Mannotriose‑containing extract of Streptomyces sennicomposti GMY01 induces the apoptosis of HeLa cells and exhibits binding affinity with BCL‑2 and MCL‑1 proteins

Introduction

Almost all cases of cervical carcinoma originate from the ectocervical or endocervical mucosa in the transformation zone, the area of the cervix between the old and new squamocolumnar junction (1). In 2020, the number of new cases of cervical cancer reached 604,127, with related deaths ranking ninth among all types of cancer worldwide (341,831 individuals) (2). New cases of cervical cancer in Indonesia alone reached 36,633, with the third highest number of deaths from total cancer in Indonesia (21,003 individuals) in the same year (2). Chemotherapy has been applied for follow-up therapy following surgery since the 20th century (2); however, the high mortality rate still indicates that existing chemotherapeutic drugs are not optimal. This description provides the basis for the importance of the development of novel cancer drugs, particularly for cervical cancer.

Candidate targets for the development of new drugs can be determined through the hallmarks (main markers) of cancer, one of which is the avoidance of apoptosis (3). Apoptosis is programmed cell death that aims to maintain homeostasis. The process of apoptosis in cancer cells is inhibited, and the homeostasis of the body becomes chaotic (4). The loss of control over apoptosis causes cancer cells to survive for longer time periods and allows for the accumulation of various mutations that occur, increasing the invasiveness of cancer cells during tumor development, stimulating angiogenesis, deregulating cell proliferation and interfering with the differentiation process (5). One of the proteins that regulate the apoptotic mechanism of the intrinsic pathway is the B-cell lymphoma-2 (BCL-2) family, which includes pro-apoptotic effector proteins, pro-apoptotic BH3-only and anti-apoptotic BCL-2(4). This finding supports the notion that these two molecules have the potential to become targets for the development of novel cancer drugs.

Streptomyces sennicomposti (S. sennicomposti) GMY01 is an actinomycetes bacteria obtained from Krakal Beach, Gunungkidul, Yogyakarta, Indonesia (6). The results of genome mining have revealed that there are functional genes from S. sennicomposti GMY01 that encode mannotriose (7). The presence of the mannotriose extract of GMY01 cell biomass was previously confirmed by liquid chromatography tandem mass spectrometry, and optimization of the isolation and purification method of mannotriose from S. sennicomposti GMY01 was successfully carried out, although it has not been 100% purified (7). The methanolic extract containing mannotriose has exhibited moderate anticancer activity against HeLa cervical carcinoma cells (IC50, 27.31 µg/ml) and HTB-179 lung carcinoma cells (IC50=33.75 µg/ml) and exerts minimal toxic effects on mammalian normal cells (Vero cells) with an IC50 of 823.3 µg/ml (7). The results of that study indicate the potential of mannotriose to be developed as a novel cancer drug that induces apoptosis.

Mannotriose is an oligosaccharide that is derived from raw materials that are easily obtained from various plants; therefore, its development as a novel cancer drug targeting anti-apoptotic proteins, one of which is BCL-2, has promising developmental potential. In silico methods are computational approaches that examine the molecular structure and quantitative structure-activity relationship; therefore, such studies can promote drug discovery and development and reduce preclinical research (8). The in silico docking analysis carried out in the present study was performed to examine the binding of mannotriose molecules from S. sennicomposti GMY01 to BCL-2 and MCL-1, in order to determine their potential to induce apoptosis through the inhibition of these two anti-apoptotic proteins. It was hypothesized that mannotriose of S. sennicomposti GMY01 has the potential to be developed as a novel cancer drug. The present study also performed an in vitro assay to examine the induction of apoptosis by the methanol extract (ME) and n-hexane-free methanol fraction (HMF) using HeLa cells. The present study aimed to examine the potential of mannotriose from S. sennicomposti GMY01 to induce the apoptosis of cervical cancer cells by inhibiting BCL-2 and MCL-1.

Materials and methods

Reagents

The isolate of S. sennicomposti GMY01 was obtained from a previous study and deposited in the Indonesian culture collection (WDCM 769), National Research and Innovation Agency, Indonesia, encoded as InaCC A147(7). The HeLa (ATCC CRM-CCLL-2) and C2C12 (ATCC CRL-1772) cell lines were obtained from the Cell Culture Laboratory of the Pharmacology Department, Faculty of Medicine, Public Health and Nursing, Gadjah Mada University, Yogyakarta, Indonesia. HeLa cells were grown in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco, Thermo Fisher Scientific, Inc.); hence, the C2C12 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Gibco, Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Inc.).

Fermentation of S. sennicomposti GMY01

The production and extraction of metabolites from the cell biomass of S. sennicomposti GMY01 were carried out using an optimized method described in a previous study (9). The bacterial isolates were re-cultured in ISP-2 (Difco, BD Biosciences) solid media and incubated at 30˚C for 7 days or until there were numerous spores. The cultures were then transferred to TSB (Difco, BD Biosciences) medium to produce starter cultures. The GMY01 culture in ISP-2 was then sampled with a transfer loop, placed in 100 ml of starch nitrate buffer (SNB; modified medium), and incubated in a rotary shaker incubator at 225 rpm for the following 3 days. The SNB medium contained 0.5 g NaCl, 1 g KNO3, 0.5 g K2HPO4, 0.5 g MgSO4, 0.5 g 7H2O, 0.01 g FeSO4 7H2O, and 20 g soluble starch in 1,000 ml distilled water. All chemical reagents were obtained from Merck KGaA. Subsequently, 5% (v/v) starter culture was inoculated into nine 250 ml Erlenmeyer flasks (Pyrex®, Merck KGaA) containing 100 ml SNB medium, followed by incubation on a rotary shaker incubator at 225 rpm for 11 days at 30˚C.

Extraction and fractionation

The fermentation culture was then transferred into 50 ml conical tubes (BD, Corning Life Sciences) and centrifuged at 1,000 x g, at 4˚C for 10 min. The cell biomass was separated from the supernatant following centrifugation by removing the supernatant. The cell biomass was macerated using 300 ml absolute methanol for 30 min, with slow stirring, at room temperature. Following completion, it was filtered using Whatman paper (Merck KGaA) and the methanolic extract was collected. The ME was concentrated using a vacuum rotary evaporator. The HMF was prepared by liquid-liquid extraction at a 1:1 ratio of methanol:n-hexane as the solvent. The fraction dissolved in n-hexane was then removed, leaving the remaining fraction in methanol. Furthermore, the ME and HMF were evaporated using a rotary evaporator, as previously described (7). The HMF was then prepared at up to 100,000 ppm of the stock solution in the same manner as the ME.

Cytotoxicity assay using HeLa and C2C12 cells

A stock solution of 100,000 ppm was prepared for in vitro testing by dissolving 100 mg of the extract in DMSO to a volume of 1 ml. The two extract stock solutions were then diluted into seven series of concentrations for MTT assay using HeLa (3.13-200 µg/ml) and C2C12 cell lines (31.25-20,000 µg/ml). The HeLa cell culture suspension (100 µl) at a density of 1x105 cells/ml and 100 µl of each extract at various concentrations was added to a 96-well plate and incubated for 24 h at 37˚C. The culture medium was then replaced with 100 µl fresh complete medium (RPMI for HeLa cells and DMEM/FBS 10% for C2C12 cells) and 10 µl of MTT solution following the incubation period. The plates were incubated for 4 h at 37˚C. A total of 100 liters of 10% SDS solution in 0.01 M HCl was added after incubation was completed followed by a further incubation in room temperature for 24 h. The absorbance of each well was measured at a wavelength of 595 nm using a microplate reader (Bio-Rad Laboratories, Inc.). The percentage of cell viability was calculated by comparing the absorbance values of the wells treated with each extract with the absorbance of the wells that were not treated (negative control). The percentage of cell viability was used to calculate the concentration of the extract that could inhibit 50% of cell growth (inhibition concentration 50% or IC50). The selectivity index (SI) was calculated from the ratio of IC50 on normal to cancer cells (10).

Apoptosis assay

The apoptotic activity of the extract and fraction was assessed in vitro using flow cytometry with Annexin V-PI staining. After each well containing cells was induced with both types of extracts (or only culture media for the negative control group, and doxorubicin was used as the positive control), 5x105 cells were collected by centrifugation at 1,000 x g at 4˚C for 5 min. Cell resuspension was then carried out by the addition of 500 µl Annexin V binding buffer solution, then 5 µl Annexin V-FITC, and 5 µl PI, and finally incubated at room temperature for 15 min (in a dark room). Annexin V-FITC binding was then analyzed by flow cytometry (BD FACSCanto™ II, BD Biosciences) (Ex: 488 nm; Em: 350 nm) with an FITC signal detector, while PI staining analysis was carried out with a phycoerythrin emission signal detector. The results of flow cytometry indicate the distribution of cell counts in each of the four quadrants in the Annexin V-FITC test results, which were then analyzed. Quadrant I (P1) contained a collection of cells that were still alive, quadrant II (P2) contained a collection of cells that underwent early-phase apoptosis, quadrant III (P3) contained a cell that underwent late-phase apoptosis, and quadrant IV (P4) contained a collection of cells that underwent necrosis. The results in the control and treatment groups (the given extract) were then compared to determine the proportion of cells that were still alive, underwent early-stage apoptosis, advanced-stage apoptosis, or necrosis.

In silico molecular docking analysis

In silico molecular docking was initiated by determining the most optimal conformation of the protein-ligand complex of PDB for the following steps. The native ligand was removed to provide space (pocket/cavity) for the ligand-binding area. The validation of the molecular docking method and the determination of the grid box for docking were then carried out before docking with the ligand of the compound to be tested, namely mannotriose. Validation was performed by re-docking native ligands on proteins whose native ligand were removed using the AutoDock program. The Root Mean Square Deviation (RMSD) of the complex of native ligand and analyzed protein was calculated for docking validation. The protein can be used further for docking if the RMSD value is <2.0 Amstrong (Å) in the re-docking process. In the case that the RMSD value is >2.0 Å, the certain conformation cannot be used for the docking; thus, it must be replaced with another conformation with a different PDB ID. The grid box was determined after obtaining BCL-2 and MCL-1 proteins, which were used for subsequent docking. The grid is an area where the binding of ligand and its protein/macromolecule supposed to be, which is expressed in box coordinates in the AutoDockTools application. Grids for the binding areas of BCL-2 and MCL-1 were obtained after the re-docking of proteins associated with their respective native ligands. Mannotriose 3D structure preparation was performed using the online application MolView (https://molview.org/) based on the 2D structure. This 3D structure was then docked to BCL-2 and MCL-1 proteins, which were separated from their native ligands. Furthermore, the docking process was performed on BCL-2 and MCL-1 proteins using AutoDockTools. A ligand-receptor complex with a binding energy below 0 kcal/mol indicates that the ligand can bind to the associated protein. Both 2D and 3D visualizations of amino acid residues interacting with the ligand were processed using Discovery Studio, as previously described (7,9).

Statistical analysis

Statistical analyses were performed using IBM SPSS 16.0 in Windows PC operating system (IBM Corp.). The mean IC50 values of ME and HMF in both cell lines were compared using the t-test. The mean comparison of AI for each group was analyzed using one-way ANOVA with Tukey's post hoc test (GraphPad Prism 10.0.2, Dotmatics). The IC50 values are displayed as averages, whereas SI and AI are presented as ratios. The comparison of the ratio of early and late apoptosis was performed using the t-test, apart from the HMF 0.5xIC50 group, which was analyzed using the Mann-Whitney U test. A value of P<0.05 was considered to indicate a statistically significant difference.

Results

Cytotoxicity assay

The results of MTT assay revealed that the HMF had a lower IC50 value than the ME, which was 93.75 compared to 78.81 µg/ml; however, the difference was not statistically significant (P=0.843). In addition, the SI of HMF in the HeLa cells was significantly higher (117.04) than in C2C12 cells (6.23). The IC50 values of ME and HMF in HeLa cells were still lower than those of the positive control, doxorubicin (2.88 µg/ml), although doxorubicin had a very low SI value of 0.135. The results of MTT assay are presented in Table I.

Table I IC50 and SI of Streptomyces sennicomposti GMY01 extract in HeLa and C2C12 cell lines.

Table I

IC50 and SI of Streptomyces sennicomposti GMY01 extract in HeLa and C2C12 cell lines.

			IC50
No.	Cell line	Treatment	Estimated	Lower	Upper	SI	P-value
1.	HeLa	Methanol extract	93.746	50.850	281.292	6.232	0.843
		n-hexane free fraction	78.813	43.113	218.592	117.036
		Doxorubicin	2.876	1.007	5.502	0.135	-
2.	C2C12	Methanol extract	584.314	411.826	915.158	-	0.001a
		n-hexane free fraction	9223.977	4100.032	55213.755
		Doxorubicin	.388	.043	1.028		-

[i] ^aStatistically significant difference, as determined using the Mann-Whitney U test. SI, selectivity index;

The results of flow cytometry analysis of Annexin V PI revealed that the mean apoptotic index (AI) differed significantly between each test group, with ME 1xIC50 having the highest value at 4.01±0.39. Analysis with the post hoc test revealed that the AI in the ME 1xIC50 and positive control (doxorubicin) was significantly higher than that in the negative control (culture medium; both P=0.001). The positive control also had a significantly higher AI than the two extracts at a concentration of 0.5xIC50 (P=0.004 and 0.003, respectively). Higher AI values at 1xIC50 compared to 0.5xIC50 were found with the ME and HMF, although a significant difference was only found with ME (P=0.005). The HMF had a lower AI than the ME, but both differences were not statistically significant (2.32±0.48 compared to 2.33±0.44 at 0.5xIC50 levels with P=1.000, and 2.99±0.27 compared to 4.03±0.39 at 1xIC50 levels with P=0.151). The results of flow cytometry are presented in Figs. 1 and 2.

Figure 1 Flow cytometry plots of: (A) Methanolic extract 0.5xIC50, (B) methanolic extract 1xIC50, (C) free n-hexane fraction 0.5xIC50, (D) free n-hexane fraction 1xIC50, (E) doxorubicin 1xIC50, (F) culture medium only. The quadrants are as follows: PI, viable cells; P2, early apoptotic cells; P3, late apoptotic cells; P4, necrotic cells.

Figure 2 Comparison of apoptotic index in each group.

Additional analyses were then performed to compare the means using a t-test (or Mann-Whitney test for non-parametric data) between the proportions of apoptotic phases (early vs. late phase) in each test group. ME and HMF, at concentrations of 0.5xIC50 and 1xIC50, exhibited the same tendency in the apoptotic phase, which was significantly dominated by the early phase of apoptosis. Moreover, the positive and negative controls exhibited the opposite tendency, where late-phase apoptosis dominated, although the difference was not statistically significant. The results of this analysis are illustrated in Figs. 1 and 3.

Figure 3 Comparison of the apoptotic phase in the test and control groups.

In silico molecular docking

In the present study, the in silico analysis used several protein conformations of BCL-2 (PDB-ID: 4IEH, 7LHB, 6O0K and 4LVT) and MCL-1 (PDB-ID: 6OQB and 6FS0). These conformations were selected as all proteins were derived from Homo sapiens, had a good wwPDB validation value, and bound to native ligands in the form of inhibitor compounds, as shown in Fig. 4 and Table II. The docking results of mannotriose from S. sennicomposti GMY01 for each protein in all these conformations are presented in Table III.

Figure 4 wwPDB validation of molecular structure with PDB-ID: (A) 4IEH, (B) 7LHB, (C) 6O0K, (D) 4LVT, (E) 6OQB, (F) 6FS0.

Table II Protein codes used in docking of cancer proteins.

Table II

Protein codes used in docking of cancer proteins.

Protein	PDB-ID	Native ligand	Original species	Mutation
BCL-2	4IEH	IE9	Homo sapiens	No
BCL-2	7LHB	ABBV-167	Homo sapiens	Yes
BCL-2	6O0K	Veneto-clax	Homo sapiens	No
BCL-2	4LVT	Navito-clax	Homo sapiens	Yes
MCL-1	6OQB	N0J	Homo sapiens	No
MCL-1	6FS0	E4W	Homo sapiens	No

Table III Ligand-protein docking results of cancer-related proteins.

Table III

Ligand-protein docking results of cancer-related proteins.

					RMSD
Protein (PDB-ID)	Ligand	Molecular formula	Molecular weight (kDa)	Binding energy	Upper	Lower
BCL-2 (4IEH)	IE9	C42H44Cl N7O6S3	874.49	-6.95	0	1.64
	Mannotriose	C18H30O16	502.42	-1.10	0	1.96
BCL-2 (7LHB)	ABBV-167	C46H53ClN7O11PS	978.45	-10.29	0	1.97
	Mannotriose	C18H30O16	502.42	-0.54	0	1.97
BCL-2 (6O0K)	Venetoclax	C45H50ClN7O7S	868.44	-6.30	0	1.71
	Mannotriose	C18H30O16	502.42	-0.42	0	2
BCL-2 (4LVT)	Navitoclax	C47H55ClF3N5O6S3	974.62	-11.08	0	1.91
	Mannotriose	C18H30O16	502.42	-0.27	0	1.95
MCL-1 (6OQB)	N0J	C32H39ClN2O5S	599.18	-15.03	0	0.23
	Mannotriose	C18H30O16	502.42	-0.75	0	1.99
MCL-1 (6FS0)s	E4W	C35H34ClN5O3S2	672.26	-16.21	0	0.07
	Mannotriose	C18H30O16	502.42	-1.33	0	1.94

[i] Binding energy is presented in in kcal/mol. RMSD, root mean square deviation.

The value of binding energy <0 in the complex between mannotriose and all proteins in each conformation (the lowest binding energy values with BCL-2 and MCL-1 were -1.10 and -1.33 kcal/mol, respectively), indicating that mannotriose could bind to BCL-2 and MCL-1 in all the selected conformations. However, these binding energy values were still higher than those of the native ligand, indicating that the native ligand is better at binding to the related protein. The complete docking results of mannotriose on BCL-2 and MCL-1 are presented in Table III, while the visualization of the most stable bond between mannotriose and BCL-2 and MCL-1 and the redocking results with native ligands is presented in Figs. 5 and 6, respectively.

Figure 5 Visualization in 2D (left) and 3D (right) of the complex between BCL-2 (4IEH) and (A) IE9 or (B) mannotriose.

Figure 6 Visualization in 2D (left) and 3D (right) of the complex between MCL-1 (6FS0) and (A) E4W or (B) mannotriose.

Discussion

In the present study, an in vitro MTT assay was carried out to determine the viability, IC50, and SI values of both the ME and HMF from the S. sennicomposti GMY01 cell biomass. The results of the probit analysis of the viability values obtained indicated that ME and HMF were inhibited in HeLa cells, which means that they are in accordance with the proposed hypothesis. The HMF in the present study was found to be more potent than the ME, as indicated by its lower IC50 value, although the difference between the two extracts was not statistically significant. These results were considered to be due to the higher proportion of manotriose in the HMF compared to the ME, as was the previous hypothesis, which suggested that manotriose acts as an inducer of apoptosis in cancer cells. On the other hand, the analysis of C2C12 cell viability revealed that HMF was non-toxic, while ME was moderately toxic.

The IC50 values of cell lines extracted from S. sennicomposti GMY01 in the present study were compared with those of a previous study (7) as that study also carried out an in vitro assay with ME and HMF from the same cell biomass. The IC50 results for the HeLa cell line due to the administration of ME in the present study were still higher than those in HTB cells (33.56 µg/ml) (7); this suggests that the cytotoxic activity of this extract was more potent in lung cancer cell line HTB. Moreover, the IC50 value of HMF and ME in the present study was lower than of the MCF-7 cells, which were 858 and 245 µg/ml (11), but also lower than of HeLa cells in another study with an IC50 of 27.31 µg/ml (7). This was the basis for the assumption that ME cytotoxicity was more potent in HeLa cells than in MCF-7 cells, but less potent than HTB. The HMF in the present study was more potent in terms of its cytotoxic effect on HeLa cells than on MCF-7 in a previous study (11). The SI values of ME and HMF tested were better than those of the positive control, namely doxorubicin, where HMF had the highest SI of 117.036. This means that it takes as much as 117,036 times its concentration in cancer cells, in this case HeLa cells, to induce an inhibitory effect on normal cells, such as C2C12 cells, as shown in the present study. This result is in accordance with the hypothesis that ME had an SI of 6.232 times. ME and HMF had a higher SI value than the positive control, doxorubicin; thus, it could still be said that these two had a good SI value.

Flow cytometry was carried out to determine whether the cell death due to treatment belonged to the early or late phase of apoptosis, or even necrosis. Annexin V is a protein complexed with phosphatidylserine on the outer surface of cell membranes undergoing apoptosis, even in the early phase of apoptosis. Necrotizing cells, although also dead, did not react with Annexin V as there was no phosphatidylserine on the outer surface of the membrane in cells that did not undergo apoptosis. PI is a nuclear dye. This molecule can bind when the cell is lysed as the molecule is large and cannot penetrate the membrane. This underlies the fact that PI can provide an overview of cells that have been lysed, either due to late-phase apoptosis or necrosis. The results of the present study indicate that apoptosis caused by the administration of the extract and fraction was significantly dominated by the early phase of apoptosis. Escalating the concentration to 1xIC50 increased the number of cells that underwent apoptosis, although the dominance of the early phase was still present. This is different from the culture medium (negative control) and doxorubicin (positive control), where the apoptosis that occurs in cells is dominated by the advanced phase, although the difference was not statistically significant.

The results of the present study could not prove the hypothesis as the results revealed that the ME had a higher AI than HMF, both at 0.5xIC50 and 1xIC50 levels, although the difference was not statistically significant. The present study demonstrated that the administration of ME and HMF had a significantly higher AI than the negative control, indicating that both ME and HMF could induce apoptosis. The extract levels of 0.5xIC50 in both treatments and the HMF 1xIC50 did not exhibit a statistically significant difference compared to the negative control, although both had higher AI values. The ME at 1xIC50, as well as the positive control were found to have a significantly higher AI than the negative control, indicating that ME at these levels significantly induced apoptosis. The AI value of the ME at 1xIC50 was not significantly different from that of the doxorubicin group (positive control). This indicates that the extract has the potential to induce apoptosis similar to drugs already on the market, doxorubicin.

The results of the AI calculation also revealed that 1xIC50 ME was significantly higher than its 0.5xIC50 concentration, which indicated that there was an increase in apoptotic induction activity with increasing extract concentrations. The HMF at 1xIC50 also had a higher AI than the same extract at a concentration of 0.5xIC50, in line with the ME extract, although the difference was not statistically significant. This is in accordance with the concentration-effect theory, which states that the concentration of a substance is directly related to its effect (12).

The mannotriose mechanism of action of BCL-2 and MCL-1 was analyzed in silico by molecular docking. All the docking results of mannotriose on BCL-2 and MCL-1 revealed a negative binding energy, although the value was still higher than that of the native ligand. The most stable conformation of the mannotriose complex with BCL-2 was with the protein with the code of 4IEH, with a binding energy of -1.10 kcal/mol (RMSD range, 0-1.96). Van der Waals bonds were found between mannotriose and BCL-2 in the following six amino acids: GLU(54), PHE(57), THR(60), PHE(79), VAL(82) and LEU(86). Conventional hydrogen bonds were found in 3 amino acids: PHE(53), ARG(56) and SER(83). Both Van der Waals and hydrogen are classified as weak intermolecular interactions (12), while the inhibitors that have been developed (native ligands) have multiple covalent bonds with BCL-2, which are strong intermolecular bonds. This resulted in, although still able to bind stably, a lower affinity of the mannotriose complex with BCL-2 than that of the existing inhibitors. Inhibitors with the code name IE9 have covalent Pi-sigma, Pi-sulfur, Pi-Pi, alkyl, and Pi-alkyl bonds, which are much stronger than van der Waals and hydrogen bonds.

The most stable mannotriose complex with MCL-1 was with protein corresponding to the 6FS0 code, with a binding energy of -1.33 kcal/mol. Van der Waals bonds were found in the following 12 amino acids: PHE(228), MET(231), LEU(235), VAL(249), MET(250), PHE(254), ASP(256), THR(266), LEU(267), PHE(270), LEU(290) and ILE(294). In addition, there were conventional hydrogen bonds in the 3 amino acids as follows: LEU(246), VAL(253) and ARG(263). The arginine amino acid, ARG(263), in MCL-1 also binds to manotriose via unfavorable donor-donor bonds. Van der Waals and hydrogen are classified as weak intermolecular interactions (13), whereas inhibitors that have been developed (native ligands) have multiple covalent bonds with MCL-1 which is a strong intermolecular bond. This explains why the affinity of mannotriose for MCL-1 is lower than that of existing inhibitors. The inhibitor with the code name AZD5991 has covalent bonds of Pi-sigma, Pi-sulfur, Pi-Pi, alkyl and Pi-alkyl types, which are much stronger than the van der Waals, hydrogen and ionic bonds (Fig. 6).

Van der Waals bonds are intermolecular bonds formed owing to the electrical interactions between very close molecules. This bond is the weakest type of bond, with a bond strength of 0.4-4.0 kK/mol and an intermolecular distance of 0.3-0.5 nm. This type of bond, although weak, can lead to very strong interactions between two molecules if there are a large number of them. The most stable complex conformation in this study had 7 van der Waals bonds between manotriose and BCL-2 and 12 bonds with MCL-1, and the number of bonds was sufficient to cause strong interactions between manotriose and BCL-2 and MCL-1. On the other hand, two molecules that are too close, on the other hand, will repel each other strongly, repel each other, even though they have Van der Waals bonds. This causes the bond to require a large amount of energy to prevent the interaction surface between the two molecules from sterically influencing each other (13). This energy requirement explains why the binding energy of manotriose with BCL-2 and MCL-1 is less negative than of the existing inhibitors because more energy is needed to maintain the number of van der Waals bonds between the ligand and the protein.

Hydrogen bonds is an attractive interaction involving a hydrogen atom in a molecule with another atom that has a high affinity for electrons (electronegativity). The difference in electronegativity causes this bond to have a higher strength than normal dipole-dipole interaction but less than covalent bonds, with a bond strength of 10-30 kJ/mol (14). Hydrogen bonds had the second highest number after van der Waals bonds between mannotriose and BCL-2 and MCL-1 in the present study, each involving 3 amino acids. This bond plays a role in stabilizing the complex formed, so that the bond affinity is still negative, which means that manotriose can bind to both proteins.

In conclusion, the results of the present in vitro study demonstrated that the HMF mannotriose concentration was higher than that in the ME and was more potent towards the HeLa cancer cell line, with higher selectivity, as well relative to the C2C12 cell line. However, it was found that the apoptotic indices of HMF at both 0.5xIC50 and 1xIC50 were lower than those of the corresponding ME group, although the differences were not statistically significant. The results of in silico analysis revealed that mannotriose from S. sennicomposti GMY01 could bind to BCL-2 and MCL-1. Therefore, it can be concluded that the ME and HMF from S. sennicomposti GMY01, which was suspected to contain mannotriose, have the potential to induce apoptosis in cancer cells, presumably through the mechanism of inhibition of the anti-apoptotic proteins, BCL-2 and MCL-1. However, the present study had certain limitations, particularly in the testing of pure mannotriose compounds isolated from S. sennicompostii GMY01. The isolation process for these compounds is protracted and requires a substantial quantity of samples. Future studies are thus required to focus on optimizing the production of mannotriose from S. sennicomposti GMY01 and evaluating its potential as an anticancer agent.

Acknowledgements

The authors would like to thank Miss Mosa Rini Nurul Hidayati at the Cell Culture Laboratory and Farid Abdullah at the Clinical Pathology Laboratory, Faculty of Medicine, Public Health, and Nursing, Universitas Gadjah Mada, Yogyakarta, Indonesia for technical assistance with the anticancer assay.

Funding

Funding: The present study was funded by the Indonesian Ministry of Research, Technology and Higher Education, Basic Research Funding Scheme (grant no. 2128/UN.1.P.III/DIT-LIT/PT/2021).

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

MP, ED, WRP and M conceived and designed the study. MP, ED, JW and M were involved in data acquisition. MP and ED were involved in data analysis and interpretation. MP, ED and M were involved in the drafting of the manuscript. MP, ED, WRP and M critically revised the manuscript. MP performed the statistical analysis. M was involved in the provision of funding. ED, JW and M provided administrative, technical and material support. ED, WRP and M supervised the study. All authors confirm the authenticity of all the raw data All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References