Anti-Proliferative and Pro-Apoptotic Activities of Hederacolchiside A1 On MCF-7 Cells Via ROS Regulation and JAK2/STAT3 Inactivation

Many studies have shown that hederacolchiside A1 (HA1) is an important anticancer saponin, although its mechanism of action and in vivo investigations are still lacking. Our previous results revealed that HA1 may have the potential to treat breast cancer. Therefore, we attempted to verify the potential anti-breast cancer effect of HA1 in vitro and in vivo. MTT, ow cytometry, DCFH-DA uorescence microscopy, and western blotting were used to evaluate the activities and mechanisms of action of HA1. Athymic nude mice were used to demonstrate the antitumor activity of HA1 in vivo. HA1 exhibited signicant cytotoxic effects on HepG2, MCF-7, MDA-MB-231, SKBr-3, HT-29, and HCT-116 cells, especially MCF-7 cells. HA1 blocked the sub-G1 and G0/G1 phases, induced apoptosis, promoted reactive oxygen species (ROS) generation, and decreased the mitochondrial membrane potential of MCF-7 cells. HA1 upregulated Bax and downregulated Bcl-2 levels and activated caspase-9 and caspase-3 in MCF-7 cells Meanwhile, HA1 inhibited the phosphorylation of JAK2/STAT3 in MCF-7 cells. In addition, 50 and 100 mg/kg HA1 signicantly inhibited the growth of transplanted tumors with inhibition rates of 46.95 ± 26.72% and 48.45 ± 22.36%, respectively. This preliminary study demonstrated that HA1 could inhibit proliferation and induce the apoptosis of MCF-7 cells via ROS-mediated activation of the mitochondrial apoptotic pathway and JAK2/STAT3 inactivation. HA1 may therefore be developed as a novel agent for breast cancer therapy. cells by inducing ROS generation and inactivating JAK2/STAT3 pathway.


Introduction
Breast cancer is the most common malignant tumor in women globally and is the main cause of cancerassociated deaths in women (Park et al. 2020). The anticancer active ingredients isolated from Traditional Chinese Medicine (TCM) have shown good e cacy after validation via extensive experiments and clinical studies (Wang et al. 2021). The signal transducer and activator of transcription 3 (STAT3) signaling pathway, which can be activated by Janus kinase (JAK), was found to be strongly related to numerous tumorigenesis processes, including cell proliferation, sensitivity to cytotoxic agents, apoptosis, and cell death (Sansone et al. 2012). The activation of the JAK2/STAT3 signaling pathway could promote tumor growth by affecting the expression of cell cycle regulatory and antiapoptotic genes, and STAT3 inhibitors may exhibit more promising effects on breast cancer therapy (Ma et al. 2020).
Oleanane-type saponins, such as -hederin, β-hederin, hederacolchiside A1 (HA1), and Pulsatilla saponin A/D, have received considerable attention because of their strong antitumor activity in vitro and in vivo Guan et al. 2020). HA1, as the main active ingredient, widely exists in the genera Anemone L. and Hedera L. These include Pulsatilla chinensis (Bunge) Regel, Anemone raddeana Regel, and Anemone accida Fr. Schmidt. HA1 and its derivatives reportedly exhibit strong antitumor activities against various cell lines in vitro, with IC 50 values of less than 10 µM Liang et al. 2016;Fang et al. 2016). Although the anticancer potential of HA1 has been extensively reported, investigations of its antitumor activity in vivo and the mechanism of action are lacking.
Our previous results showed that HA1 exhibited strong cytotoxicity in a variety of cancer cell lines with IC 50 values of 0.29-3.48 µM (Zhao et al. 2018). We also reported that TSS (the alkaline hydrolysis product of A. raddeana) showed excellent antitumor activity against breast cancer both in vitro and in vivo. HA1 is the main active ingredient and mainly contributes to the anticancer effect of TSS (Zhang et al. 2020). In addition, previous studies have reported that the growth and metastasis of 4T1 cell (mouse breast cancer cells) xenografts could be inhibited by 5 mg/kg HA1 after intraperitoneal injection (Liang et al. 2016). These data suggest that HA1 might play a crucial role in the proliferation and progression of breast cancer, although these results are still inconclusive. As a part of our continued investigation of HA1 (Zhang et al. 2017a;Zhang et al. 2020), this study rst attempted to summarize the anticancer effects of HA1 in vitro and in vivo. Next, we investigated the mechanism of the antiproliferation and proapoptotic activities of HA1 on MCF-7 cells. Moreover, the antitumor effect of HA1 was further con rmed in the MCF-7 xenograft tumor model.

Cell cycle detection
The cell cycle was detected as previously reported without any change (Zhang, 2020). After treatment with HA1 (4, 8, and 12 µM) for 24 h, the cells were digested, collected, centrifuged, washed, and xed overnight. After centrifugation and washing with phosphate-buffered saline (PBS), 100 μL RNase A was added to MCF-7 cells. Then, the cells were incubated at 37 °C for 30 min, and 5 μL PI was added and mixed. Finally, the DNA content was detected using the same ow cytometer that we used previously (Zhang et al. 2020).
Annexin V-FITC/PI double staining assay MCF-7 cells were inoculated in 6-well plates and exposed to different dosages of HA1 (4,8,and 12 µM) for 48 h. Then, following the manufacturer's instructions, cells (1 × 10 5 ) were collected, centrifuged, and washed with PBS. PI and Annexin V-FITC (5 µL each) staining were then performed room temperature for 15 min in the dark. Finally, the cells were analyzed using a ow cytometer.

ROS generation detection
Following the manufacturer's instructions, ROS generation in MCF-7 cells after treatment with HA1 was detected by uorescence staining using 2',7'-dihydro uorescein-diacetate (DCFH-DA). Cells were rst incubated with fresh medium containing 10 μM DCFH-DA for 20 min in the dark in 6-well plates. The cells were then exposed to different concentrations of HA1 (4, 8, and 12 µM) for 30 min in an incubator. Finally, after washing with PBS three times, the cells were photographed using a uorescence microscope (Olympus, Tokyo).
Mitochondrial membrane potential (MMP) detection MMP was detected using Rhodamine 123 dye, as previously described (Zhang et al. 2015). After treatment with HA1 for 24 h, the MCF-7 cells were washed twice with PBS. Then, the cells were stained with Rhodamine 123 (1 mM) and incubated for 30 min. Finally, cells with uorescence were photographed using the same uorescence microscope (Olympus, Tokyo, Japan).

Western blot analysis
Following a previously established protocol (Zhang et al. 2017b), MCF-7 cells were collected and lysed in RIPA buffer (containing 10% PMSF) after incubation with HA1 for 48 h. Total protein was quanti ed with the BCA protein assay kit after centrifugation according to the manufacturer's instructions. Then, 40 µg total protein was loaded onto 10% SDS-PAGE and separated by electrophoresis. After isolation, the proteins were transferred to PVDF membranes. The membranes were blocked with 5% blocking buffer and incubated with Bcl-2, Bax, caspase-3/9, cleaved caspase-3/9, JAK2/STAT3, and β-actin at appropriate dilutions overnight. Secondary antibodies were incubated for 2 h, and the membranes were washed three times the next day. The chemiluminescence method was used to detect the protein content following the manufacturer's instructions (Beyotime, Jiangsu, China).
In vivo antitumor activity BALB/c-nude female mice (16-20 g) were loaded subcutaneously with MCF-7 cells (0.1 mL of 5×10 6 cells/mouse) into the ank. After 6 days of tumorigenesis, 18 mice were randomly chosen and divided into three groups with six mice per group: the control group was treated with saline (10 mL/kg/mouse/d) and the HA1 groups were administered HA1 (50 and 100 mg/kg/mouse/d) for 28 days. After all mice were sacri ced, all tumors were obtained and measured. The tumor volume and inhibition rate (IR) of tumor weight were calculated.

Statistical analysis
All data are expressed as the mean ± standard deviation (SD) of three experiments using GraphPad Prism software. One-way analysis of variance (ANOVA) with Dunnett's post-hoc test was used to compare the signi cant differences (p < 0.05) between groups.

HA1 exhibited cytotoxic effects on several cancer cell lines
As summarized in Table S1, HA1 exhibited signi cant antitumor activities against dozens of cell lines in vitro, and the IC 50 values were less than 10 µM. In this study, we evaluated the cytotoxicity of HA1 using an MTT assay on six cell lines (HepG2, MCF-7, MDA-MB-231, SKBr-3, HT-29, and HCT-116). As shown in Table 1, HA1 exhibited signi cant inhibitory activity against these cells (IC 50 < 10 µM). HA1 exhibited the strongest antiproliferative effect on MCF-7 cells, with IC 50 values of 4.90 ± 0.50 µM. The structure of HA1 is shown in Fig. 1A. As shown in Fig. 1B, HA1 signi cantly inhibited the cell viability of MCF-7 cells when the dose was higher than 2.50 µM. The effect was di cult to observe when the HA1 dose was less than 1.25 µM; however, the cytotoxic activity was too high when the dosage was higher than 15 µM. We concluded that the effective dose range for HA1 was narrow. After constant exploration, HA1 dosages of 4, 8, and 12 µM were chosen for the follow-up experiments in vitro.

HA1 induced cell cycle arrest in MCF-7 cells
The cell content in the sub-G1 and G0/G1 phases increased signi cantly when the dosage of HA1 was 12 µM, while the S phase tended to decline. Meanwhile, HA1 induced the arrest of the sub-G1 phase in MCF-7 cells after 24 h of treatment ( Fig. 2A). These ndings suggest that HA1 signi cantly blocked the cell cycle at the sub-G1 and G0/G1 phases, thus affecting DNA replication and inhibiting MCF-7 cell proliferation (Fig. 2B).
HA1 promoted ROS production and reduced MMP in MCF-7 cells HA1 (8 and 12 µM) could signi cantly increase the levels of intracellular ROS in MCF-7 cells from 101.23 ± 2.05% to 207.93 ± 21.75% and 264.32 ± 32.58%, respectively, after 30 min incubation. However, NAC signi cantly decreased the ROS production to 103.27 ± 23.76% ( Fig. 4A and C). Considering that the proapoptotic activity was mediated by ROS, excessive ROS might have a large impact on the integrity of MMP. We detected the changes in MMP of MCF-7 cells after treatment with HA1 (8 and 12 µM) for 24 h.

HA1 induced MCF-7 cell apoptosis through the mitochondrial pathway
The results of western blot analysis revealed that HA1 (4, 8, and 12 µM) caused a signi cant decrease in the levels of caspase-3/9 and a remarkable increase of cleaved caspase-3/9 after 48 h of treatment. HA1 also caused a signi cant decline in Bcl-2 levels and increased Bax levels (Fig. 5A), which caused a signi cant increase in the ratio of Bax/Bcl-2 ratio after HA1 treatment (Fig. 5B). These results preliminarily con rmed that HA1 induced apoptosis in MCF-7 cells by activating the mitochondrial apoptotic pathway.

HA1 inactivated the JAK2/STAT3 signaling pathway
After treating cells with HA1 (4, 8, and 12 µM) for 48 h, the expression levels of JAK2, STAT3, p-JAK2, and p-STAT3 were determined. The levels of total JAK2 and STAT3 did not signi cantly change compared to the control; however, the levels of p-JAK2 and p-STAT3 were signi cantly reduced (Fig. 6A). HA1 (8 and 12 µM) signi cantly inhibited the phosphorylation ratio (Fig. 6B). These results suggest that HA1 could inhibit the activation of the JAK2/STAT3 pathway.

HA1 exhibit antitumor effects on mice
To further con rm the antitumor effect of HA1 in vivo, tumor-bearing nude mice were subjected to experiments. At the end of the experiments on day 28, HA1 (50 and 100 mg/kg) signi cantly inhibited tumor growth in these tumor-bearing nude mice (Fig. 7A). The volume (Fig. 7B) and weight of tumors (Fig. 7C) of the 50 and 100 mg/kg HA1-treated groups were signi cantly inhibited compared to those of the control group. The tumor volume decreased from 1,446.77 ± 216.72 mm 3 to 661.36 ± 120.65 mm 3 and 588.45 ± 191.32 mm 3 in the 50 and 100 mg/kg HA1 groups, respectively, and the tumor weight decreased to 43.36 ± 26.72% and 48.45 ± 22.36% of the control group.
These activities include cytotoxic, antiproliferative, antitumor, antimicrobial, antifungal, anthelmintic, and antileishmanial effects (Gumushan-Aktas et al. 2016;Wang et al. 2017). As shown in Table S1, previous studies have reported the cytotoxic activity of HA1 on numerous cell lines with IC 50 values lower than 10 µM. Our previous results showed that HA1 exhibited strong cytotoxicity in a variety of cell lines with IC 50 values of 0.29-3.48 µM (Zhao et al. 2018). HA1 has also been reported to exhibit signi cant inhibitory effects on breast cancer proliferation, invasion, and metastasis in vitro and in vivo (Liang et al. 2016;Wang et al. 2018). Similar to these results, this study further con rmed that HA1 exhibited a signi cant inhibitory effect on HepG2, MCF-7, MDA-MB-231, SKBr-3, HT-29, and HCT-116 cells and superior antiproliferative ability in MCF-7 cells.
As previously mentioned, cell cycle arrest is an effective measure to inhibit the abnormal proliferation of tumor cells (Zhang et al. 2020). A previous study reported that HA1 exhibited a signi cant inhibitory effect on the proliferation, invasion, and metastasis of many types of cancer cell lines by inducing cell cycle arrest (Debiton et al. 2004;Gerkens et al. 2007;Liang et al. 2016). Among them, HA1 could inhibit proliferation and induce apoptosis by arresting the cell cycle at the G0/G1 phase, accompanied by a decrease in cyclin D1, cyclin A, cyclin E, and p21 and p27 (Liang et al. 2016). Similar to these results, we found that HA1 inhibited the proliferation of MCF-7 cells by blocking the sub-G1 and G0/G1 phases. Therefore, HA1 may affect the mitosis of MCF-7 cells, thus terminating the normal operation of the cell cycle and inhibiting tumor cell proliferation.
Apoptosis is accompanied by changes in cell morphology, such as cell volume reduction, nuclear membrane shrinkage, chromatin concentration and aggregation, and apoptotic bodies (Zhang et al. 2015). HA1 has a signi cant inhibitory effect on many types of tumor cell lines (NCI-H460, ZR75-1, HL-60, SMMC-7721, Bel-7402, MCF-7) by inducing apoptosis (Gerkens et al. 2007;Liang et al. 2016;Wang et al. 2018;Guan et al. 2020). HA1 can reportedly induce the apoptosis of breast cancer cells (ZR75-1 and MCF-7) and increase the expression of cleaved PARP and caspase-3 (Liang et al. 2016). In addition, one of the derivatives of HA1 can induce cell apoptosis via mitochondria-mediated intrinsic pathways in HepG2 cells . Our results showed that HA1 could induce the concentration and aggregation of chromatin, cause cell shrinkage, and induce the production of apoptotic bodies in MCF-7 cells. Meanwhile, HA1 signi cantly increased the percentage of apoptotic cells. These results preliminarily con rmed the proapoptotic effect of HA1.
In addition, in our previous study, TSS with HA1 as the main active component did not affect the expression of Fas, suggesting that TSS may not induce MCF-7 cell apoptosis through an exogenous signaling pathway (Zhang et al. 2020). We further con rmed that HA1 could signi cantly downregulate Bcl-2 and pro-caspase-3/9, as well as upregulate the expression of Bax and cleaved caspase-3. These results indicate that the caspase cascade reaction was triggered by HA1, which eventually led to the apoptosis of MCF-7 cells. This may be related to the effects of increased ROS levels and decreased MMP. These results suggest that the induction of apoptosis may be one of the mechanisms underlying the antiproliferative effect of HA1.
The molecular mechanisms of the antitumor effects of HA1 remain uncertain, except for its reported membrane damage and high a nity for melanin (Delmas et al. 2000;Debiton et al. 2004). To date, research on the molecular mechanism is mainly limited to the inhibition of the PI3K/AKT/mTOR signaling pathway Guan et al. 2020;Zhang et al. 2020). It has been shown that STAT3 is crucial in numerous tumorigenesis processes, and activation of the JAK-STAT pathway plays a crucial role in the antiproliferation and apoptosis induction of breast cancer cells (Sansone et al. 2012;Ma et al. 2020). Similar saponin compounds, such as raddeanin A and α-hederin, have been reported to inhibit the JAK2/STAT3 pathway (Sun et al. 2018;Wang et al. 2019). Similarly, our results revealed that HA1 inhibited the phosphorylation of both JAK2 and STAT3. These results revealed that the inhibitory effect of JAK2/STAT3 pathway by HA1 may be one of its mechanisms of action.
Although many studies have indicated that HA1 has high antiproliferative activity in vitro, research on its antitumor effects in vivo is still lacking. HA1 has been reported to exhibit a potent antitumor effect in H22 xenografts (Fang et al. 2016a;Fang et al. 2017;Wang et al. 2018). Most in vivo studies have focused on anti-breast cancer; for example, the growth and metastasis of 4T1 cell xenografts were signi cantly inhibited after intraperitoneal injection with HA1 (5 mg/kg) (Liang et al. 2016). HA1 (15 mg/kg) exerted an inhibitory effect on transplanted MCF-7 cells in nude mice . We also reported that HA1 was the main active ingredient of TSS (the alkaline hydrolysis product of A. raddeana), and TSS showed excellent antitumor activity against breast cancer both in vitro and in vivo (Zhang et al. 2020). Similar to these results, we preliminarily con rmed that HA1 (50 and 100 mg/kg) could inhibit the growth of transplanted MCF-7 cells in vivo. These results reveal that HA1 may possess therapeutic potential in breast cancer. The difference in the tumor inhibition rate between 50 and 100 mg/kg HA1 was not obvious; thus, in comparison to the dosage used in other studies (Fang et al. 2016;Fang et al. 2017;Liang et al. 2016;Wang et al. 2018;Kang et al. 2018), our dosage may have been too high. The dosage was set below 50 mg/kg in further experiments.
In conclusion, this study further con rmed that HA1 was cytotoxic to HepG2, MCF-7, MDA-MB-231, SKBr-3, HT-29, and HCT-116 cells. HA1 signi cantly inhibited the proliferation of MCF-7 cells by promoting cell cycle arrest and inducing apoptosis, which may be mediated by promoting ROS generation, decreasing MMP, and activating the mitochondrial apoptosis pathway. Meanwhile, HA1 inhibits the activation of the JAK2/STAT3 pathway, which may contribute to the apoptosis-inducing effect of HA1. Finally, the antitumor activity of HA1 was con rmed in vivo. These results suggest that HA1 has the potential to be developed as a new agent for breast cancer therapy.  HA1 induced the apoptotic response of MCF-7 cells by activating mitochondrial pathway. (A): The levels of Bax, Bcl-2, caspase-3/9 and cleaved-caspase-3/9 were signi cantly decreased by HA1; (A'): The bar chart exhibited the values of mean ± S.D. of three times. β-actin was the internal control. *P < 0.05 vs. Control; **P < 0.01 vs. Control; ***P < 0.001 vs. Control.