citral. In addition, PTEN gene is another gene that was highly up-regulated in an apoptosis-related pathway. Phosphatase and tensin homolog deleted from chromosome 10 (PTEN) is a tumor suppressor that can inhibit cellular proliferation, survival, and growth by inactivating PI3-kinase-dependent signaling pathway 22 . Over-expression of PTEN in cancer cells results in cell cycle arrest, cell death, and metastasis. The study reported that PTEN has therapeutic potential for ovarian cancer through the inhibition of angiogenesis in mice tested 23 . Additionally, it has been documented that PTEN could enhance the sensitivity of anticancer agents in ovarian cancer and in human bladder cancer cells 24,25 . Intriguingly, in this case, pure citral has significantly induced the up-regulation of PTEN in MDA MB-231 treated cells than NLC-Citral. This implies that NLC-Citral and citral may regulate differently in different types of a gene in the induction of apoptosis signaling pathway. The previous study has demonstrated that the level of hTERT and PinX genes expression were different in Calu-6 cells treated with nanoparticle loaded curcumin and pure curcumin at the same concentrations and conditions 26 .
Subsequently, proteome profiler analysis proved that NLC-Citral and citral increased the protein expression of Bax, cleaved Caspase 3, TRAIL R1 and Cytochrome C in MDA MB-231 cells while decreased the Pro-Caspase 3, Bcl-2, Bcl-X and Survivin proteins. The Bcl-2 family is consist of a number of protein that plays a crucial role in the regulation of apoptosis where some members of the Bcl-2 family such as Bax appear to induce apoptosis while the other like Bcl-2 and Bcl-X can inhibit apoptosis 27 . Bax is a pro-apoptotic protein that resides in the outer mitochondrial membrane 28 . The up-regulation of Bax triggered a release of Cytochrome C from mitochondria and induced caspases activation into the cytosols 29 . Several studies had suggested that Bax could reduce the mitochondrial membrane potential by creating pores in the membrane for Cytochrome C to escape and eventually cell death 30,31 . Moreover, the release of Cytochrome C was upstream of the initiation of Caspase cascade including Caspase 8, 9 and 3 in enhanced apoptosis 32 . The activation of Caspase 9 may amplify the mitochondrial disruption for the efficient apoptosis executor during intrinsic apoptosis 33 . Besides that, Caspase 8 is involved in the extrinsic pathway of apoptosis and usually recruited by the ligand binding of death receptors and tumor necrosis factor 34 . The activation of death receptor ligand, TRAIL R1 lead to the recruitment of Caspase 8 which triggers a Caspase cascade and subsequently induced apoptosis of susceptible cells 35 . Meanwhile, the induction of cleaved Caspase3 protein expression and the inhibition of pro Caspase 3 in the cells directly targeted the mitochondria for cell death to occur 36 . In the previous study, citral has been found to be a potent inducer in the activation of Caspase 3 enzymatic activity in tumor cell lines 37 . Interestingly, the NLC-Citral was seen to significantly stimulated Caspase 9 and Caspase 8 activities in a dose-dependent manner. Therefore, this result suggested that the apoptogenic effect of NLC-Citral on MDA MB-231 cells is via both the extrinsic and intrinsic pathways.
Furthermore, Bcl-2 and Bcl-X are anti-apoptotic proteins that highly expressed in the chronic cancers. Over-expression of Bcl-2 and Bcl-X prevent the Cytochrome C to initiate the apoptosis by blocking it to be released from the mitochondria thus the apoptosis mechanism to take action 38,39 . Survivin protein was observed to be down-regulated in the NLC-Citral and citral treated cells. Survivin was demonstrated to inhibit the apoptosis by interfering with the function of Caspase 3, 7 and 9 in apoptosis-related pathway either directly or in indirect ways 40 . The study had revealed that high expression of Survivin was associated with the inhibition of apoptosis in 167 of breast cancer patients 41 . In cancer cells, the induction of Survivin is related to the increase in cells proliferation, inhibition of apoptosis, and resistance to chemotherapy 42 . Hence, the regulation of proteins and genes studies represents an important molecular mechanism by which NLC-Citral induced breast cancer apoptosis. Cellular mechanism studies showed that NLC-Citral affected the MDA MB-231 cell cycle machinery. One of the factors in antiproliferative activity of NLC-Citral on MDA MB-231 is the target of cell division by arresting cells at the G2/M phase which is analyzed through flow cytometer system. The key part of cell cycle machinery is the cyclin-dependent kinases and the regulatory protein called cyclins. Superficially, the connection between the cell cycle and cancer is obvious as the cell cycle machinery controls cells proliferation. Several therapeutic agents include flavopiridol, AZD5438, bryostatin-1 and indisulam are targeting the regulation of expression and activity of cyclin-cyclin dependent kinases complexes in cell cycle process 43 . This occurrence was further confirmed with the microarray as the result showed that NLC-Citral and pure citral has regulated the cell cycle signaling pathway. Microarray analysis indicated the down-regulation of PLK-1 gene in NLC-Citral treated cell while this was not significantly detected in citral treated cells. The study revealed that over-expression of PLK-1 contributed to many tumor types, including breast cancer, colorectal cancer, ovarian cancer, non-small lung cancer and many more 44 . The depletion of PLK-1 gene expression was directly promoted G2/M phase arrest in the human cancer cells 45 . In addition, pre-clinical studies using small interfering RNAs (siRNA) inhibited PLK-1 resulting in the G2/M arrest, apoptosis and tumor growth inhibition 44 . As depicted in the microarray analysis, NLC-Citral managed to inhibit the expression of PLK-1 gene in the treated cells. Cell cycle checkpoints help the DNA replication and divisions. Checkpoints are regulated by a family protein kinases, CDKs. Mutation in G2 checkpoint prevents the entry of mitosis in cell cycle process and undergo apoptosis 46 . Cyclin-dependent kinases, on the other hand, are activated by cyclin binding and inhibited by CKD inhibitors 47 . Cyclin B is another gene associated with the cell cycle deregulation in cancer cells. CCNB1 gene was down-regulated by the NLC-Citral and pure citral treatments. CCNB1 is known as Cyclin B1 gene that required for the activation of mitosis factor upon the binding of CDK-1 for mitotic initiation 48 . The study reported that down-regulation of Cyclin B1 promoted G2/M phase arrest in the leukemia cells treated with berberine 49 . Prystowsky et al. 50 studied that the inhibition of Plk1 and Cyclin B1 expressions results in panobinostat induced G2 delay and mitotic defects in the head and neck carcinoma cells. Additionally, CDKN1B gene was regulated in both NLC-Citral and pure citral treated cells. CDKN1B is a cyclin-dependent kinase 1 beta inhibitor and widely known as p27. P27 is a major key player in the cell cycle arrest that prevents the activation of Cyclin E or Cyclin D complexes in the cell cycle machinery 51 . In conclusion, these results have an agreement with previous studies showing that citral has an anti-cancer effect on MCF-7 and NB4 cells via the induction of cell cycle arrest at G2/M phase 52,53 .
Approximately, 90% of cell cancer deaths are related to metastasis, therefore the prevention of cancer cell distribution and secondary tumor formation is a major goal of cancer therapy 54 . Anti-metastasis activity of NLC-Citral on MDA MB-231 cells was elucidated through several assays including wound healing. This method is suitable to study cell-matrix interactions and cell migration activities 55 . The results demonstrated thatNLC-Citral has potent activities in inhibiting migration and invasion of breast cancer cells in vitro. Metastatic cascade can be divided into several crucial steps that include migration and invasion of the cancer cells into the circulatory system 56 . From mouse aorta ring assay, the number of microvessels was significantly decreased in NLC-Citral. The previous publication suggests that ascorbate has anti-angiogenesis effects through the decreased number of outgrowths of endothelial tubes from aorta ring treated groups 57 . Further, substantiate the metastasis effect of NLC-Citral on breast cancer cells, the metastasis-related genes expression analysis from microarray was also measured.
Angiogenesis is a critical step in tumor growth. Several chemotherapeutic agents were developed by targeting the angiogenesis machinery in breast cancer cells 58 . From the microarray analysis, GJA1 or also known as Connexin 43 (CX43) was the most highly up-regulated gene in the NLC-Citral treated cells. GJA1 gene is a component of gap junctions that allow the transfer of small molecules between adjacent cells. The study has reported that the regulation of gap junctions was pivotal for quantum dot cytotoxicity effects on human mesenchymal stem cells 59 . In addition, the study was revealed that down-regulation of CX43 gene inhibited an anti-angiogenic gene THRSB1 and increased the VEGF gene in the breast cancer cells 60 . Taken together, this result suggests that NLC-Citral and pure citral managed to induce the anti-angiogenic effects on MDA MB-231 cells. Silver nanoparticle was examined to increase the gap junction intercellular communication (GJIC) in the human lung adenocarcinoma cells, A549 through the significant increase of CX43 expression in the cells 61 . On the contrary, the result of the present study showed that Peroxidasin (PXDN) was the most down-regulated gene in MDA MB-231 treated cells. PXDN was on the list of most frequently mutated genes in the third stage pancreatic cancer patients through the prevalence screened 62 . It is an adhesion molecule involved in the extracellular matrix formation (ECM). PXDN was demonstrated as one of the highly regulated genes in plasma protein changes associated with tumor progression study in human breast cancer cells 63 . Besides that, the PXDN expression was relatively high in a few studies including breast, melanoma, ovarian as well as brain tumor samples and it is absent from the normal tissues [64][65][66] . In addition, microarray result also showed the down-regulation of SNAIL gene in the treated cells. SNAIL is crucial in the embryonic development and cancer progression. Over-expression of SNAIL enhances the metastasis activity of breast cancer cells through the increase of cell motility and invasiveness 67 . In the late stage of cancer, TGF-β has been studied to regulate the EMT by the induction of SNAIL expression in response to tumor progression 68 . It has also been reported that the over-expression of SNAIL in the prostate cancer cells is associated with the secretion of PXDN gene in the cell's extracellular matrix 69 . According to the microarray analysis, TGF-βI gene was down-regulated in the treated MDA MB-231 cells. Alteration in TGF-βI may induce the angiogenesis stimulation and immune system suppression by promoting the favorable environment for cancer cells development 70  induced cell migration through the lymphatic vessels 71 . On the other hand, the up-regulation of THSB1 gene was also observed. Studies have demonstrated that the up-regulation of THSB1 is an inhibitor of angiogenesis and tumor growth through the direct effects on cell migration and survival 72 . The result of gene expression analysis has supported the anti-proliferative and anti-metastasis effects that were previously measured through cytotoxicity assays and metastasis assays. Microarray technology allows us to perform high-throughput screening changes in gene expression level. In this study, targeted RNA sequencing assay (Truseq) and quantitative real-time PCR (qPCR) were employed to validate the gene expression from microarray data. Truseq is capable of examining a wide dynamic range of gene expression in a single experiment from a very small amount of sample, thus beneficial for the validation of gene expression level regulated in microarray 73 . Nevertheless, qPCR was done to double confirm the microarray and Truseq data as Truseq is a new method employed with so little references to refer. In fact, most of the studies were use qPCR to validate a few genes regulated in microarray 74 . All selected genes expressed similar gene expression patterns that were found in microarray analysis. Furthermore, to ascertain the potential of antitumor effect of NLC-Citral, the oral efficacy of NLC-Citral as compared to the citral in the inhibition of breast cancer tumor growth was tested in vivo with 28 days of treatment in 4T1 challenged mice. The size and weight of the tumors in the NLC-Citral treated mice were found to be substantially reduced compared to the NLC-Blank and citral groups reflecting that NLC-Citral is more effective compared to citral alone. A study has reported that the formulation of citral into biocompatible PEG micelles has significantly inhibited 4T1 tumor growth compared to the citral and control 75 . The results from this study provide critically important experimental evidence to suggest that NLC-Citral may be a novel efficacious and safe agent to inhibit the growth and metastasis of breast cancer.
In conclusion, this study suggests that NLC-Citral holds a promising cytotoxic activity targeting of breast cancer. Overall, the cytotoxic effect of NLC-Citral occurs via the induction of apoptosis and the effect is in a dose-dependent manner. Loading the citral into NLC system did not inhibit the in vitro cytotoxic, anti-proliferation and anti-metastatic effects of citral on breast cancer cells. However, NLC-Citral and citral have shown to regulate the different level of gene expression in MDA MB-231 treated cells. The current study shows that the NLC can be used as a carrier for delivery of citral in the breast cancer therapy for the first time. Hence, NLC-Citral can be developed into an ideal carrier for the delivery of citral in combating breast cancer in cancer therapeutics. Nonetheless, further in depth analysis will be done in our future study for in vivo experiments to test this hypothesis and also to investigate further roles of the NLC-Citral as the potential anticancer agent.

Materials and Methods
All methods and procedures used were carried out in accordance with relevant guidelines and regulations. Additionally, all experimental protocols used were approved by Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia. All data are compared to NLC-Blank in order to confirm that the anticancer effects of NLC-Citral are due to the citral and not to the components of NLC 10 . Preparation and characterization of NLC-Citral. Citral was loaded into NLC by high-pressure homogenization technique using hydrogenated palm oil (HPO), lipoid S-100 (Merck Millipore, Germany), olive oil (Basso, Italy), thimerosal, D-Sorbitol, Tween-80 and citral 95%. Characterization of NLC-Citral was determined using Transmission Electron Microscope (TEM), zeta potential (ZP), entrapment efficiency (EE), and drug loading capacity (DLC). TEM and zeta sizer analyses revealed that NLC-Citral was a nano-size particle with an average diameter of 54.12 ± 0.30 nm. Meanwhile, ZP, EE, and DLC of NLC-Citral were −12.73 ± 0.34 mV, 98.9 ± 0.124%, and 9.84 ± 0.041% respectively 76 . MTT assay. MTT assay was conducted in accordance with the previous study with slight modifications 78 .
Briefly, MDA MB-231, MCF-10A, and 4T1 cells were harvested, counted and seeded at 1 × 10 4 cells per well in the 96-well plate for 24 hours. The following day, cells were treated with various concentrations of the sample and incubated for 48 and 72 hours at 37 °C with 5% CO 2 . After that, 20 µL of 5 mg/mL MTT (Merck, USA) reagent was added to each well and incubated for 3 hours. Next, the solution was removed and 100 µL of dimethyl sulfoxide (DMSO) was added to the wells. Then, the plate was read at 575 nm by using the microtiter plate reader (µQuant, Bio-Tek Instrument). The results were analyzed as the percentage proliferation of thecells in respect to the concentration of the samples treated.
Annexin V fluorescence assay. The apoptosis effect of MDA MB-231 cells was studied using an Annexin V fluorescence Kit (BD Pharmingen, USA) according to manufacturer's instruction without any modifications. Briefly, cells were seeded for 24 hours at a density of 1.8 × 10 5 cells per well in 6-well plate. Then, the seeded cells were treated with the designated concentration (Table 1). At 48 hours post-treatment, the treated cells were harvested and collected as a pellet. Next, the pellet was resuspended in 400 µL binding buffer and stained with 5.0 µL of FITC-AnnexinV and 5.0 µL of PI provided in the kit. Afterward, the cells were analyzed by BD FACS Calibur (Becton Dickinson, USA). The result was analyzed using CellQuest 3.3 software. The experiment was performed in triplicates 79 .  (Table 1). After 48 hours post-treatment, cells were collected and the pellet was resuspended in 200 µL of trypsin (solution A) and RNase (solution B) for 10 minutes each prior to 250 µL of solution C (propidium iodide (PI)) for 15 minutes. After that, flow cytometric analysis was performed using BD FACS Calibur (Becton Dickinson, USA) within 3 hours. The result was analyzed using CellQuest 3.3 software. The experiment was performed in triplicates 80 .
JC-1 mitoscreen assay. JC-1 assay was used to detect the depolarization of the mitochondrial membrane potential of the cell after being treated with the selected doses of NLC-Citral for 48 hours using BD MitoScreen Kit (BD Pharmingen, USA). In brief, the cells were seeded in 6 wells plate for 24 hours at a density of 1.8 × 10 5 cells per well. Then, the cells were treated with various concentrations described in Table 1. This assay was performed based on the manufacturer's instruction provided with the kit without any modifications. After 48 hours of incubation, the cells were harvested, collected, and incubated with 500 µL of JC-1 stock solution for 15 minutes. . Cells were seeded in 6 wells plate at a density of 1.8 × 10 5 cells per well. The next day, cells were treated at desired concentrations as described in Table 1.
Then, cells were harvested at 48 hours post-treatment and the pellet was stained with 1.0 µL of RED-IETD-FMK for Caspase 8 and 1.0 µL of RED-LEHD-FMK for Caspase 9. The stained pellet was then incubated in the 37 °C incubator for 30 minutes. Finally, the cells were washed twice with PBS prior to being analyzed using Fluorometer (Thermo Fisher Scientific, USA). The procedure was followed according to the manufacturer's instructions without any modifications.
In vitro wound healing assay. This assay was performed according to the protocol defined in a previous study with slight modifications 82 . MDA MB-231 cells were plated to create 80% confluent monolayer overnight. After 24 hours incubation, the cell monolayer in a straight line scraped with a sterile p200 pipette tip to create scratches. Subsequently, the media was replaced with the treated media. Cells were then incubated for 24 hours. The image of wound closure was captured at 0, 12 and 24 hours of incubation time (Nikon, Japan). Ex vivo mouse aorta ring assay. The aorta ring assay provides a better understanding of angiogenesis process, in which it allows the analysis of developing microvessel branching and migration over the timescale 84 . Thoracic aorta was removed from the 7 weeks old BALB/c mice and immediately transferred to a petri-dish containing sterile PBS. The aorta ring was rinsed thrice before being cut into 1.0 mm long. The sectioned aorta was then transferred into 48-well plate pre-coated with 100 µL of matrigel (BD, USA). Next, another layer of matrigel was placed on top of the aorta subsequently. Once the gel was solidified, Opti-MEM media (Life Technologies, USA) with the desired concentration of NLC-Citral was added into the well. The plate was incubated at 37 °C with 5% CO 2 for 7 days and viewed under an inverted microscope (Nikon, Japan). The aorta was photographed and analyzed based on the number of protruding vessels from the aorta. This study was performed in accordance with relevant approval by the Animal Care and Use Committee (ACUC), Faculty of Veterinary Medicine, Universiti Putra Malaysia (UPM/IACUC/AUP-R098/2014). Data validation by targeted RNA sequencing (TruSeq). The targeted RNA sequencing for gene expression study was conducted to measure the expression level of selected genes of interest in MDA MB-231 cells after being treated with NLC-Citral at selected dose. The targeted RNA expression project was first designed using Design Studio software (Illumina, USA). In brief, 75 ng of total RNA was reversed transcribed into the cDNA. The ProtoScript II Reverse Transcriptase and RCS1 were mixed with the total RNA and placed on the pre-programmed thermal cycler for 30 minutes. After that, the hybridization of the oligo pool was done by adding the Targeted Oligo Pool (TOP) mixture to the tube containing cDNA. Next, OB1 was poured and placed on the pre-programmed thermal cycler to anneal. AM1 wash buffer was added to remove the unbound oligos and the supernatant was discarded using the magnetic stand. Then, UB1 was immediately added and the supernatant was removed. Subsequent, the ELM4 was poured and incubated at 37 °C for 45 minutes. For PCR amplification, diluted HP3 was added and incubated for 8 minutes at room temperature. Then, the mixture of TDP1 and PMM2 were spilled and placed on the magnetic stand until the liquid appears clear. Then placed on the thermal cycle for the PCR amplification process. Afterward, the AMPure XP Beads was added to the tube and dispersed completely before the supernatant was transferred into the new tube by using the magnetic stand. About 80% of ethanol was poured into the tube while standing on the magnetic stand without disturbing the beads. The supernatant was discarded, the tube air dried for about 15 minutes and the RSB was inserted before it was placed on the magnetic stand. Finally, the library pooling and quantitation were done in preparation for the sequencing process. Pooled libraries were run on a Miseq instrument (Illumina). All data were analyzed using the Miseq Reporter Targeter RNA application (BaseSpace).

Gene Expression Profiling
Data validation by quantitative real-time polymerase chain reaction (qPCR). Total RNA extracted from the treated cells was converted to cDNA using Maxima First Strand cDNA synthesis kit (Thermo Scientific, USA) and was run into a thermal cycler (Labnet, USA). Then, the qPCR reaction was performed using SYBR Select Master Mix (Life Technologies, USA) in the Eco Illumina (Illumina, USA) to quantify the differential expression of the selected genes. The PCR parameters were, 95 °C for 10 minutes, 40 cycles of 95 °C for DNA denaturation, and 55 °C for 30 seconds. The forward and reverse sequences of the target and housekeeping genes (GAPDH and ACTB) were obtained from Primer-Blast NCBI ( Table 3). The quantity of the genes was calculated and measured by Eco study software (Illumina, USA). The expression levels of NLC-Citral and citral were compared to the NLC-Blank control group.  Proteome profiling of apoptosis-related proteins. The expression profile of apoptosis-related proteins was investigated on the MDA MB-231 cells treated with NLC-Citral for 48 hours using the Human Apoptosis Array Kit (R & D Systems, USA). Protein from the cells was extracted using 600 µL of RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1.0% TritonX-100, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with 10 mg of protease inhibitor cocktails (Roche, Canada). The protein was then quantified using Bradford reagent (Sigma, USA). Concisely, the membrane was incubated for 1 hour in the 2.0 mL array buffer 1 and the protein sample was prepared with the addition of 1.0 mL lysis buffer 17 to the protein lysate. Next, prepared sample was loaded into the membrane and incubated at 4 °C overnight. Then, the membrane was washed 3 times with 20 mL wash buffer, transferred into the 4-well multi-dish containing detection antibody cocktails and incubated for 1 hour. Two mL diluted Streptavidin-HRP with array buffer was poured, incubated for 30 minutes,and washed 3 times before 1.0 mL of Chemi Reagent Mix was pipetted onto the membrane for viewing. The membrane was scanned using the ChemiDoc XRS (Bio-Rad, USA).
Data validation by Enzyme-Linked Immunosorbent Assay (ELISA). The quantified serum proteins including survivin, Cytochrome C, Bax and Bcl-2 concentrations were determined in triplicates and validated using the quantitative human colorimetric ELISA kit (Abcam, USA). Then, relative expression was calculated by dividing sample against NLC-Blank and a standard curve was generated for each protein separately. Animals grouping and treatment. The mice were grouped into 3 groups (n = 6 for each group): (1) NLC-Blank group, (2) NLC-Citral treated group, and (3) Citral treated group. Mice in all 3 groups were inoculated with 1 × 10 5 4T1 cells in RPMI-1640 media via subcutaneous injection. All treated mice were fed orally with NLC-Citral and citral with 50 mg/kg daily using oral gavage for 28 days starting on the first day of inoculation.

Animals.
The dose was selected based on the studies conducted previously 85 . Whereas, the negative control group was treated with the vehicle used to encapsulate the citral (NLC-Blank).
Measuring the tumor weight and size. The 4T1 tumor harvested from the mice after 28 days of treatment with NLC-Blank, NLC-Citral, and citral were weighted using a weighing balance (Mettler Toledo, Switzerland) while the tumor size were measured using a caliper and the tumour volumes were calculated using the formula V = 1/2 (width2 × length).

Statistical analysis.
All experiments were done in triplicates and the average values were obtained. The statistical analysis was performed using the GraphPad Prism 6.0. One way ANOVA was selected for the experimental analysis with Tukey's post hoc test. The significance was set at p < 0.05 when comparing to NLC-Blank as a control and NLC-Citral and citral as treated groups.