Annona muricata Graviola Induces Apoptosis in Two Osteosarcoma Cell Lines and Downregulates the Cytokines IL-6 and TGFβ1 Which Are Implicated in Tumour Growth and Metastasis

Abstract

Background

Osteosarcoma (OS) is an aggressive bone tumour which affects mostly young children. Despite advances in chemotherapy regimens there is still high fatality and cure rates remain low. Annona muricata Graviola (GR) is a tropical fruit bearing tree whose leaves, stems and fruits have indigenous medicinal properties. Studies have shown that GR has anti-tumour effects on breast, liver and prostate tumours.

Objectives

The aim of this study was to investigate the effect of GR on bone cancer cell lines of the OS lineage.

Methods

Two human OS cell lines; HOS and MG63 were cultured with GR (300 mg/mL) for 24 h at 37°C. After which supernatants and cell pellets were collected and analyzed for apoptosis by microscopy and flow cytometry, using annexin staining. Culture supernatants were analyzed for IL-6 and transforming growth beta (TGFβ) using ELISAs. Label free proteomics was used to evaluate changes in protein expression. We also investigated the effect of combining blocking antibodies to p53 and BCL-2 with GR treatment on the cell lines. The effect on TGFβ expression was then measured using flow cytometry.

Results

Treatment of HOS and MG63 cells with GR increased the expression of annexin compared to untreated cells. GR treatment also caused a dysregulation in the secretion of the cytokines IL-6 and TGFβ. Proteomics showed that GR induced apoptosis in OS cells through multiple pathways triggering an alteration in the expression of key proteins involved in cellular respiration, cell cycle, motility, DNA synthesis and cell death. Further, GR was shown to downregulate TGFβ through BCL-2 and p53 mediated pathways.

Conclusion

The data suggest GR has anti-tumour effects on OS cell lines therefore the efficacy of GR should be tested clinically in OS patients.

Introduction

Osteosarcoma (OS) is the most common primary bone tumour in children and adolescents. It has an incidence of three million cases per year and accounts for nearly five million deaths worldwide. ¹ OS occurs in growing bone tissue and consists of immature undifferentiated osteoclast cells. OS can become rapidly metastatic with tumours spreading to the lungs, lymph nodes and other bones. ² Treatments involve mainly invasive often crippling surgery alongside chemotherapy but unfortunately despite recent treatment advances, the survival rate has not increased. Currently the cure rate is around 60% but drops to only 20% for patients diagnosed with metastasis. ³ The incidence of recurrent disease and chemotherapy resistance ⁴ is also high. One of the reasons OS is difficult to contain is that the tumour affects the rapidly growing bone ⁵ cellular environment. Furthermore, OS tumour cells are capable of forming a layer of osteoid tumour matrix which can secrete multiple cytokines tumour necrosis factor-alpha (TNFα), IL-11, IL-6, and receptor activator of nuclear factor kappa (RANK). ⁶ This can also activate unnecessary healthy bone degradation and contribute to an inflammatory microenvironment. ⁷ The role of the surrounding cells and factors around tumours is becoming increasingly important in predicting the outcome of cancer. ⁸ Numerous studies have shown that cell growth promoters such as transforming growth factor-β (TGFβ), ⁹ insulin growth factor-1 (IGF-1) and IL-6⁹ when present around tumours can enhance their growth, spread and speed up angiogenesis ¹⁰ . Therefore, early diagnosis and new treatments which can target angiogenic factors and tumour spread are essential to contain and combat this aggressive bone tumour. Cancer treatment often involves therapy which targets tumour apoptosis or programmed cell death in which cells are called to self-destruct when stimulated by an appropriate trigger. ¹¹ Graviola (Annona muricata L.) is a species of plant belonging to the Annonaceae family and is often referred to as soursop, custard apple and guanabana. The evergreen GR fruit tree is native to South America and the Caribbean rainforest. It is commonly used to prepare teas, syrups, ice creams, drinks and sweets. ¹² Traditionally, multiple parts of GR including the seeds, leaves, bark, fruit-pulp, roots and stems have been used to treat a vast range of maladies such as diabetes, ¹³ arthritis, ¹⁴ bacterial or fungal infections ¹⁵ and tumours ¹⁶ . GR has many ethnomedicinal uses including anticancer properties that have been reported on breast, ¹⁷ prostate, ¹⁸ pancreatic ¹⁹ and colorectal cancers ²⁰ cell lines causing loss of clonogenicity and cell death Interestingly a study showed that mice with Ehrlic solid tumours reported a significant decrease in their tumour size after 30 days of oral GR treatment. It was found that GR interfered with the tumour cell cycle and initiated cell apoptosis. ²¹ However, the effect of GR on osteosarcoma cells has never been explored and there have been few advances in terms of new treatments for OS. Therefore in this study we investigated the effect of GR on OS cell lines and aimed to identify the underlying mechanisms involved.

Methods

Consumables: Cells: HOS (RRID:CRL-1543, STR 15), MG63 (RRID:CRL-1427, STR 15) and U937 (RRID:CRL-3253, STR-8) were all purchased from the American type tissue collection (ATCC) 1081 University Blvd, Virginia, USA. Annexin 5 FITC labelling kit (ab14085) was purchased from Stockport, Abcam, UK Fetal calf serum (ES-009B), dithiothreitol (DTT, D0632), iodoacetamide (IAA, CAS144489) and glass beads (#G4649-100G) were purchased from Merck life science Ltd, Dorset, UK. IL-6 (#EH2IL6), TGFβ₁ ELISA kits (#BMS249-4) anti-TGFβ (#TB21), cell media DMEM (#12491015), penicillin streptomycin (#15240062) and Rabbit anti- mouse IgG FITC antibody (#A18904) were purchased from Thermofisher Scientific, Milton Keynes,UK. L-glutamine (gluta-max #A1286001) was purchased from Gibco, UK. Anti-BCL-2 (#Ib5b9), anti-p53 (#21891-1-AP) and rabbit isotype IgG (#30000-0-AP) were purchased from Proteintech, Manchester, UK.

Cell Culture

HOS and MG63 cells were cultured in 24 well bottom plates at a concentration of 4 × 10⁵ per mL and incubated at 37°C with 5% carbon dioxide. GR was tested at concentrations of 4, 8,16, 32, 75, 150 and 300 mg/mL with both OS cell line cultures. For all control experiments, the HOS and MG63 cells were cultured 4 × 10⁵ per mL in media containing no stimulants and incubated at 37°C with 5% carbon dioxide for 24 h. Cells were incubated for period of 24 h in co-culture with GR. Following incubation, cultures were, centrifuged at 1200g for 10 min and supernatant and cell pellet were collected and stored at −70^°C prior to analysis. Initial experiments involved collecting the cell supernatants at 2, 4, 6, 12, 24, and 48 h after the addition of GR. This established that the optimal time for cell death and maximum cytokine release occurred at 24 h and with a GR concentration of 150 mg/mL. All further experiments were carried out using these optimal conditions.

Human Mononuclear Cell Culture

The human monocyte cell line U937 was used as a control cell line to compare the effects of GR. The U937cells were grown to confluence and seeded at a density of 4 × 10⁵/mL. The cells were cultured either with GR (150 mg/mL) or in media without any stimulation (control) for 24 h at 37°C with 5% CO₂. After which supernatants and cell pellets were collected for analysis of cell viability and assays.

Graviola Preparation

Graviola (Annona muricata, 600 mg, 100 capsules) was purchased from Raintree, Europe ( www.raintree.com) and consisted of the leaves and stem of the plant. For the experiment, the capsule contents were emptied and dissolved in warm cell culture medium without supplements to prepare a stock solution of GR at level of 300 mg/mL. The solution was cooled to room temperature and double filter sterilized using a 0.4 µm filter to remove particulates and microbes prior to addition to cells. Any required dilutions were prepared in culture medium without supplements. For each experiment GR was freshly prepared using the above methodology for all experimental testing throughout the study.

Cell Culture Preparations for Bottom-Up Proteomics

OS cells lines: HOS and MG63 were incubated in media with and without GR (150 mg/mL) for 24 h at 37^°C (n = 6 in both groups). Cells were then washed with DMEM followed by three washes with cold PBS, and placed on ice. A small amount of glass beads (200 mg) was added to each sample together with 200 µL of 50 mM ammonium bicarbonate buffer containing 8 M urea. The samples were then placed at −80ºC for 10 min. Samples were thawed on ice and subsequently cells were lysed using a Fastprep® 24 homogeniser (MP Biomedicals, Eschwege Germany, (#116004500) at 6000 rpm for 45 s. This step was repeated two more times with samples being placed on ice for one min between each cycle. Then samples were centrifuged at 12 000g for 15 min. Supernatants were transferred to a clean tube and protein concentration was measured using a NanoDrop^TM spectrophotometer (ThermoFisher Scientific, UK) set at a wavelength of 280 nm.

Protein Sample Preparation and LC-MS/MS Analysis

An aliquot of protein extract containing 100 µg of protein was diluted with 50 mM ammonium bicarbonate to a final volume of 200 µL. The samples were reduced with 10 µL of 95 mM dithiothreitol (DTT, D0632) for 1 h at 56ºC and subsequently alkylated by incubating with 10 µL of 300 mM iodoacetamide (IAA, #I1149) at room temperature in the dark. The sample volume was adjusted to 1.0 mL with ammonium bicarbonate. Proteins were digested by incubating the samples with trypsin at a ratio of 1:50 enzyme/protein at 37°C overnight. The tryptic digests were vacuum-centrifuged to dryness using a SpeedVac vacuum centrifuge (ThermoFisher, Hemel Hempstead, UK). The samples were then reconstituted in 100 µL of 50 mM ammonium bicarbonate. Tryptic digests were subsequently analyzed using a Dionex Ultimate 3000 RSLC Nano ultra-high-performance liquid chromatography system coupled to a Q Exactive (ThermoFisher Scientific, Loughborough, UK). Aliquots (15 µL) were desalted and concentrated using a 5 mm × 300 μm i.d. C18 trap cartridge (ThermoFisher Scientific, Hemel Hempstead, UK). Solvents consisted of a mixture of water and acetonitrile (98:2%, v/v) containing 0.05 % TFA (loading solvent A) and a mixture of acetonitrile and water (80:20%, v/v) (loading solvent B). A flow rate was 20 μL·min⁻¹ was used. Concentrated sample was separated using a 150 × 300 µm C18 Pepmap column (ThermoFisher Scientific, Hemel Hempstead, UK) with a binary gradient elution profile composed of a mixture of water and acetonitrile (95:5 %, v/v) containing 0.1 % formic acid (eluent A) and a mixture of acetonitrile and water (80:20 %, v/v) (eluent B) containing 0.1 % formic acid at a flow rate of 6 μL·min⁻¹. Overall acquisition time was 60 min. The autosampler and column oven temperature was set to 4 and 40°C, respectively. The Q Exactive was operated in a data dependent mode. MS survey scans were acquired from m/z 350 to 2000 at resolution of 70 000 with AGC of 3e6 and maximum IT of 100 ms. The 10 most abundant ions were subjected to MS/MS and measured with a resolution of 17 500 and AGC of 1e5 and maximum IT of 200 ms.

Protein Identification and Quantitation

The raw MS data files were analysed using Proteome Discoverer® v2.5 (ThermoFisher Scientific, Hemel Hempstead, UK). Database matches were made against Human downloaded from National Centre for Biotechnology Information (NCBI) using the Sequest search engine. For identification of proteins the following search parameters were used. Precursor mass tolerance of 10 ppm and a fragment mass tolerance of 0.02 Da, enzyme-trypsin, static modification-carbamidomethyl (CAM: + 57.021 Da), variable modification-methionine oxidation (MetOx: + 15.995 Da) and acetylation of N-terminus (Acetyl: +42 Dal) and maximum missed cleavages of two. For quantification of proteins, protein abundances were normalised to the same total peptide amount per channel and scaled as defaults, so that the average abundance per protein and peptide is 100. To ensure the robustness of the results, six biological replicates were performed with precursor quantification set to 66% (proteins detected in at least four out of six replicates). Label-free quantification (LFQ) was conducted with a stringent target false discovery rate (FDR) of 0.01 for both proteins and peptides, ensuring high confidence in the data. Proteins showing a fold change greater than 2.0 or less than 0.5, along with a percentage coefficient of variation for group abundance of ≤20%, were selected for further analysis, ensuring that only consistently regulated proteins were considered. Proteins were deemed exclusive to one condition if they were consistently detected in at least four out of six replicates in one condition and completely absent in the other. This approach, combined with rigorous statistical thresholds and replicate consistency, validates the reliability of the identified condition-specific proteins.

Bioinformatics

Functional analysis was conducted using databases such as Gene Ontology (GO) (http://www.geneontology.org/), PANTHER version 16.0, and DAVID (http://david.abcc.ncifcrf.gov/) to explore the involvement of identified proteins in biological processes, molecular functions, protein categories, and cellular components. Additionally, up- and down-regulated proteins were analyzed using STRING (http://string-db.org/). This analysis provided a deeper understanding of the identified proteins, their biological context, and their involvement in various pathways affected in cell lines following GR treatment.

Apoptosis Detection using Annexin with Flow Cytometry

HOS and MG63 cells were either co-cultured with GR or without GR for 24h at 37°C, 5% CO₂. The cells were then collected and washed with PBS by gentle pipetting. 200 µL pre-diluted binding buffer (50 mL binding buffer mixed with 150 mL of distilled water, 1:4, v/v) was used to resuspend the cells, adjusting to a cell density of 2-5 × 10⁵ cells/mL. The cell suspensions (195 µL) was incubated with 5 µL of Annexin V: FITC in the dark for 10 min at room temperature. Cells were washed with 200 µl of pre-diluted binding buffer and subsequently resuspend in 190 µL of the same buffer. A small volume (10 µL) of the propidium iodide solution was added to the cells, which were then analyzed using flow cytometry.

The flow cytometer was set so that the mean fluorescence intensity of the annexin V negative population was between 1 and 10. For the positive control, OS cells were incubated with 3% formaldehyde in buffer for 30 min on ice after which cells were washed and resuspended in cold binding buffer at 2-5 × 10⁵ cells/mL. Annexin labelling was then carried out on control cells, as described above.

Blocking antibodies against BCL-2 and p53 as well as the isotype matched controls, were added to both OS cell lines (2-5 × 10⁵ cells/mL) prior to treatment with GR at 300 mg/mL dilution. The cultures were incubated at 37^°C for 24h. Following incubation, the cells were detached, labelled with anti-TGFβ and the percentage expression was determined through flow cytometry.

Visual Microscopic Analysis of Cell Viability

After 24h incubation at 37^°C with 5% CO₂ and GR at various concentrations, Trypan blue was added to cells. The cells were then counted and analyzed using a light microscope. Images were captured using a light microscope with an optical camera visualizer with a 14% compensation.

Statistical Analysis

All results are representative of at least three to six repeats of independent experiments [specific replicate numbers (n) have been shown in figure legends] with results expressed as the duplicate means ± SD of the independent repeat of each experiment. Statistical analysis was carried out using Excel tools and two-way analysis of variance (ANOVA).

Results

The primary effect of GR on OS and monocyte cells lines was examined after 24 h using a light microscope to show any changes in cell structure, shape and viability.

Figure 1 shows light microscopic images depicting confluent cells in media and compares images of the effects of GR on the OS cells after 24 h. GR was able to induce cell death in HOS and MG63 cell lines following co-culture with GR after 24 h. This was clearly visible under light microscopy which revealed cells were visibly clumping and detaching from the culture flask in aggregates. After staining the cells with trypan blue, the percentage of cell death was evaluated under a light microscope. Treatment with 300, 150 and 75 mg/mL of GR resulted in 86 ± 2%, 76.3 ± 1.5% and 27.7 ± 4% cell death in HOS cells respectively. In MG63 cells the corresponding cell death percentages for 300, 150 and 75 mg/mL were 83 ± 2%, 74.3 ± 2.0%, 31 ± 1% (Figure 2d and e). In comparison, untreated control cells showed less than 5.0% positive staining (Figure 2a and b). The images shown are for results at 150 mg/mL of GR treatment on all cells.

Figure 1.

Light microscopic images of human osteosarcoma and monocyte cell viability after co-culture with Graviola (GR). 4 × 10⁵/mL of cells were incubated at 37^°C for 24 h in media or GR at 300 mg/mL for all experiments. (a) HOS in media (b) MG63 in media; c) U937 monocyte cells in media (d) U937 monocytes with GR; (e) HOS with GR (f) MG63 with GR. All cells were stained with trypan blue before analysis using a light microscope and photographed at ×100 magnification. Images 1e and 1f have been zoomed in at 175%.

Figure 2.

Effect of graviola (GR) on human osteosarcoma cell viability after co-culture for 24 h. 4 × 10⁵/mL of (a) MG63 or (b) HOS cells incubated at 37^°C for 24 h in media or with different concentrations of GR diluted in cell line media at 4, 8, 16, 32.5, 75,150 and 300 mg/mL of GR. After which cells were stained with trypan blue and assessed using a light microscope for cell viability. Results are representative of mean % counts of dead cells and ±SD of three independent repeats of the experiments.

Strikingly, when healthy human monocytes were cultured in the presence of GR (150 mg/mL) overnight there was no evidence of cell apoptosis. Monocytes stained 5.0 ± 1.4% (control cells cultured without GR treatment) compared to 5.4 ± 1.5% (cells cultured with GR) positive with trypan blue.

Cell death was further verified by Annexin staining. HOS cells exposed to 300 mg/mL of GR showed 77.8 ± 8% annexin positive staining (P < 0 .0001), compared to only 1.31 ± 0.07% in untreated control cells. Similarly, MG63 cells exposed to 300 mg/mL of GR showed 82.5 ± 1% annexin positive staining (P = 0.0001). These results represented the mean % positive stain in HOS and MG63 cells after co-culture with GR for 24h, compared to control cells not exposed to GR (Figure 3). Also, the degree of annexin staining in cells was found to be dose dependent.

Figure 3.

Evaluation of apoptosis on osteosarcoma cell lines after culture with graviola (GR). 4 × 10⁵/mL of OS cells were incubated at 37^°C for 24 h in either media (control) or with GR at concentrations of 32.5, 75 and 150 mg/mL. Cells were subsequently stained with annexin FITC and analyzed using flow cytometry. Results show the mean fluorescence percentage % of annexin stain compared to controls. All results are representative of mean positive annexin fluorescence ±SD of three independent repeats of the experiments. Bar charts (a) and (b) represent the % annexin stain at 32.5, 75 and 150 mg/L of GR treatment on MG63 and HOS respectively. Histograms (c) and (d) show the overlays results for annexin stain of control versus GR at 150 mg/mL on MG63 and HOS cells respectively.

Downregulation of Osteosarcoma Cell Cytokine Secretion by Graviola

Analysis of cell culture conditioned supernatants collected 6 h following GR treatment from the cells showed that IL-6 and TGFβ were constitutively detected in supernatants from both osteosarcoma cell lines. However, treatment with GR (300 mg/mL), resulted in a significant downregulation of both TGFβ₁ and IL-6 in both HOS and MG63 cells (P < .0001). Cytokine reduction was found to be concentration dependent (Figure 4). The mechanism by which GR induces apoptosis in tumour cells is still not fully understood. To gain deeper insights into the molecular changes involved, we conducted a label-free quantitative proteomic study. This approach allows for an unbiased and comprehensive analysis of protein expression, enabling the identification of key proteins and pathways affected by GR treatment (300 mg/mL) in OS cell lines after 24 h.

Figure 4.

The effect of graviola (GR) treatment on cytokine secretion by osteosarcoma cell lines. 4 × 10⁵/mL of MG63 and HOS cells were incubated at 37^°C for 6 h in either media or with different concentrations of GR diluted in media at 4, 8, 16, 32, 75, 150, 300 mg/mL of GR. Supernatants were collected and analysed for cytokines content using ELISAs. Results represent mean ± SD of three independent repeats of the experiments. The bar charts show the effect of GR on a) TGFβ1 secretion by MG63 (b) IL-6 secretion by MG63 (c) TGFβ1 secretion by HOS (d) IL-6 secretion by HOS.

Differentially Expressed Proteins in MG63 and HOS Cells Following Treatment with GR

Following extraction of proteins from MG63 and HOS cells treated and untreated they were analysed using UPLC with high resolution mass spectrometry. To determine any changes in expression of proteins Proteome Discoverer^® 2.5 was used to was used to quantify the results and find the altered expressions of protein. In case of the MG63 a total of 641 proteins were detected. There were 421 proteins that were similarly expressed in both sets of samples and 195 were only found in the control set and 25 to the treated group. In the case of the HOS cell lines 757 were commonly expressed in treated and untreated group and 212 were unique to the control group and 37 to the treated. To understand the identified proteins and their expressions in each group, all the peptide intensities and their significance levels were quantified, and the abundance levels were compared using Proteome Discoverer® 2.5 software with a stringent filtering criterion. There were many proteins that were found to be altered. However, we filtered the differentially expressed proteins from our study by using the false discovery rate (FDR) of 0.1%, abundance ratio of 2.0 or higher for upregulated proteins and 0.5 and below for down regulated proteins, P value of <.05, percentage coefficient of variation of ≤ 20 % and the identification of at least 2 unique peptides. Our results for MG63 showed that among the 170 altered proteins, 162 were down regulated and 8 were up-regulated. The collated spectroscopy data is summarised and available as Supplemental Figure S9 and Supplemental Tables S2 to S5. In the case of HOS cell line 37 proteins were altered with 33 down regulated and 4 were up regulated. These results are presented in pie charts for each cell line (Figure 5a and b).

Figure 5.

Gene Ontology (GO) analysis of differentially expressed proteins in (a) MG63 and (b) HOS cell lines. The charts categorise proteins based on their associated biological processes, molecular functions, and cellular components, providing understanding into their functional roles within the cells. Six samples of six independent experiments from each cell line were used (4 ×10⁵/mL) after culture either in media with no treatment or with GR (300 mg/mL). Cells pellets were collected after 24 h of incubation at 37^°C with 5% CO_2.

Functional Enrichment and Pathway Analysis

The proteins that were significantly altered following treatment with GR were further examined through functional enrichment analysis using Gene Ontology (GO) terms.

Protein-Protein Interaction Network

To gain insights into the identified proteins and their potential interactions that were down-regulated in MG63 and HOS cells following treatment with GR, we conducted a protein-protein interaction (PPI) network analysis using the STRING database. A high confidence threshold of 0.7 was applied, ensuring a high level of accuracy in identifying relationships between the altered proteins. The resulting PPI networks demonstrated strong associations between numerous proteins. As shown in Figure 6a and b, the analysis for MG63 cells revealed several key interactions, particularly among ribosomal proteins and translation factors, RNA processing and splicing proteins, as well as stress response and chaperone proteins. Additionally, there were notable associations between cytoskeletal and cell adhesion proteins. In the case of HOS there were significantly fewer interactions. There was strong association among chaperons.

Figure 6.

Protein–Protein Interaction (PPI) network analysis of altered proteins in (a) MG63 and (b) HOS Cell lines. The PPI network generated using STRING, illustrates a highly interconnected network of proteins with multiple functional roles.

In light of these results we wished to investigate the apoptosis process further by examining the effect of neutralizing antibodies to p53 and BCL-2 on TGFβ₁ expression. We found that the constitutive expression of TGFβ₁ decreased by 28 ± 3.3% in MG63 cells (P < .001) and by 17.1 ± 3.6% in HOS cells after 24 h of GR treatment. Figure 7a and d shows both the constitutive expression of TGFβ₁ and the decrease in expression after GR treatment. Essentially when GR was used in combination with p53 and BCL-2 antibodies, reconstitution of TGFβ₁ expression occurred. Figure 7b and e represents flow cytometry overlays for the effects of anti-p53 on MG63 and HOS respectively. Figure 7c and f shows the effects of anti-BCL-2 on MG63 and HOS cells respectively. Isotype matched controls for both antibodies displayed TGFβ expression which was comparable to that of in untreated constitutive expression on (Figure 7a and d. control). Table 1 summarizes the detected changes in percentage TGFβ₁ in both cells lines as measured by flow cytometry. Collectively, the results indicate that GR impacts numerous pathways, affecting the structural components of the crucial cell proteins which are involved in multiple cellular processes. We identified that BCL-2 and p53 seem to be involved with TGFβ₁ regulation. The mechanism seems to converge on distinct pathways, leading to powerful cell apoptosis through multiple pathways. Fundamentally, Figure 8 demonstrates how GR and its compounds could facilitate OS tumour apoptosis. This depicts the pathway of mitochondria apoptosis in humans sourced from Kegg pathway database at www.Genome.jp/kegg/pathway/html. The circled proteins represent key molecules identified in the modulation of OS cells survival following GR exposure.

Figure 7.

The effect of graviola (GR) treatment and inhibitors on TGFβ expression by flow cytometry on 2 OS cell lines. 4 × 10⁶/mL of MG63 and HOS cells were incubated at 37^°C for 24 h in media or with GR diluted in media at 300 mg/mL and in the presence of neutralising antibodies to BCL-2 and p53. After which cells were stained for surface TGFβ expression Results represent % mean TGFβ1 ± SD of 3 independent experiments. Flow cytometry histogram overlays showing TGFβ FITC as % positivity. (a) MG63 control with overlay of MG63+GR (b), MG63+IgG with overlay of MG63 with anti-BCL-2 + GR, (c) MG63 with IgG with overlay of MG63 with anti-p53 +GR, (d) HOS control with overlay of HOS+GR (e) HOS+IgG with overlay of HOS with anti-BCL-2 + GR (f) HOS+IgG with overlay of HOS with anti-p53 + GR respectively.

Table 1.

Flow Cytometry Summary Data Showing Mean % Expression of TGFβ₁ After Graviola (GR) and Inhibitor Treatment Expressed as Mean ± SD of three Independent Experiments.

Cell type and conditions	Control unstimulated cells constitutive expression of TGFβ1	+Graviola	+anti-p53+Graviola	+anti-BCL-2+ Graviola
Mean % TGFβ1 expression by HOS cells mean ± SD	71.8 ± 3.3	62.7 ± 5.3	78.7 ± 6.6	75.7 ± 3.7
Mean % TGFβ1 expression by MG63 cells mean ± SD	61.8 ± 3.9	34.3 ± 3.1	85.2 ± 4.1	79.1 ± 3.6

Figure 8.

Mitochondria apoptosis pathway sourced through KEGG pathway database-Genome Net. BCL-2 and p53 are pro-apoptotic proteins which were highlighted as involved in the cell death pathway. This figure represents a proposal of the mechanism involved in GR mediated OS cell line apoptosis.

Discussion

The aim of this study was to investigate whether GR has an anti-tumour effect on OS cell lines. Our results demonstrate that GR treatment is able to trigger OS cell apoptosis verified by both upregulation of annexin and the expression of genes and proteins associated with activation of the apoptotic process. GR also mediated a downregulation of IL-6 and TGFβ1 secretion. Tumour metastasis consists of a series of events involving localized migration and invasion of surrounding extracellular matrix, vessel intravasation leading to blood stream and lymphatic involvement. ²² Further the surrounding tumour microenvironment can play a critical role particularly in OS progression. ²³ Any local inflammation is often characterized by cytokines IL-1 and TNFα secretion which can amplify IL-6 signals. IL-6 causes a positive feedback loop amongst these pro-inflammatory cytokines. IL-6 expression and action even at autocrine level has been shown to stimulate epithelial mesenchymal transition.^24,25 In addition, IL-6 downstream effects include the modulation of key immune cells which combat tumours such as neutrophils, T lymphocytes and natural killer cells. Since these cells surround the tumour they can promote an immunosuppressed environment and tumour tolerance. ²⁶ IL-6 also upregulates T regulatory cells and myeloid derived suppressor cells which contribute to an immunocompromised tumour environment encouraging tumour escape from natural immune defences. ²⁷ Therefore IL-6 could be an ideal target for new OS therapies. A study by Bellavial et al. 2024 showed that blocking IL-6 promoter demethylation caused a decrease in IL-6 expression ²⁸ and inhibition of epithelial mesenchymal transition in three different OS cell lines. Another study by Cortini et al. 2016 showed that the migratory potential of tumour activated mesenchymal stromal cells is potentiated via IL-6 secretion which then drives OS tumour spread. In addition, Gross and colleagues found that IL-6 and CXCL8 ²⁹ mediate tumour lung interactions in OS. Furthermore, TGFβ plays an important role in regulating tumour progression by driving cell invasion and metastasis by increasing tumour cell motility, ³⁰ actin cytoskeleton reorganization and a decrease in intracellular adhesion. ³³ Also, TGFβ₁ is used as a serum tumour marker in OS for monitoring patient treatment efficacy. Elevated levels of serum TGFβ₁ has been predictive of high-grade OS tumours which display chemotherapy resistance. ³¹ Studies have shown that OS patients with lung metastasis express much higher levels of serum TGFβ compared to patients with localized tumours.^32,33 Essentially, TGFβ₁ release by OS cells has been linked to drug resistance, tumour progression and the suppression of immunosurveillance. ³⁴

Hence compounds which can target or neutralize IL-6 and TGFβ could be of significant therapeutic value in new OS treatments. Our results indicate that GR has the potential to downregulate two key angiogenic cytokines (TGFβ and IL-6) which are directly secreted by OS cells, making GR a promising therapeutic for OS.

The proteomic analyses corroborated our immunoassay results and provided information on the proteins and pathways which could be involved in how GR affected OS cells.

In the HOS and MG63 cell lines, 33 and 162 proteins, respectively, were found to be downregulated following treatment with GR. These downregulated proteins belonged to various functional groups, including antioxidant/defence proteins, chaperones, and cytoskeletal proteins. Other affected categories included enzymes and proteins involved in metabolism, heat shock proteins, ion channels, proteases, ribosomal proteins, RNA splicing, binding, and processing proteins, as well as transcription factors and transport proteins. Proteins of particular interest were those associated with apoptosis and the cytokines IL-6 and TGF-β-mediated signalling pathways.

In the MG63 cell line, P09429-High Mobility Group Protein B1 (HMGB1) was found to be downregulated. This protein is known to have a range of functions, including roles in DNA repair, chromatin remodelling, and the regulation of gene expression. It is involved in processes that lead to cancer and inflammation such as oxidative stress, autophagy, and immune response. ³⁵ HMGB1 is often related to the promotion of cell survival and proliferation. By regulating genes involved in cancer cell growth, cell cycle progression, and inhibition of apoptosis, HMGB1 can contribute to the survival of cancer cells. Therefore, it’s downregulation could reduce the MG63 cells' ability to proliferate and avoid apoptosis. Downregulation of this protein following treatment with GR may reduce the cells' ability to use autophagy as a protective mechanism, making them more vulnerable to drug-induced apoptosis or necrosis. Also, the reduction of this protein could promote apoptosis by reducing its protective effects against oxidative stress and preventing it from playing a role in DNA repair, which is crucial for cancer cell survival.

P26583-High Mobility Group Protein B2 (HMGB2), another member of the High Mobility Group Box (HMGB) family, was also found to be downregulated in MG63 cells following treatment with GR. Like HMGB1, its downregulation may impair the ability of MG63 cells to repair damaged DNA, making them more susceptible to the cytotoxic effects of GR, which may cause oxidative stress or DNA damage. Also, high levels of HMGB2 have been related to poor OS progression. ³⁶

P13693 - Translationally-controlled tumour protein (TCTP) was found to be downregulated in MG63 cells treated with GR. TCTP plays a crucial role in cellular functions, such as regulating cell growth, the cell cycle, and stress responses. It is also involved in controlling apoptosis and promoting cell survival. The downregulation of TCTP could be part of the stress response, aiming to limit cell proliferation by inducing cell cycle arrest and enhancing apoptosis in tumour cells. This downregulation may serve as a mechanism to slow tumour progression and growth in response to the stress caused by GR. TCTP is a key drug treatment target in OS shown by both in vivo and in vitro studies. ³⁷

Cell migration is a complex, multi-step process that begins with the formation of protrusions on the cell membrane, triggered by migratory and chemotactic signals. This process is mainly driven by the polymerisation of actin filaments beneath the membrane, which facilitates cell movement. Recent research has highlighted that certain proteins responsible for connecting migratory signals to the actin cytoskeleton are often overexpressed in cancer cells, ³⁸ particularly those with invasive and metastatic traits. The downregulation of the proteins P61158, Q16643, P15311, P47756, O75369, P06396, Q9UHB6, P40121, P26447, P35241, P16949, Q9Y490, P37802, P18206, and Q15942 in MG63 cells following treatment with GR may be associated with the inhibition of cell migration, invasion, and metastasis. Many of these proteins are involved in important processes such as cytoskeletal organisation, actin filament dynamics, and cellular signalling that promote cell movement. For instance, P61158 (Actin-related protein 3) and P15311 (Ezrin) are involved in actin cytoskeleton remodelling, ³⁹ which is essential for the formation of cell protrusions that enable migration. Downregulation of these proteins would reduce the cell's ability to extend protrusions, impairing its movement.

Q16643 (14-3-3 protein sigma, SFN) is a regulatory protein that plays a role in cell cycle regulation and apoptosis; its downregulation may reflect a cellular response to maintain stability and prevent unwanted cell migration under stress conditions, such as treatment with GR. Proteins like P47756 (F-actin-capping protein subunit beta), O75369 (Filamin-B), and Q9UHB6 (Peptidyl-prolyl cis-trans isomerase FKBP14) are involved in stabilising the actin cytoskeleton and regulating actin filament dynamics. Their downregulation can disrupt the structural integrity of the cytoskeleton, thus impairing cellular motility. ⁴⁰

P06396 (Gelsolin) regulates actin filament assembly and disassembly, and its downregulation could hinder the formation of the dynamic structures necessary for cell movement. P40121 (Macrophage capping protein) and P26447 (Protein S100-A4) are associated with cell signalling and cytoskeletal interactions. Their downregulation suggests that the cell may be reducing migratory and invasive behaviour in response to GR. Similarly, P35241 (Radixin) and P16949 (Stathmin) contribute to cytoskeletal dynamics and cell shape changes, and their downregulation could lead to reduced cell motility. In OS, gelsolin has been found to be upregulated in tumour tissue and correlates with tumour size. Research has shown that gelsolin knockdown resulted in significant reduction in tumour proliferation and size in OS cell lines therefore it is an attractive therapy target. ⁴¹

The downregulation of Q9Y490 (Talin-1) and P37802 (Transgelin-2) also points to a reduction in cellular interactions with the extracellular matrix, further limiting the ability of the cells to migrate and invade. Lastly, P18206 (Vinculin) and Q15942 (Zyxin). are involved in cell adhesion and mechanical signalling; their downregulation could damage the adhesion and signalling pathways that support cell migration. ⁴² Overall, the downregulation of these proteins shows a coordinated response to prevent excessive migration and invasion induced by GR, which is a common cellular response to cancer treatment aimed at limiting tumour progression and metastasis. ⁴³

In HOS cells treated with GR, Q15691 (CAP1), P35241 (Radixin), and O75347 (Filamin-B) were found to be downregulated. This may be a mechanism to inhibit processes such as metastasis, migration, and invasion. CAP1 regulates cytoskeletal dynamics and cAMP signalling, ⁴⁴ and its downregulation may limit tumour spread. Radixin and Filamin-B are actin-binding proteins that maintain cell structure and motility, and their downregulation can disrupt the cytoskeleton, impairing cell migration and proliferation. Together, this reduction in key proteins helps prevent cancer cell invasion and promotes tumour cell death, enhancing the therapeutic effects of GR.

The downregulation of Q9NYF8-Bcl-2-associated transcription factor may be part of a wider cellular response. BclAF1 is involved in regulating apoptosis by interacting with the BCL-2 family, which controls the mitochondrial pathway of cell death. ⁴⁵ When cancer cells are treated with GR, it may reduce BclAF1 levels, enhancing pro-apoptotic signals and making the cells more susceptible to apoptosis. This downregulation could disrupt the survival pathways typically helped by BclAF1, sensitizing HOS cells to the treatment and pushing them towards cell death.

Q96EY1-DnaJ homolog subfamily A member 3, mitochondrial (DNAJA3), a protein that is involved in maintaining mitochondrial function and regulating apoptosis, ⁴⁶ was also found to be downregulated in HOS cells. Downregulation of DNAJA3 could enhance apoptosis by disrupting mitochondrial homeostasis, making cells more susceptible to cell death pathways that are activated by GR.

Following treatment with GR, significantly fewer proteins were upregulated in MG63 and HOS cells, with eight upregulated in MG63 and only four in HOS. Additionally, P21926 (CD9 antigen) was upregulated in MG63 cells. CD9 has been linked to drug resistance ⁴⁷ by promoting cell survival through modulating signalling pathways, helping cancer cells resist apoptosis. It also influences immune interactions within the tumour microenvironment, possibly enhancing immune evasion. ⁴⁸

Another protein that was upregulated in HOS cells that should be highlighted here is P60953 (cell division control protein 42 homolog, CDC42). CDC42 is a key regulator of cell cycle progression and division and its upregulation in HOS cells following treatment with GR may be an adaptive response to maintain or increase cell proliferation under stress. In addition, CXCR4 tumour migration and invasion is thought to be inhibited by downstream CDC42 expression in OS tumours. ⁴⁹

To note there is a difference in the proteomic response to GR by the two cell lines used in this study. This is not surprising as the OS cells can be extremely heterogenous and both cells types have different immunophenotypic profiles as well as growth and proliferation rates. ⁵⁰

We have identified that GR targets cell proteins which function in cell division, mitochondrial respiration, protein synthesis (ER function), transcription, cytoskeletal cell structure, enzyme function and cytoplasm transportation. Our observations are supported by the findings by Tawaill et al. In this study, rats with induced tumours were treated with GR for 30 days which resulted in a decrease in expression of CD44, TGFβ, BCL-2 and tumour volume. ⁵¹ Recently, there has been a renewed interest in p53 based therapy as a gene target especially as an immunotherapy targeting bone diseases. ²¹

An in vivo study which analysed the blood of rats with induced breast cancer after a 30-week period of GR treatment, found that the breast cancer was ameliorated. ⁵² The blood work showed an upregulation of the pro-apoptotic gene BCL-2, increases in antioxidants, a decrease in the estrogen receptor α gene and lowered lipid peroxidation levels.

All the GR tree parts, including the leaves, stems, seed, fruit and pulp of GR have been proven to possess a range of therapeutic potential when used in the form of tablets, tinctures or fruit teas, smoothies or juices.^53,54 The chemical constituents of GR include alkaloids, flavanols, triglycerides, phenols, cyclopeptides, megastigmones and over one hundred acetogenins (ACG).^55,56 These are C35-C37 fatty acid derivatives which are biosynthesized through the polyketide pathway. One mode of ACG action is these compounds are able to deprive tumour cells of adenosine triphosphate (ATP) by disrupting mitochondrial transport ⁵⁷ which activates tumour apoptosis. They have also been found to suppress multi-drug resistant cancer cell growth which is also a major complication of OS disease by inducing cell death. ⁵⁸ New therapies for OS are required as it is one of the most aggressive and deadly primary bone tumours and despite intensive chemotherapeutic regimens and new antibody mediated immune therapy the cure rate remains extremely low. Also, the use of invasive surgeries can remove the bone tumour but ultimately sacrifices healthy bone tissue leaving young children with life changing mobility issues. Essentially, children whom have been treated with anthracyclines which are included in OS treatment regimens are at high risk of developing irreversible heart issues such as cardiomyopathy and myocardial contraction problems. ⁵⁹ Other side effects also include infertility, secondary cancers and chemotherapy resistance resulting in relapse. ⁶⁰ GR does not have any known side effects although its use should be monitored in diabetics since it can lower insulin levels.^61,62

The tumour microenvironment including surrounding cells, tissues, proteins and secreted cytokines can have a major impact on cancer outcomes especially in OS in which bone cells rapidly multiply. ⁶³ Hence, therapies which can modulate both the tumour and secreted cytokines are crucial as they could help slow down OS tumour progression.^8,64 Limitations of the study are that the results are from the use of 2 osteosarcoma cell lines. Future research should test GR and its extracts clinically on OS patients and monitor disease outcomes.

Conclusion

In conclusion, we have found that GR induces apoptosis in OS cell lines and downregulates the secretion of two key angiogenic cytokines IL-6 and TGFβ which are also known to promote tumour cell growth and mobility. This downregulation of TGFβ by GR was modulated by p53 and BCL-2. The bioinformatic data suggests that GR treatment resulted in a decrease in the relative abundance of proteins essential to cell cycle division, mitochondria respiration, gene transcription and cytoskeletal mobility. Together these results suggest that GR should be strongly considered in clinical trials for the integrative management of OS.