Identification of Aloperine as an anti-apoptotic Bcl2 protein inhibitor in glioma cells

Data acquisition and analysis using different bioinformatics methods

Bioinformatics analysis for drug target discovery was performed using several public web-based tools, which are summarized in Table S1.

The bioinformatics analysis tool for molecular mechanism of traditional Chinese medicine (BATMAN-TCM) is a user-friendly online tool to evaluate the functions of TCM in human pathophysiological processes (Liu et al., 2016). PubChem is a chemical information resource for biomedical research (Kim et al., 2019). Using these two algorithms, we screened possible molecular targets of the natural compound ALO. After PPI network of these targets was constructed by STRING v11 (Szklarczyk et al., 2019), we performed GO and KEGG pathway enrichment analysis by WEB-based Gene Set Analysis Toolkit (WebGestalt) (Wang et al., 2017) and FunRich (Pathan et al., 2015), respectively. Next, molecular docking was conducted by systemsDock (Hsin et al., 2016) to evaluate interactions between ALO and its predicted targets. Protein sequence alignment of Bcl2 Arg63–Ala108 was conducted in Clustal Omega to identify sequence divergence.

Cell culture

Glioma cell lines, including T98G, U87, U118, SK-N-AS, U251 and Hs683, were obtained from the Cancer Research Institute, Central South University, China. Glioma cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Invitrogen, Carlsbad, CA, USA) with 10% fetal bovine serum (Gibco, Invitrogen, Carlsbad, CA, USA) and 100 U/ml penicillin-streptomycin (Gibco, Invitrogen, Carlsbad, CA, USA) at 37 °C and 5% CO₂.

Reagents and plasmids

Aloperine was purchased from Selleck Chemicals (S2420). The Bcl2 inhibitor ABT199 was purchased from MedChemExpress (HY-15531). All reagents were dissolved in dimethylsulfoxide (Amresco, Solon, OH, USA). Bcl2 over-expression plasmid was purchased from MiaoLingBio (P8816).

RNA extraction, cDNA synthesis and real-time PCR

Total RNA from renal tissues was isolated using Trizol reagent (Invitrogen, San Diego, CA, USA). Total RNA was then reverse-transcribed to cDNA using the PrimeScript™ strand cDNA synthesis kit (6210; Takara, Dalian, China). Densitometry was analyzed using 200TM-Image software (Bio-Rad, Hercules, CA, USA), and actin served as an internal reference gene. Forward and reverse primer sequences were as follows: Actin: 5′-CATGTACGTTGCTATCCAGGC-3′ and 5′-CTCCTTAATGTCACGCACGAT-3′; Bcl2: 5′-GGTGGGGTCATGTGTGTGG-3′ and 5′-CGGTTCAGGTACTCAGTCATCC-3′. Expression levels were evaluated using iTaqTM Universal SYBR green Supermix (Bio-Rad, Hercules, CA, USA) under the following conditions: 95 °C for 5 min, 40 cycles of 95 °C for 15 s, 56 °C for 30 s. Data analysis was conducted using the 2^−ΔΔCt method.

Western blot

Briefly, approximately 5 × 10⁶ glioma cells were washed in PBS and lysed in lysis buffer (Thermo Scientific, Waltham, MA, USA). Then, lysates were incubated on ice for 30 min and centrifuged at 10,000×g for 30 min at 4 °C. Protein concentrations were calculated by a BCA kit (Thermo Scientific, Waltham, MA, USA). After sample separation by 15% SDS-PAGE, proteins were transferred onto NC membranes (HATF00010; Burlington, MA, Millipore). Membranes were incubated overnight at 4 °C for 1 h after blocking with primary antibodies as follows: Bcl2 (1:1,000, #2872; Cell Signaling Technology, Danvers, MA, USA) and tubulin (1:2,000; Proteintech, Chicago, IL, USA). Then, membranes were probed with secondary antibodies (1:10,000; Proteintech, Chicago, IL,USA). Bands were detected using Immobilon western chemiluminescent reagents (WBKLS0500; Millipore, Burlington, MA, USA).

Apoptosis analysis

Apoptosis assays were conducted using the Annexin V-fluorescein isothiocyanate apoptosis kit (BD Biosciences, San Jose, CA, USA) according to the manufacturer’s protocols. The results were analyzed with a Dxp Athena™ flow cytometer (Cytek, Fremont, CA, USA).

Statistical analysis

All data in this study were presented as means ± standard deviations and experiments mentioned above were repeated at least three times. The significance of difference among the means was calculated using t-tests for two groups using SPSS 19.0 software. Statistical significance was defined as p < 0.05 in all tests.

Prediction of potential ALO targets using online bioinformatics analysis tools

Two public bioinformatics databases, BATMAN-TCM and PubChem Project, were used to identify potential targets of the natural compound ALO. As shown in Table 1, we observed that Fas, Casp3, anti-apoptosis protein Bcl2, CXCR4 and SLC6A2 were potential molecular targets of ALO. To assess the involvement of these targets in the biological behaviors of human biology, PPI for these five genes was conducted by a frequently used algorithm, STRING v11 (Fig. 1A). Furthermore, the WebGestalt database was used for analyzing GO enrichment of these five genes. Interestingly, several well-known processes involved in human pathology and physiology were identified in our top-ranked GO processes, including cell communication and response to stimulus (ontology: biological process), membrane and protein-containing complex (ontology: cellular component), protein binding (ontology: molecular function) (Figs. 1B–1D). Significant pathways of these targeted genes were compared using KEGG analysis to further illuminate their function. KEGG pathway analyses by FunRich proteomics software revealed that apoptosis was obviously enriched among identified biological signaling pathways (p-value < 0.01 was recommended as the cut-off) (Fig. 1E). Next, according to data from BATMAN-TCM, we evaluated the roles of ALO in human disease. Expectedly, ALO showed significant therapeutic effects on several human cancer and other pathological processes, such as late-stage solid tumor (Fig. 2; Table 2). Taken together, these findings suggest that agent ALO might exert anti-tumor effects by inducing cell apoptosis.

Molecular docking identifies Bcl2 as a target of ALO

By analyzing the macromolecular structure archive from the RCSB Protein Data Bank (PDB) (Burley et al., 2017), we confirmed preferred 3D protein structures of ALO target molecules (Fas: 3HHD, Casp3: 2DKO and Bcl2: 2W3L) in homo species. However, no 3D shape of CXCR4 and SLC6A2 proteins could be reliably generated in the PDB database. Then, using the molecular docking platform systemsDock, we detected protein-ligand binding profiles between ALO and these potential targets. As shown in Figs. 3A and 3B, we found that ALO could be well docked into the 3D protein structure of Bcl2, with the highest docking score values of 5.960 pKd/pKi. Next, we evaluated the possible activity pocket for ALO within the Bcl2 protein. The ligand-receptor binding pattern revealed that the native ligand could be docked into the Phe63–Val192 domain of Bcl2 in homo species, indicating the Phe63–Val192 domain is the active binding pocket of Bcl2 protein. Moreover, we found that ALO was also docked into the consistent pocket as expected, which includes amino acids Phe63, Tyr67, Phe71, Met74, Val92, Leu96 and Ala108 (Figs. 3C–3F). In addition, the phylogenetic analysis using the Clustal Omega tool (Madeira et al., 2019) revealed that the Bcl2 is highly conserved from homo sapiens to other species, with several conservative amino acid sites, particularly Val92–Ala108 in homo sapiens, suggesting its important function in biological processes (Fig. 3G). Thus, these data indicate good affinity of ALO for the Bcl2 protein that occurs by forming intermolecular forces between their polar moieties and conservative amino acid residues.

ALO induces apoptosis via Bcl2 inhibition in glioma cells

Glioma, a highly invasive and fatal cancer, is one of the most highly malignant human solid tumors, accounting for approximately 80% of all primary brain tumors (Chen et al., 2019). Studies have shown that natural compounds provide a large source of bioactive compounds with excellent therapeutic efficacy and have been investigated for their anticancer properties (Xu et al., 2018). After determining the relationship between ALO and solid tumors, we further investigated the effect of ALO on the inhibition of Bcl2 in glioma cells. Protein levels of Bcl2 were obviously higher in U118 and SK-N-AS compared to T98G, U87, U251 and Hs683 glioma cells (Fig. 4A). We then assessed the effect of ALO on Bcl2 compared to the Bcl2 inhibitor ABT199. After treatment with ALO and ABT199, we observed that protein levels of Bcl2 were obviously reduced in both groups, while mRNA expression in the ALO group was not significantly reduced in U118 and SK-N-AS cells (Figs. 4B and 4C). Using a concentration gradient, expression of Bcl2 consistently declined in response to treatment with 0.5 and one mM ALO or 50 and 250 μM ABT199 in glioma cells (Figs. 4D and 4E). Pathways regulated by Bcl2 play key roles in apoptosis, including controlling mitochondrial membrane permeability, forming feedback loop systems with caspases and attenuating inflammation (Kontos, Christodoulou & Scorilas, 2014). To investigate whether Bcl2 inhibition caused by ALO is associated with cell apoptosis in glioma cells, we examined the apoptosis rate in response to ALO and ABT199 treatment. Flow analysis indicated consistently induced apoptosis in U118 and SK-N-AS cells, further supporting the phenotype of Bcl2 inhibition (Fig. 5). To further confirm the targeting relationship between ALO and Bcl2, the effects of ALO and Bcl2 over-expression plasmid single or combined treatment were analyzed in glioma cells. Western blot experiments showed ALO-induced Bcl2 inhibition was significantly decreased in combination treatment (Fig. 6A). Additionally, over-expressed Bcl2 suppressed the pro-apoptosis effect of ALO in glioma (Figs. 6B–6K).