Establishment of lentivirus-mediated G-cleave LC3B biosensor in MDA-MB-231 breast cancer cells

Autophagy is a cellular mechanism with dual roles that regulate tumor initiation, progression, and chemoresistance [32]. It promotes the therapeutic resistance of cancer cells to toxic stress or activates cell death mechanisms to increase treatment efficiency in tumor cells [33]. Several methods and pathways have been reported for detecting autophagy in cancer cells [19]. Still, a responsive and rapid autophagy detection with high-throughput potential among live-cell detection remains to be established. To effectively monitor autophagy induction in cells, we modified the previous NIADS sensor [25, 26, 28] and inserted three repeats of the autophagy cleavage sequence (3X TFGMKLS) between pepA-N’luc and pepB-C’luc fusion proteins, with the repeats of glycine serving as a linker [29, 34]. Next, to utilize this G-cleave LC3B autophagy biosensor (LC3BpepABLuc) for cell normalization, we fused the enhanced green fluorescent protein (EGFP) gene with a self-cleaving peptide P2A/T2A at the amino termini of pepABLC3B (pEGFP-LC3BpepABLuc, Fig. 1A).

Fig. 1

Establishment of G-cleave LC3B biosensor on MDA-MB-231 breast cancer cells. A Schematic diagram of the pEGFP-LC3BpepABLuc construct expressing EGFP- pepABLuc LC3B cleavage peptide. B Lentivirus-mediated pEGFP-LC3BpepABLuc expression in MDA-MB-231 cells was visualized by fluorescence microscope. Scale bar: 50 μm. C MDA-MB-231 cells with pEGFP-LC3BpepABLuc expression were collected, and EGFP and luciferase protein expressions were determined. D IVIS images show the luciferase activity of MDA-MB-231 cells with or without lentivirus-mediated expression of pEGFP-LC3BpepABLuc. Quantification of total photons flux was analyzed (n = 3 replicates; student-t test; ****, p < 0.0001; bars represent mean ± SE). E Schematic representation of the EGFP and luciferase expression of the G-cleave LC3B biosensor. Upon expression of the fusion peptide EGFP-P2A-pepABLC3B in cells, self-cleavage at the P2A/T2A domain releases EGFP from the subsequent pepABLC3B peptide. During autophagy, autophagy signaling cleaves the specific autophagy cleavage sequences (3X TFGMKLSV) and enables the formation of cleaved pepA-N’luc2 and pepB-C’luc2 due to the strong interaction of pepA and pepB fragments. The luciferin addition is, therefore, able to determine autophagy activity (photons influx) through IVIS. F Cell lysates from pEGFP-LC3BpepABLuc expression cells were collected with or without 30 μM CQ. The expressions of EGFP, full-length luciferase, LC3B lipidation, and autophagic degradation were determined by immunoblot. β-actin was used as a loading control

After introducing the pEGFP-LC3BpepABLuc sensor into MDA-MB-231 breast cancer cells via a lentiviral delivery system, strong EGFP signals were observed under fluorescence microscopy (Fig. 1B). The expression of EGFP and luciferase proteins was further validated through immunoblot analysis (Fig. 1C), confirming EGFP as a transfection marker of the pEGFP-LC3BpepABLuc sensor in MDA-MB-231 cells, alongside the expression of the luciferase fusion protein. To explore the broader applicability of the pEGFP-LC3BpepABLuc sensor, we also transfected it into MDA-MB-468 and MDA-MB-468 TNBC cells. The expression of both EGFP and luciferase proteins was similarly validated in these cell lines (Supplementary Fig. 1).

We also assessed the enzyme activity of luciferase by adding luciferin as the substrate in MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc. The IVIS detection showed strong and significant luciferase activity in MDA-MB-231 cells transfected with the pEGFP-LC3BpepABLuc sensor (p≦0.0001), compared to the parental MDA-MB-231 cells (Fig. 1D). Based on the above result, we displayed a brief schematic model of how this G-cleave LC3B biosensor functions in MDA-MB-231 cells (Fig. 1E). First, EGFP is released from the pepABLC3B fusion protein via cells self-cleaving at P2A/T2A. Second, autophagy signaling cleaves the specific autophagy cleavage sequences and enables the formation of cleaved (N’luc2 and C’luc2) due to high protein–protein interaction affinity (pepA and pepB). Finally, luciferin exposure activates the luciferase protein, and bioluminescence represents the autophagy LC3B cleavage activity during autophagy events.

It is well known that the LC3B lipidation and autophagy cargo protein (SQSTM1) degradation are commonly used to assess autophagy events [35]. To understand whether the pEGFP-LC3BpepABLuc sensor might activate autophagy, we measured the LC3B lipidation and cargo protein degradation of pEGFP-LC3BpepABLuc in MDA-MB-231 and parental cells treated with/without the autophagy degradation inhibitor CQ (Fig. 1F). Immunoblot analysis revealed that LC3-II expression were increased in both MDA-MB-231 (p≦0.01) and cells expressing pEGFP-LC3BpepABLuc (p≦0.01) by CQ, while SQSTM1 accumulation occurred in cells expressing pEGFP-LC3BpepABLuc in the presence (p≦0.05) and absence (p≦0.05) of CQ, suggesting the expression of pEGFP-LC3BpepABLuc did not alter the levels of LC3B lipidation or SQSTM1 degradation in MDA-MB-231 cells compared to the parental cells, indicating that this pEGFP-LC3BpepABLuc biosensor construct does not interfere with endogenous autophagy process. This finding also suggests that the pEGFP-LC3BpepABLuc (G-cleave LC3B biosensor) in MDA-MB-231 breast cancer cells is ideal for exploring autophagy activation without intruding into endogenous autophagy activity.

Activation of autophagy in MDA-MB-231 breast cancer cells expressing G-cleave LC3B biosensor

Autophagy initiation is a multi-step process comprising nucleation, elongation, fusion, and degradation. To elucidate the duration and efficiency of autophagy detection using the G-cleave LC3B biosensor, we evaluated the activation of autophagy in MDA-MB-231 cells expressing the biosensor by nutrient depletion (Fig. 2A). The levels of LC3B lipidation and SQSTM1 degradation were evaluated under EBSS (Fig. 2A, upper panel) or serum starvation (Fig. 2A, lower panel) conditions, both of which led to a time-dependent increase (p≦0.05 and p≦0.001) in LC3B-II (lipidated LC3B) expression. Correspondingly, the time-dependent degradation of SQSTM1 occurred, with notable compensation in autophagic degradation 48 h after serum starvation, indicating intensified cycles of autophagic degradation. To ensure that the observed increase in LC3B-II was not due to inhibition of autophagy degradation, we co-treated the cells with EBSS/nutrient depletion and CQ. Upon adding CQ, LC3B lipidation induced by either EBSS (Fig. 2B, upper panel) or serum starvation (Fig. 2B, lower panel) was further enhanced (p≦0.05 and p≦0.01), while the EBSS-mediated SQSTM1 degradation was slightly inhibited (p≦0.05). These findings suggest that EBSS/nutrient depletion strategies can be effectively used to explore the G-cleave LC3B biosensor in MDA-MB-231 cells.

Fig. 2

EBSS and serum starvation induce autophagy in MDA-MB-231 cells expressing the G-cleave LC3 biosensor. A MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc were subjected to either Earle balanced salt solution (EBSS) or serum starvation for varying durations as indicated. B MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc were subjected to EBSS or serum starvation in the presence or absence of 30 μM CQ for 24 h. Cell lysates were collected and subjected to immunoblotting analysis to assess the levels of autophagic lipidation (LC3B) and degradation (SQSTM1). GAPDH and β-actin were used as loading controls. C MDA-MB-231 cells expressing pmCherry-EGFP-LC3B were treated with EBSS or serum starvation medium in the presence or absence of 30 μM CQ for 24 h to determine autophagic flux. The formation of EGFP (green), mCherry (red), and merge (yellow/orange) puncta was observed by the confocal microscopy. Scale bar: 10 μm. D Quantitate analysis of mCherry (red), EGFP (green), and merged (yellow/orange) puncta per cell are represented as means ± SEM in 20 to 30 cells in three independent experiments

To monitor the autophagic flux by visualizing the autophagosomes fusion with lysosomes to form autophagolysosomes, we observed the fluorescent punctate signals in MDA-MB-231 cells transduced with mCherry-EGFP-LC3B. The formation of mCherry (red) and EGFP (green) puncta was visualized in the structure of autophagosomes, shown as yellow/orange puncta in the merged images. Due to the pH sensitivity of EGFP fluorescence (autophagosome formation), which is quenched in the acidic environment of autophagolysosomes, red puncta appear in merged images, representing the lysosome-autophagosome fusion stage. Fluorescence microscopy showed that nutrient depletion increased the number of green puncta, indicating autophagosome formation (Fig. 2C). This was followed by an increase in red puncta, representing autophagosome-lysosome fusion, after EBSS or serum starvation treatment. Furthermore, yellow/orange puncta accumulated, particularly in cells treated with CQ, compared to controls (p≦0.05, Fig. 2D), indicating that EBSS/nutrient depletion effectively activated autophagic flux in MDA-MB-231 cells. These results demonstrate that nutrient depletion enhances LC3B lipidation, SQSTM1 degradation, and autophagic flux, all of which can be observed in MDA-MB-231 cells or cells expressing the G-cleave LC3B biosensor using traditional immunoblotting and fluorescence-based autophagy assays.

EBSS/nutrient depletion stimulates luciferase activity of G-cleave LC3B biosensor in a proteasome degradation-dependent manner

To investigate whether nutrient depletion efficiently activates the G-cleave LC3B biosensor, we employed a bioluminescence-based assay to monitor the luciferase degradation activity of the biosensor in MDA-MB-231 cells. We revealed that EBSS and serum starvation significantly decreased the bioluminescence activity in MDA-MB-231 cells (p≦0.01, Fig. 3A), which refers to the previous finding that autophagy elevated LC3B lipidation and cargo protein degradation. The maximum luciferase degradation activity of the G-cleave LC3B biosensor occurred four hours after EBSS or serum starvation, with no further increase after prolonged treatment (Supplementary Fig. 2). We further assessed the applicability of pEGFP-LC3BpepABLuc biosensor in MDA-MB-453 and MDA-MB-468 TNBC cells, where EBSS and serum starvation also significantly decreased bioluminescence activity (p≦0.01, Supplementary Fig. 3A, B), with autophagy activity peaking at four hours post-treatment (p≦0.01, Supplementary Fig. 3C, D). Additionally, the G-cleave LC3B biosensor exhibited a titratable response in a dose-dependent manner with higher EBSS exposure (p≦0.01, Fig. 3B), demonstrating the biosensor's sensitivity to varying degrees of autophagy induction. In contrast, serum depletion resulted in a mild time-dependent increase in autophagy activity over 1–4 h (p≦0.01, Fig. 3C) we found that FBS concentrations between 1 and 5% did not produce significant changes in bioluminescence, suggesting that low FBS levels may sufficiently support cell survival and prevent autophagy activation. It is speculated that to surpass the threshold for autophagy induction, FBS concentrations should be reduced to less than 1% in cell culture conditions. Importantly, the biosensor showed no response in cells cultured with 10% serum within four hours, confirming that the observed signal loss was due to autophagy-specific cleavage rather than nonspecific degradation (Supplementary Fig. 4).

Fig. 3

Enhanced autophagic flux and G-cleave LC3B biosensor luciferase activity in MDA-MB-231 cells through proteosome-related degradation. A IVIS images show the decline of bioluminescence in MDA-MB-231 cells expressed pEGFP-LC3BpepABLuc with 4 h of EBSS and serum starvation treatments. Quantification of total photons flux was analyzed (n = 3 replicates; student-t test; **, p < 0.01; bars represent mean ± SE). MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc were exposed to varying concentrations of B EBSS and C serum starvation for 1 to 4 h, or pre-treatment with or without 10 μM MG132 followed by indicated incubation with D EBSS and E serum starvation. The luciferase degradation activity (convert to autophagy activity) was assessed as described in materials and methods (n = 3 replicates; student-t test; *, p < 0.05; **, p < 0.01; error bars represent mean ± SE). The lysates of 4 h F EBSS-treated and G serum starvation-treated were also collected and immunoblotted with anti-luciferase antibody. The full-length luciferase appeared at around 63 KDa, while the cleavage form was around 48 KDa on SDS-PAGE. H The level of cleaved luciferase and autophagy degradation was evaluated by immunoblotting in MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc, treated with EBSS in the presence or absence of 30 μM CQ for 24 h. GAPDH was used as a loading control. I A schematic diagram of the G-cleave LC3B biosensor is presented. The fusion protein of pepA-N’luc2 and pepB-C’luc2 were linked with a 3X LC3B cleavage sequence (TFGMKLS). The bioluminescence expression of the cleavage form of luciferase mediated by ATG4B was detected upon the addition of luciferin. This bioluminescence is predominantly degraded in a proteosome-dependent manner during the activation of autophagy. The luciferase activity reflects the degradation ratio of luciferase upon complete autophagy activation

The proteolytic processing of LC3 by the ATG4 cysteine protease is essential for the initiation of autophagy membrane conjugation [36] and the ubiquitin-mediated autophagic degradation [37]. To determine whether the autophagy-triggered reduction in G-cleave LC3B bioluminescence in MDA-MB-231 cells requires a protease degradation process, we exposed cells to MG132, a potent proteasome inhibitor to examine the bioluminescence changes upon EBSS/serum starvation treatments. We found that the high autophagy activity of EBSS/serum starvation treatments was significantly reduced during MG132 exposure (p≦0.01 and p≦0.05), whereas MG132 treatment only maintained low autophagy activity (Fig. 3D, E). Since it is known that the UPS is the primary mechanism that degrades the short-lived protein [37], we proposed that luciferase, a cleaved form of LC3BpepABLuc, is degraded through a proteasome-associated pathway upon autophagy activation. Indeed, immunoblot results showed that both EBSS-(Fig. 3F) and serum starvation-(Fig. 3G) mediated luciferase expression can be enhanced (p≦0.05) in cells pre-treated with MG132. To investigate whether autophagy degradation responsible for the long-lived protein is involved in this luciferase degradation process, we treated cells with the autophagy lysosome-fusion inhibitor chloroquine (CQ) for 24 h. Unlike the effects of MG132, CQ did not attenuate the degradation of luciferase in response to EBSS or serum starvation (Fig. 3H). However, long-term luminescence degradation was partially reduced when CQ inhibited autophagy activity (p≦0.05, Supplementary Fig. 5A), indicating that the lysosome-mediated long-term degradation pathway is not the primary mechanism driving the G-cleave LC3B biosensor activity (Fig. 3I).

We further evaluated the biosensor response to non-canonical autophagy modulators, specifically Monensin (autophagy activator) and Bafilomycin A1 (autophagy inhibitor), which are involved in the CASM pathway [38, 39]. The G-cleave LC3B biosensor was found to be highly sensitive to canonical autophagy induction by EBSS (Fig. 3B and Supplementary Fig. 5A), but showed minimal response to non-canonical autophagy modulators, including Monensin (Supplementary Fig. 5B), CQ (Supplementary Fig. 5C) and Bafilomycin A1 (Supplementary Fig. 5D). Unlike nutrient depletion, which markedly increased luciferase degradation, non-canonical autophagy stimulation did not lead to a decrease in luciferase degradation, as confirmed by immunoblotting (Supplementary Fig. 6). The quantified immunoblot results indicated that the expression level of either full length or cleaved forms of luciferase modulated by Monensin were not significantly altered by bafliomycin A1 exposure (Supplementary Fig. 12).

Taken together, the above results provide evidence that autophagic stimuli such as EBSS and serum starvation trigger the rapid degradation of short-lived cleaved-luciferase and increase the autophagy activity of the G-cleave LC3B biosensor in MDA-MB-231 cells primarily via proteasome-related degradation mechanisms. The sensor's selectivity is driven by its responsiveness to short-lived, canonical autophagic degradation rather than long-lived non-canonical autophagy. This is evident as non-canonical autophagy stimuli do not enhance cleaved-luciferase degradation or LC3 lipidation, even in the presence of autophagy inhibitors (Fig. 3I & Supplementary Fig. 6).

LC3B conjugating enzyme ATG4B is required for nutrient depletion-mediated activation of autophagy activity in G-cleave LC3B biosensors

The post-translational modification of LC3B is an essential step in autophagy, in which the C-terminus of the soluble LC3-I is cleaved by ATG4B, an autophagy-related cysteine protease, to expose the C-terminal glycine residue required for the formation of the membrane-bound LC3-II [40]. The G-cleave LC3B biosensor used in this study was designed to incorporate an autophagy cleavage sequence that contains the critical C-terminal glycine residue needed for LC3B-PE conjugation [29, 34]. To investigate the role of ATG4B in regulating the luciferase degradation (autophagy activity) of G-cleave LC3B biosensor, we used lentivirus-mediated CRISPR-Cas9 gene editing to establish ATG4B gene-edited MDA-MB-231 cells with pEGFP-LC3BpepABLuc sensor (Fig. 4A). Through Tracking of Indels by DEcomposition (TIDE) analysis, we noticed that the average indels rate was 91.2% in ATG4B gene edited pool MDA-MB-231 cells with pEGFP-LC3BpepABLuc sensor (Fig. 4B), whereas the most abundant indel was + 1 nucleotide insertion with 42.9% of all population (Fig. 4C). On the other hand, ATG4B protein expression was completely abolished in the ATG4B gene-edited pool MDA-MB-231 cells with pEGFP-LC3BpepABLuc sensor, compared with SC parental cells (Fig. 4D). In addition, we found that the EBSS- and serum starvation- triggered autophagy activity of the G-cleave LC3B biosensor was significantly inhibited in ATG4B gene-edited cells (p≦0.01 and p≦0.05, Fig. 4E), indicating that the autophagic lipidation enzyme ATG4B is essential for initiating autophagy process and downstream proteasome-related degradation of G-cleave LC3B biosensor during autophagic stimulation.

Fig. 4

The specific autophagic roles of ATG4B and RSV sensitized the autophagy process of the G-cleave LC3B biosensor in MDA-MB-231. A The schematic of the CRISPR/Cas9 sgRNA targeting sequence located in the human ATG4B chromosome. TIDE analysis displayed the B gene editing efficiency and C gene indel spectrum of ATG4B DNA locus in MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc. D Immunoblotting verified the depletion of ATG4B protein expression. E MDA-MB-231 cells expressing the pEGFP-LC3BpepABLuc were transduced with lentivirus-mediated CRISPR/Cas9 targeting ATG4B or scrambled control (SC) and then treated with either EBSS or serum starvation for 4 h. The luciferase degradation activity (convert to autophagy activity) of the G-cleave LC3B biosensor was assessed. F MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc were treated with or without 100 μM RSV for 4 h. IVIS images were obtained, and quantification of total bioluminescence photon flux was analyzed. G MDA-MB-231 cells expressing pEGFP-LC3BpepABLuc were treated with RSV at indicated concentrations for 4 h. Luciferase degradation activity (convert to autophagy activity) was determined as described in the materials and methods (n = 3 replicates; student-t test; *, p < 0.01; **, p < 0.01; error bars represent mean ± SE). H MDA-MB-231 cells were treated with 1 to 100 μM RSV for 24 h, with or without 30 μM CQ co-treatment. Cell lysates were collected, and the accumulation of LC3B-II and SQSTM1 was examined by immunoblotting. β-actin was used as a loading control. I MDA-MB-231 cells with pmcherry-EGFP-LC3B expression were treated with 25 μM RSV with or without 30 μM CQ for 24 h. Autophagic flux was determined by observing the formation of EGFP (green), mcherry (red), and merge (yellow/orange) puncta under the confocal microscopy. Scale bar: 10 μm. J Quantitate analysis of mCherry (red), EGFP (green), and merged (yellow/orange) puncta per cell are represented as means ± SEM in 20 to 30 cells in three independent experiments

Screening of autophagy-modulating drugs by G-cleave LC3B biosensor in MDAMD231 breast cancer cells

To identify potent drugs that regulate autophagy in breast cancer, we performed a drug screening assay using the G-cleave LC3B biosensor, including clinical anti-breast cancer drugs, autophagy modulators, and flavonoids. The drug details used to validate the luciferase degradation activity (autophagy activity) of the biosensor in MDA-MB-231 cells are listed in Supplementary Table 2. We found that EBSS and serum starvation increased the autophagy activity of the biosensor by 2.82 ± 0.48 and 2.04 ± 0.19 folds, respectively (Supplementary Table 1). The autophagy-inducing peptide Tat-Beclin1 L11, which specifically activates autophagy via interaction with the autophagy suppressor GAPR-1/GLIPR2 [41], increased the autophagy activity by 3.89 folds (Supplementary Table 2). The autophagy inhibitors CQ and 3-MA did not alter the autophagy activity of the biosensor in MDA-MB-231 cells. Additionally, we found that Resveratrol (RSV), a natural anti-tumor phenol isolated from grapes, achieved the highest autophagy activity (4.31 folds) compared to other drugs as determined by the G-cleave LC3B biosensor in MDA-MB-231 cells. The IVIS assay also demonstrated significant biosensor degradation after RSV treatment (p≦0.0001, Fig. 4F).

Based on the drug pre-screening results (Supplementary Table 2), we found that RSV, a natural phenol known for its anti-tumoral properties [42], increased the autophagy activity of MDA-MD-231cells much more than other drugs. It was evident that RSV treatment at 10 to 100 µM increased autophagy activity dose-dependently (p≦0.01, Fig. 4G). Additionally, 24 h after RSV exposure, MDA-MB-231 maintained autophagy characteristics, such as LC3B lipidation and SQSTM1 degradation in cells with or without CQ (p≦0.05, Fig. 4H). RSV also promoted autophagy flux, as indicated by the formation of green (autophagosome) and red (autophagolysosome) puncta in pmCherry-EGFP-LC3B-transfected MDA-MB-231 cells (Fig. 4I). Additionally, CQ treatment led to enhanced accumulation of yellow/orange (autophagolysosome) puncta compared to RSV treatment alone (p≦0.05), as quantified from MDA-MB-231 cells (Fig. 4J). These findings highlight the effectiveness of the G-cleave LC3B biosensor in identifying promising autophagy-targeting drugs for potential TNBC treatments.

RSV enhances the drug-sensitivity of DOX in MDA-MB-231 breast cancer cells

The anti-tumoral role of autophagy has also been reported in breast cancer [43]. Dysregulation of autophagy caused by the loss of the autophagy-related gene BECN1 facilitates tumor formation and progression in TNBC. Despite numerous studies aimed at verifying potential cancer therapeutic drug combinations by inhibiting or stimulating autophagy, no authorized pharmaceuticals are currently designed to manipulate autophagy for addressing TNBC. To explore whether selected autophagy agents synergize with clinical chemotherapy agents, such as doxorubicin (DOX) to improve the anti-cancer effect on TNBC, we combined RSV and DOX to treat MDA-MB-231 cells. The cytotoxicity determination showed that the IC50 concentration of RSV and DOX were 90 µM (Supplementary Fig. 7A) and 2.0 µM (Supplementary Fig. 7B), respectively. Since the development of drug resistance in malignant breast tumors is frequently observed, the combination of DOX with other anti-neoplastic agents is therefore required in clinical settings [44, 45]. Hence, we evaluated the anti-cancer potential of RSV in combination with DOX to suppress breast cancer cell survival and growth.

RSV is a natural compound that exhibited health benefits wildly, including its anti-cancer properties. The effective concentration of resveratrol for the anti-tumor study can vary depending on the specific type of cancer, the experimental conditions, and the study design. Typically, the effective anti-cancer concentration of RSV ranges from 50 to 100 µM, depending on the specific cancer. Research shows that long-term RSV treatment, or its combination with other chemotherapeutic agents, enhances therapeutic outcomes by promoting apoptosis and inhibiting cancer cell proliferation. This is particularly true for aggressive cancers like TNBC, where RSV has demonstrated efficacy in reducing tumor growth and increasing treatment sensitivity [46, 47]. RSV has been shown to inhibit the mechanistic target of rapamycin complex 1 (mTORC1), which facilitates early autophagosome formation by promoting the interaction between unc-51 like autophagy activating kinase 1 (ULK1) and the class III phosphatidylinositol 3-kinase (PtdIns3K) complex [48,49,50]. By inhibiting mTORC1, RSV activates autophagy and prevents the overactivation of the PI3K/Akt/mTOR pathway, a known contributor to DOX chemotherapy resistance. This dual action ultimately enhances apoptosis [48, 49]. Therefore, RSV may promote DOX-induced apoptosis through mTOR-dependent autophagy activation, offering potential therapeutic benefits in combating aggressive breast cancer progression.

As shown previously, the IC50 concentrations of RSV and DOX were determined to be 90 µM and 2.0 µM. To evaluate the potential synergistic cytotoxicity of RSV and DOX in MDA-MB-231 cells, we used the Combination Index (CI) algorithm to categorize drug combinations as additive (CI = 1), synergistic (CI < 1) or antagonistic (CI > 1), respectively (Fig. 5A). Among these combinations, co-treatment with 19.03 µM RSV and 0.5 µM DOX (red circle) achieved the greatest cytotoxic effect (CI = 0.462), indicating a strong synergy. Other combinations, such as reducing RSV or increasing DOX concentrations, did not result in better CI values. To confirm this finding, we tested fixed concentrations of RSV (25 µM) with varying concentrations of DOX (0.5–25 µM). The results demonstrated that the co-treatments of 25 µM RSV with 0.5 µM or 1 µM DOX showed the most significant cytotoxicity in MDA-MB-231 cells (p≦0.05 and p≦0.01, Fig. 5B). Additionally, in a time-dependent analysis, the combination of 25 µM or 50 µM RSV with 0.5 µM DOX led to a more pronounced reduction in cell viability compared to 0.5 µM DOX alone after 48 h of treatment (p≦0.05, Fig. 5C). Even after 72 h, the combination of 50 µM RSV with 0.5 µM DOX maintained a significant inhibitory effect on cell viability compared to DOX treatment alone (p≦0.05).

Fig. 5

RSV potentiates DOX-induced cytotoxicity and apoptosis in MDA-MB-231 breast cancer cells. A The combination index (CI) of RSV and DOX was determined for 48 h on MDA-MB-231 cells by CCK-8 cell viability assay. An isobologram was generated to represent the synergistic/antagonistic effect of each combination. B MDA-MB-231 cells were treated with DOX (0.5 to 25 μM) with or without RSV (25 μM) for 48 h, and cell viability was measured by CCK-8 assay. C MDA-MB-231 cells were treated with a low concentration of DOX (0.5 μM) in combination with increasing concentrations of RSV (25 to 50 μM) for 24 to 72 h, and cell viability was measured by CCK-8 assay. (n = 3 replicates; student-t test; *, p < 0.05; **; error bars represent mean ± SE). MDA-MB-231 cells were treated with DOX (0.5 μM) with or without RSV (50 to 100 μM) for 48 h. D The activity of caspase 3 and PARP cleavage was examined by Immunoblot. β-actin was used as a loading control. E Flow cytometry determined the apoptosis cell population of the sub-G1 phase, and the percentage of sub-G1 cell populations are listed (n = 3 replicates; student-t test; *, p < 0.05; **; error bars represent mean ± SE)

Lastly, we confirmed the above anti-cancer finding with immunoblotting. It was evident that the combined treatment with 100 µM of RSV and 0.5 µM DOX significantly induced caspase 3 and PARP cleavages on MDA-MB-231 cells (p≦0.05 and p≦0.001, Fig. 5D). In addition, we used flow cytometry to confirm the above synergic apoptosis activity of RSV and DOX exposures. After the combination drug treatments for 2 days, 100 µM of RSV and 0.5 µM DOX obtained the most effective apoptosis event (sub-G1 phase) than DOX monotherapy and control group (p≦0.01, Fig. 5E). These findings indicate that RSV, the anti-cancer compound identified from the G-cleave LC3B biosensor cells, enhances the cytotoxic and apoptotic effects during DOX exposure to MDA-MB-231, implying this potential anti-cancer drug combination may be used in clinical breast cancer therapy.

The toxicity of DOX is known to decrease in acidic conditions (pH 6.3) due to reduced cell membrane permeability [50]. Additionally, an acidic extracellular environment can hinder the efficacy of anti-cancer drugs by inhibiting autophagy [51]. However, autophagy can help cancer cells survive in acidic environments. All conditions were maintained at pH 7.3 to pH 7.6 to minimize pH interference in our experiments-the impact of acidic tumor microenvironments on drug efficacy identified by our biosensor warrants further investigation. Our findings suggest the potential use of biomaterials to mitigate acidic environmental effects, enhancing autophagy-based therapies in breast cancer.

The advantage of G-cleave LC3B biosensor in monitoring of autophagy

The G-cleave LC3B biosensor offers several distinct advantages in the monitoring of autophagy. We summarized these advantages and compared them to traditional autophagy detection methods commonly employed in cancer research (Table 1). Firstly, conventional techniques such as immunoblotting and flow cytometry provide insights into the autophagy process by monitoring changes in specific autophagy-related proteins, allowing researchers to identify the stage of autophagy, whether it involves elongation, fusion, or protein degradation. However, these methods have significant drawbacks, including their time-consuming and expensive cost requirements, primarily due to the need for various antibodies. Secondly, imaging-based approaches, such as puncta formation and electron microscopy (EM), are frequently used to detect the autophagic flux and formation of the double-membrane autophagosome structure. While these techniques are more cost-effective than immunoblotting and flow cytometry, they demand a deep understanding of molecular cloning techniques, access to sophisticated fluorescence or electron microscopes, and, most importantly, the expertise of trained personnel to obtain reliable imaging results. Furthermore, the data obtained from these traditional methods, often image-based analysis, are either non-quantitative or semi-quantitative, limiting their utility for large-scale drug screening due to the substantial sample requirements. In contrast, our study introduces an innovative autophagy detection approach that is quantifiable through digital scale-based analysis, rapid and high-throughput. By culturing live cells that carry the G-cleave LC3B biosensor in multi-well plates, applying autophagy-inducing drugs less than 4 h, whereas researchers can obtain digitalized quantitative experimental data using a fluorescence/luminescence microplate reader within 30 min. The G-cleave LC3B biosensor minimizes sample requirement and provides high detection capacity in the multi-well plates, make it an efficient tool for autophagy drug screening. This method also holds great promise for advanced applications in precision medicine, enables the screening of autophagy/apoptosis-targeting drugs in clinical cancer patients, paving the way for more personalized and effective cancer treatments.

Table 1 The comparison between G-cleave LC3 biosensor and other routine autophagy detection methods