Garlic (Allium sativum)-derived SEVs inhibit cancer cell proliferation and induce caspase mediated apoptosis

As a key component of the cell-to-cell communication, small extracellular vesicles (SEVs) released from various sources are known to be affecting the physiological conditions of the target cells. Although it has been suggested that edible plant-derived nanoparticles contributes to the cross kingdom communication with the mammalian cells, the effect of these particles on cancer cell progression still needs a further exploration. Here, we isolated and then characterized garlic derived SEVs by nanoparticle tracking analysis, electron microscopy and SEV surface antibodies. In order to investigate anti-cancer property of garlic SEVs A498 human kidney carcinoma, A549 human lung carcinoma were used as cell models along with the normal human dermal fibroblast cell lines. Annexin V/pI staining and analysis of apoptotic mRNA and protein expression levels suggested that garlic SEVs induced apoptosis through activation of intrinsic pathway. Furthermore, angiogenic VEGF protein expression levels significantly decreased in response to SEVs treatment in cancer cells. Our results support that garlic derived SEVs could cause apoptotic cell death among cancer cells while normal cells remain unaffected with the treatment. This study revealed for the first time that plant SEVs possess anti-cancer affects by inducing caspase mediated apoptosis and provided a new alternative for cancer treatment.

SEVs characterization via nanoparticle tracking analysis and zetasizer. Size distribution and particle number of the isolated SEVs was determined by Nanosight and zetasizer (Malvern Instruments Ltd, Malvern, Worcestershire, UK). SEVs samples were diluted in 1 ml distilled water and parameters were analyzed at room temperature. All samples were evaluated in at least 4 replicates. Blue laser beam at 488 nm was applied to the SEVs solution and a video was recorded with frame rate of 25 frames/s. The video was analyzed by NTA software version 3.3, Nanosight. Each SEVs replicate was analyzed at least 5 times, and their Brownian motion as well as their particle movement velocity were used to analyze the particle size of SEVs by using two-dimensional Stokes-Einstein equation. Zetasizer Nano ZS (Malvern Instruments) was also used to evaluate particle size of the isolated SEVs.
Cell viability assay. Cells were seeded with a concentration of of 5 × 10 3 cells/well in a 96 well plate and treated with escalating doses of A. sativum nanovesicles and garlic juice (5-50 μg/ml) for 24,48 or 72 h in a humidified incubator (5% CO 2 in air at 37 °C). The cell viability was performed with WST-1 assay (BioVision, CA) as described previously 20 . The absorbance was evaluated with the spectrophotometer at 540 nm. Three independent experiments were performed to generate the percentage of growth versus untreated control cells.
Annexin V assay. The apoptotic effect of garlic SEVs on cancer cells was detected with Annexin V Detection Kit (BioLegend, USA) according to the manufacturers' protocol. A498, A549 and HDF cells were seeded onto 6-well plates with the cell density of 1 × 10 5 and treated with 50 μg/ml of A. Sativum SEVs for 24 and 48 h. Following the incubation process, cells harvested were incubated with Annexin V FITC and PI PE conjugated antibody and Apoptosis and necrosis were evaluated using flow cytometer. Obtained data was then analyzed using CellQuest Pro program. Total  Enzyme-linked immunosorbent assay. Cells seeded at 2 × 10 5 cells/well into 6-well plates were treated with 50 μg/ml SEVs for 48 h. Following cell lysis in RIPA buffer containing 1% protease inhibitor cocktail for 5 min on ice, lysates were centrifuged at 14 000×g for 15 min at 4 °C. The concentration of proteins were measured with Lowry Assay as described in the manufacturer's instructions. (Bio-Rad USA). 50 µg protein from each sample (in 100 µl) is transferred to the precoated ELISA plates and incubated at + 4 °C overnight. The total amount of cleaved caspase 3 (Ray Biotech, USA) and Bcl-2 (Thermo Fisher Scientific, USA) protein expressions were quantified and normalized as micrograms of Cas3 or Bcl-2 per millilitre of total protein content using ELISA kits as described previously in the manufacturer's guide. For VEGF Elisa, medium of control cells and treated cells were collected and centrifuged at 10,000 rpm for 30 s in order to remove any impurities remained in the media and then the amount of VEGF secretion was quantified and normalized as micrograms per milligrams of total protein content using ELISA kit according to manufacturer's guide (Thermo Fisher Scientific, USA).

RNA extraction and real-time PCR.
Western blotting. The caspase 3 protein expression in garlic SEV-treated cells was confirmed using the Wes Simple Western apparatus (Protein Simple, San Jose, USA). Denatured 0.4-0.6 g protein was loaded onto Wes multiwell plates after being denatured in a DTT-based buffer supplied by the manufacturer. Protein isolation, antibody binding, and labeling were all accomplished using capillary cartridge separation.

Effect of garlic SEVs on the cell survival of A498, A549 and HDF cells.
The effect of various concentrations (5-50 μg/ml) of garlic SEVs on A498, A549 and HDF cell viability were determined at 24, 48 and 72 h (Fig. 2). Concentrations up to 50 μg/ml of garlic SEVs exerted significant cytotoxic effects on both cancer cell lines where the viability of A498 and A549 cells decreased down to 35% within 24 h. The cytotoxic effect of garlic SEVs increased in the following 48 h and 72 h by 28% and 22%, respectively ( Fig. 2A,B). Interestingly, increasing doses of the garlic SEVs showed no significant cytotoxic effect on the viability of normal HDF cells. Contrarily, the proliferation of HDF cells was significantly increased up to 188% when exposed to garlic SEVs at low doses (Fig. 2C). Moreover, to check the effect of garlic juice on these cancer cells, garlic juice was used in treatment of these cells on similar doses.

Treatment of garlic SEVs caused S phase cell cycle arrest in A498 and A549 cells. The effect of
garlic SEVs on cell cycle progression was also investigated on A498, A549 and HDF cells via flow cytometry in order to clarify the mechanism contributing the growth inhibition. 24 and 48 h treatment with garlic SEVs led to accumulation of both cancer cells in S phase (Fig. 3). The number of A498 cells in G0/G1 phase significantly decreased from 70 to 57% in the first 24 h. Consistently, 24 h after the treatment with garlic SEVs, A498 cells displayed an increase in the S phase from 14 to 22%. Although the treatment led to a slight increase in the G2/M phase (from 15 to 19%), at 48 h S phase accumulation reached to 25% whereas the accumulation in the G2/M phase drop down to 7% among A498 cells (Fig. 3A). Similarly, a respective decrease of 15% and 17% in the number of G0/G1 phase A549 cells were observed when treated with garlic SEVs for 24 and 48 h. An increase in S phase cell population from 12 to 26% for 24 h and to 27% for 48 h was recorded for A549 cells in the presence Apoptotic effect of garlic SEVs on A498, A549 and HDF cells for 24 and 48 h. In order to further evaluate the cell cycle arrest and cell death mechanism, Annexin V-FITC/pI assays was performed to confirm the apoptotic effect of garlic SEVs on A498, A549 cells. 50 µg/ml of garlic SEVs treatment on A498 cells showed that apoptotic cell death was nearly 7% on 24 h which was increased to 24% after 48 h treatment. No significant necrosis was observed in neither 24 nor 48 h treatment (Fig. 4A). A higher rate of apoptosis with 16% was evident in A549 cells when treated with 50 μg/ml of garlic SEVs at 24 h. The apoptotic cell death ratio was increased to 21% in A549 cells at the end of 48 h due to SEVs treatment (Fig. 4B). On the other hand, no significant induction of apoptosis was observed for normal HDF cells with the apoptotic cell ratio remaining at 4% and 5% for the 24 and 48 h treatments in response to garlic SEVs (Fig. 4C). Increased ROS levels and mitochondrial dysfunction, which are seen in many cancers, cause oxidative damage to biological systems, particularly in genomic and mitochondrial DNA, resulting in oncogene activation and inactivation of tumor suppressor genes, which contributes to cellular transformation. ROS levels were evaluated by the DCFDA / H2DCFDA-Cellular ROS Assay. Results showed that garlic SEVs treatment resulted in an increase in ROS levels.

Effect of garlic SEVs on pro-/anti-apoptotic genes and protein expression levels. Next, we
examined the effect of 50 µg/ml garlic SEVs on pro-and anti-apoptotic gene expression levels using RT-PCR as well as their respective protein expressions levels by ELISA. Anti-apoptotic Bcl-2 gene expression level was significantly decreased down nearly to 60% in A498 cells and 65% in A549 cells. Although Bcl-2 gene expression level was increased up to 150% in normal HDF cells, this change was found to be statistically not significant. The change in Bcl-2 gene expression levels was confirmed with ELISA, showing that anti-apoptotic Bcl-2 protein levels were decreased nearly 40% in both A498 and A549 cells along with a significant increase up to 20% in normal HDF cells (Fig. 5A). Consistent with decreased anti-apoptotic gene and protein expression levels, there was a significant threefold increase in pro-apoptotic Cas3 gene expression levels in both cancer cells while sevenfold decrease was detected for Cas3 mRNA levels in normal HDF cells upon treatment. Caspase-3 protein expression levels have also risen up by 20% in cancer cell lines and reduced by 10% in HDF cells (Fig. 5B). In parallel   www.nature.com/scientificreports/ respective 1.5-, 3-and 2.5-fold increase in Bax, P53 and Cas9 genes was seen in response to the SEVs treatment. Albeit to these notable changes in cancer cells, normal HDF cells exerted no significant change in both Bax and P53 gene expression levels, while Cas9 gene expression levels was substantially decreased by sevenfold in HDF cells (Fig. 5C). Lastly, the Cas3 protein expression levels were evaluated by western blotting. Results demonstrated that garlic SEVs treatment lead to increase Cas 3 protein expression in A498 and A549 cells.

Role of garlic SEVs on vasculogenesis.
In order to determine the effect of garlic SEVs on vascularization, we evaluated the soluble VEGF protein expression levels on each cell line. The levels of VEGF in cell media was found the highest in A498 cells, followed by A549 and HDF cells (Fig. 6). VEGF secretion of A498 cells significantly decreased approximately to 1000 pg/ml levels from 1500 pg/ml. Similarly amount of VEGF released by A549 cells significantly decreased by 1.5-fold to 500 pg/ml levels in response to garlic SEVs. Although no significant change was observed in VEGF release after the treatment with garlic SEVs of HDF cell line, there was a slight increase up to 200 pg/ml in VEGF levels. (Fig. 6).

Discussion
Comprehensive research for new therapeutic drugs for cancer have been an ongoing research topic during the last decades. Although several molecular compounds were studied distinctively both in vitro and in vivo for their chemotherapeutic and chemoprotective effects for various cancer types such as prostate, colon, breast and stomach carcinogenesis, the search for newer and more effective drugs is never-ending [21][22][23][24] . In this study, we characterized SEVs-like nanoparticles from garlic and evaluated their effect on both cancer and normal cells. The use of natural extracts in cancer prevention and treatment has been an increasing trend as the numerous data in the traditional medicine suggest that these natural extracts possess anti-proliferative activities on different cancer types [25][26][27] . A. sativum had been used for external tumor treatment around 1550 B.C. by ancient Egyptians 28,29 . In order to determine whether garlic SEVs possess an anti-cancer effect, we chose 2 cancer cell  30,31 . Extensive size distribution (20-300 nm in size), protein, lipid and RNA contents from plants show similarities with mammalian SEVs. Recent literature suggested that the pre-observed effects of these natural extracts could be due to SEVs found in extract constituents 32 . Herein, we examined exosomal tetraspanin family marker proteins CD9, CD63 and HSP70 via flow cytometry from SEVs derived from garlic juice 8,12,31,[33][34][35][36] . We also analyzed the size distribution and size intensity of the SEVs as well as scanning electron microscopy images of SEVs [37][38][39] .
Despite the lack of sufficient knowledge of interspecies communications of SEVs-like nanoparticles from plants, Chang et al. reported that anti-oxidant gene expression levels increased in intestinal macrophages after treatment with grape derived extracellular vesicles 12 .
Hence, we evaluated if the SEVs derived from garlic possess any effect in a cross-kingdom communication and prevent cancer cell progression through consecutive in vitro studies such as MTS assay, cell cycle analysis, Figure 5. Quantitative RT-PCR analysis of Bcl-2, (A), Cas3 (B) Bax, P53, and Cas9 (C) expression levels of human kidney carcinoma A-498, human lung carcinoma A-549 and human dermal fibroblast (HDF) cells. Expression of 18S was used as a negative control and gene expression levels were normalized to the housekeeping gene 18S. Cells were plated in 6-wells and treated with 50 µg/ml garlic SEVs. After 48 h of treatment, RNA isolation and cDNA conversion is performed as described in the "Materials and methods" section. Data are expressed as the mean ± SD. Results are presented at n = 4 for each group one-way ANOVA multiple comparisons, **p < 0.01, ***p < 0.001, ****p > 0.0001, ns not significant. www.nature.com/scientificreports/ Annexin V/pI analysis, mRNA and protein expression levels. While the SEVs treatment significantly reduced the viability of cancer cells, it did not cause a cytotoxic response on the normal cell line HDF. One of the reasons that for this observed decrease in cell proliferation among cancer cells in response to SEVs treatment was the cell cycle arrest in the S phase with a non-significant changes in G2/M phase. Evidently, the treatment caused no significant cell cycle arrest on HDF which explains lack of anti-proliferative effect in response to the SEVs treatment.
In order to examine the relationship between cell cycle arrest and the mode of cell death mechanism, annexin-V/pI staining assay was performed for each cell line upon garlic SEVs treatment. treated cancer cells treated with garlic SEVs exhibited high apoptotic activities at all time points. Our results showed that, early and late apoptotic cell percentage has increased in a time dependent manner in both cancer cell lines whereas apoptotic cell number in normal cells remained not significant with respect to the control cells. Furthermore, pro-and anti-apoptotic pathways in each cell line were evaluated to demonstrate that these nanovesicles are in fact inhibiting cancer cell proliferation via certain pathways. Our results suggest that, treatment of kidney and lung carcinoma cells with garlic SEVs led to a significant increase in gene expression levels of pro-apoptotic genes such as p53, Bax, Cas3 and Cas9 while a statistically significant decrease in anti-apoptotic gene expression levels of Bcl-2 was detected in each cancer cell lines. ELISA results for Bcl-2 levels were found to be in line with the observed changes for Bcl-2 gene expression for A498 & A549 in response to SEVs treatment. Similarly, parallel results were obtained for Cas3 activation as increase in cleaved Cas3 levels was detected for SEVs treated A498 & A549. On the other hand, the change in the level of Bcl-2, p53 and Bax gene expression and Bcl-2 protein levels remained statistically not significant in HDFs upon treatment but a significant reduction in both pro-apoptotic Cas9 and Cas3gene expression as well as cleaved Cas3 was evident. Activated mainly by Cas9, Cas3 is one of the most vital executioner of the cellular apoptosis 40 , hence this finding suggest that the effect garlic SEVs was not limited to abrogated cancer cell proliferation but also to increased cell survival in normal cells. As a potent angiogenic factor, it is well established that the increase in VEGF expression contributes to the tumor angiogenesis in cancer cells as well as protecting these cells from apoptosis 41,42 . In response to treatment garlic SEVs, A498 & A549 displayed a significant decrease in VEGF secretion though no significant change was recorded for in normal human dermal fibroblast cells after the treatment with garlic SEVs.
In summary, data presented here strongly suggest that SEVs derived from garlic display therapeutic effects on cancer cells whilst they exert minimum to none cytotoxic effect to normal HDFs. On the contrary our results showed that these nanovesicles in fact increase the viability by increasing the anti-apoptotic Bcl-2 expression levels and decreasing the basal Cas3 activity. The ability of cross kingdom communication of these SEVs-like nanoparticles from garlic could be useful to develop new anticancer therapeutics with minimum side effects. Figure 6. ELISA analysis of human kidney carcinoma A-498, human lung carcinoma A-549 and human dermal fibroblast HDF cells after the 48 h treatment of 50 µg/ml garlic SEVs. Cells were plated in 6-wells and treated with 50 µg/ml garlic exosome. Proteins of treated cells were harvested as described in the "Materials and methods" section. Data are expressed as the mean ± SD. Results are presented at n = 5 for each group one-way ANOVA multiple comparisons, **p < 0.01, ***p < 0.0001, ns not significant. www.nature.com/scientificreports/