Synergistic Combination of Luteolin and Asiatic Acid on Cervical Cancer In Vitro and In Vivo

Simple Summary This study was the first to demonstrate the anticancer effects of luteolin (Lut) combined with asiatic acid (AsA) on CaSki and HeLa cervical cancer cells. Lut combined with AsA effectively reduced cervical cancer cell viability by arresting the cell cycle in the sub-G1 phase and inducing caspase-mediated intrinsic apoptosis. Lut combined with AsA treatment downregulated PI3K/AKT signaling (PI3K, AKT and p70S6K), JNK/p38 MAPK signaling and FAK signaling (integrin β1, paxillin and FAK) and upregulated ERK signaling to induce apoptosis and inhibit cancer cell migration. In in vivo study, Lut combined with AsA markedly inhibited cervical cancer cell-derived xenograft tumor growth and the cytotoxic effect was minimal or nonexistent. Collectively, the present study demonstrated that Lut combined with AsA may be used as an anticancer agent in future clinical treatment to improve the prognosis of cervical cancer. Abstract Cervical cancer is an important issue globally because it is the second most common gynecological malignant tumor and conventional treatment effects have been shown to be limited. Lut and AsA are plant-derived natural flavonoid and triterpenoid products that have exhibited anticancer activities and can modulate various signaling pathways. Thus, the aim of the present study was to evaluate whether Lut combined with AsA could enhance the anticancer effect to inhibit cervical cancer cell proliferation and examine the underlying molecular mechanisms in vitro and in vivo. The results of a CCK-8 assay showed that Lut combined with AsA more effectively inhibited the proliferation of CaSki and HeLa cells than Lut or AsA treatment alone. Lut combined with AsA caused apoptosis induction and sub-G1-phase arrest in CaSki and HeLa cells, as confirmed by flow cytometry, mitoROS analysis, antioxidant activity measurement and western blot assay. In addition, Lut combined with AsA significantly inhibited the cell migration ability of CaSki and HeLa cells in a wound-healing assay. Furthermore, Lut combined with AsA induced apoptosis and inhibited migration through downregulated PI3K/AKT (PI3K, AKT and p70S6K), JNK/p38 MAPK and FAK (integrin β1, paxillin and FAK) signaling and upregulated ERK signaling. In an in vivo study, Lut combined with AsA markedly inhibited cervical cancer cell-derived xenograft tumor growth. Collectively, the present study showed that Lut combined with AsA may be used as an anticancer agent to improve the prognosis of cervical cancer. Indeed, with additional research to develop standardized dosages, Lut and AsA combination therapy could also be applied in clinical medicine.


Introduction
Cervical cancer is one of the major gynecologic malignancies, with more than 600,000 women diagnosed and 342,000 deaths in 2020 [1]. Conventional treatment has a limited efficacy and is prone to cause drug resistance, recurrence and metastasis, and thus the outcome is not favorable; the overall 5-year survival rate is only 66.7% for advanced cervical cancer patients [2]. Several studies have shown that approximately 90% of cervical carcinomas are associated with high-risk human papillomavirus (HPV) infection, especially HPV16 and HPV18, which cause 70-72% of invasive cervical cancers [3][4][5]. Thus, new treatment agents to inhibit HPV-linked carcinogenesis, improve the chemotherapy outcome and enhance the overall survival rate of cervical cancer patients need to be developed.
Lut (3',4',5,7-tetrahydroxyflavone, C 15 H 10 O 6 ) is a natural flavonoid present in different plants species, such as vegetables, fruits and medicinal herbs [31,32]. Numerous studies have demonstrated that Lut has various biological effects and protective functions, including anti-inflammatory, antioxidant, anti-allergic, immune-modulation, anti-microbiota and anti-atherosclerosis effects [33][34][35][36]. Additionally, Lut has been observed to possess the ability to obstruct the development of various cancers in vitro and in vivo by inhibiting cancer cell proliferation, activating cell cycle arrest, protecting carcinogenic stimuli, inducing apoptosis and suppressing cancer metastasis through different signaling pathways [37][38][39][40]. Wu et al. [41] and Tsai et al. [42] revealed that Lut can effectively inactivate the AKT/mTOR pathway and reverse epithelial-mesenchymal transition (EMT) to suppress breast cancer cell proliferation and metastasis, as well as downregulating Nrf2-mediated expression to enhance chemosensitivity in breast cancer treatment. Potočnjak et al. [43] also indicated that the antitumor activity of Lut in colon cancer SW620 cells acted through the ERK/FOXO3adependent mechanism and had anti-metastatic potential. Previous studies also revealed that Lut could inhibit the proliferation of HeLa cervical cancer cells through G2/M arrest, apoptosis and upregulation of the p16 INK4A and JNK expression and downregulation of TNF-α-induced NF-κB activation [44,45]. Lut showed a synergistic effect with TNF-related apoptosis-inducing ligand (TRAIL), which could induce apoptosis in HeLa cells by activation of death receptor 5 and caspase-8 [46]. These results suggested that Lut may be a potentially effective anticancer agent for the treatment of cervical cancer. Based on these results, this study assessed whether Lut can enhance the anticancer effect of AsA on cervical Cancers 2023, 15, 548 3 of 24 cancer cells and further evaluated the underlying mechanism using both cervical cancer cells (CaSki, HeLa and C33A) and a xenograft mouse model.

Cell Viability and Drug Combination Assay
CaSki, HeLa and C33A cells were seeded into 96-well plates with 2 × 10 4 cells per well and then treated with different concentrations of Lut (10, 50, 75 and 100 µM) with or without AsA (25, 50 and 75 µM) for 24 h. To each well, 90 µL fresh culture medium and 10 µL CCK-8 solution (cell counting kit-8, #CK04; Dojindo Molecular Technique, Inc., Rockville, MD, USA) were added and cells were incubated at 37 • C for 1 h. Absorbance was measured at 450 nm using a FLUOstar Galaxy microplate reader (BMG Labtech, Ortenberg, Germany). The effect of combination was evaluated using the combination index (CI) method of Chou and Talalay and CalcuSyn software (Biosoft, Cambridge, UK) [47]. Generally, CI < 1 indicates synergy, CI = 1 indicates additivity and CI > 1 indicates antagonism. Cells treated with 0.1% DMSO in culture medium were used as controls and were regarded as 100% viable and the viabilities of the Lut-treated or AsA-treated cells were determined. The half-inhibitory concentration (IC50) values were calculated as the drug concentration that inhibited cell proliferation by 50% compared to vehicle controls. In further experiments, cells were pretreated with z-VAD-fmk (#1009-20C, 2.5 µmol/L, BioVision, Inc.) prior to Lut and AsA treatment.

Apoptosis Analysis
CaSki and HeLa cells were seeded into 6-well plates with 1 × 10 6 cells per well and treated with 50 µM Lut with or without 50 or 75 µM AsA for 24 h. Then, an FITC Annexin V apoptosis detection kit (#556547, BD Biosciences, Bergen, NJ, USA) was used to detect 10,000 collected cells in different stages of cell death; cells were harvested and suspended in 1× Annexin V binding buffer (100 uL) and then double-stained with 5 µL FITC Annexin V (20 µg/mL) and 5 µL propidium iodide (PI, 50 µg/mL) in the dark for 15 min at room temperature (RT), adding 1x Annexin V binding buffer (900 uL) to each tube. Finally, stained cells were analyzed by Cytomics FC500 flow cytometry and CXP software (version 2.3; Beckman Coulter, Inc., Brea, CA, USA). Early and late apoptotic/necrotic cells were assessed and quantified.

Cell Cycle Analysis
CaSki and HeLa cells were seeded into 6-well plates with 1 × 10 6 cells per well and treated with 50 µM Lut with or without 50 or 75 µM AsA for 24 h. Cells were then collected and fixed with 70% ice-cold ethanol (1 mL) at −20 • C overnight (16-18 h). After fixation, the cells were centrifuged at 400× g at 4 • C for 10 min, washed with cold PBS and stained with 0.5 mL BD Pharmingen TM PI/RNases staining buffer (PI, 10 µg/mL; RNases, 300 µg/mL; #550825; BD Biosciences) in the dark for 15 min at RT (25 • C). Finally, cell cycle distribution data were collected for 10,000 collected cells by Cytomics FC500 flow cytometry and CXP software (version 2.3; Beckman Coulter, Inc.).

Mitochondrial ROS Measurement
CaSki and HeLa cells were seeded into 6-well plates with 1 × 10 6 cells per well and treated with 50 µM Lut with or without 50 or 75 µM AsA for 24 h. The cells were harvested and suspended in RPMI 1640 medium or EMEM medium (0.5 mL) and then the cells were immediately supplemented with 2.5 uL of 0.5 mM MitoSOX Red mitochondrial superoxide indicator solution (#M36008; Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA, USA; the final concentration of MitoSOX = 2.5 µM) and incubated in a 37 • C water bath for 20 min in the dark. The cells washed with 37 • C prewarmed PBS, submitted to centrifugation to pellet the cells and then resuspended in warm PBS buffer (0.5 mL). Fluorescence signal data were obtained using a Cytomics FC 500 flow cytometer (Beckman Coulter, Inc.) and analyzed using FlowJo software (version 7.6; BD Life Sciences, Franklin Lakes, NJ, USA).

Measurement of Glutathione (GSH) and Catalase Levels
The glutathione (GSH) and catalase levels of the cell lysate were determined using an Amplite TM Fluorimetric Glutathione assay kit (#10055; AAT Bioquest ® , Inc., Sunnyvale, CA, USA) and an Amplite TM Fluorimetric Catalase assay kit (#11306; AAT Bioquest ® , Inc.), respectively. CaSki and HeLa cells were seeded into 6-well plates with 1 × 10 6 cells per well and treated with 50 µM Lut with or without 50 or 75 µM AsA for 24 h. PBS was used to wash and suspend cells, followed by centrifugation for 10 min at 400× g. Collected cells were resuspended in 0.1% Triton X-100 (250 µL), followed by centrifugation for 7 min at 380× g and further centrifugation at 16,000× g for 1 min. Supernatant (50 µL) was added onto the 96-well microliter plate, followed by 50 µL Thiolite TM Green reaction mixture and the plate was incubated at RT for 30 min in the dark. Absorbance was measured at Ex/Em = 490/520 nm using a CLARIOstar fluorescence microplate reader (BMG Labtech, Ortenberg, Germany). To measure the total catalase activity, 50 µL supernatant was applied onto a 96-well microliter plate, and 50 µL H 2 O 2 (10 µM) assay buffer was added, followed by incubation at RT for 30 min in the dark, after which 50 µL catalase assay mixture (1× Amplite TM Red and 100 mU/mL horseradish peroxidase) was applied. After 30 min, the absorbance was measured at Ex/Em = 540/590 nm using a CLARIOstar fluorescence microplate reader (BMG Labtech).

Wound-Healing Assay
CaSki and HeLa cells were cultured in 6-well plates with 1 × 10 6 cells per well to form a confluent monolayer and straight wounds were made by scratching with a 200-µL pipette tip. After washing with medium to remove cell debris, the wounded monolayers were treated with 50 µM Lut with or without 50 or 75 µM AsA for 0, 12 and 24 h. The wound gaps were photographed at regular intervals (0, 12 and 24 h) under an Olympus BX61 microscope (Olympus Corporation, Shinjuku, Tokyo, Japan) at 100× magnification, and the cell-free wound areas were measured using ImageJ software (http://rsb.info.nih.gov/ij, accessed on 11 May 2022, NIH, Bethesda, MD, USA). To reduce variability in the results, multiple views of each well were documented, and wound closure was evaluated as a percentage relative to the untreated control.

Statistical Analysis
Statistical analyses were performed using the one-way ANOVA followed by Bonferroni correction's post-test using GraphPad software (version 9.3.0; Dotmatics, Inc., San Diego, CA, USA). All data representing results are presented as the mean ± standard deviation (SD) from at least three independent experiments. A p value < 0.05 was considered to indicate a statistically significant difference.

Luteolin and Asiatic Acid Inhibited the Proliferation of Cervical Cancer Cell Lines
To assess the anti-proliferation effects of both Lut and AsA on cervical cancer cells, a CCK-8 assay was performed using CaSki, HeLa and C33A cells with various concentrations of Lut (0-100 µM) with or without AsA (0-75 µM) for 24 h. We observed that Lut and AsA dose-dependently decreased the proliferation of CaSki, HeLa and C33A cells ( Figure 1, p < 0.05; and Figure S1 for 48 and 72 h), although 25 µM AsA did not, in the case of CaSki cells. Moreover, we also found that Lut had a lower IC 50 value, of 46.6 ± 2.7 µM for HeLa cells, than CaSki and C33A cells, which had values of 115.4 ± 1.8 µM and 117.0 ± 12.1 µM, respectively. The IC 50 values for the AsA treatment were found to be 47.8 ± 0.6 µM for HeLa cells, 71.4 ± 6.2 µM for CaSki cells and 63.2 ± 4.5 µM for C33A cells. Lut was less active than AsA, which showed more strong antiproliferative activity against CaSki cells and C33A cells. In addition, we further investigated cell proliferation under a combination of Lut and AsA. Compared with the control cells (100%), the proliferation of CaSki cells was significantly decreased by 28% and 57% after treatment with 50 and 75 µM AsA (71.6 ± 2.1% vs. 42.8 ± 1.8%) and inhibited by 59% and 71% when treated in combination with 50 µM Lut (40.9 ± 0.3% vs. 28.9 ± 1.5%), respectively. The 75 µM AsA combined with 75 µM Lut (26.7 ± 1.4%) also had approximately 73% inhibition efficiency. Similarly, HeLa cell proliferation significantly decreased by 52% and 91% after treatment with 50 and 75 µM AsA (47.5 ± 2.1% vs. 9.0± 0.6%) and was markedly inhibited, by 89% and 93%, when combined with 50 µM Lut (47.5 ± 2.1% vs. 9.0± 0.6%), respectively. The 75 µM AsA combined with 75 µM Lut (5.1 ± 1.0%) had 95% inhibition efficiency. Additionally, C33A cell proliferation significantly decreased, by 48% and 58%, after treatment with 50 and 75 µM AsA (52.4 ± 0.8% vs. 42.3± 6.9%) and was inhibited by 47% and 49% when combined with 50 µM Lut (53.4 ± 3.1% vs. 50.5 ± 1.0%), respectively. The 75 µM AsA combined with 75 µM Lut (48.8 ± 0.9%) only had 51% inhibition efficiency. Additionally, it can be seen from comparison of CI values, AsA combined with Lut exerted synergistic effects, with the CI being 0.910 and 0.885 for 50 or 75 µM AsA combined with 50 µM Lut for CaSki cells; and 0.894 and 0.941 for 50 or 75 µM AsA combined with 50 µM Lut for HeLa cells; and 1.145 and 1.264 for 50 or 75 µM AsA combined with 50 µM Lut for C33A cells, respectively. Therefore, AsA combined with Lut was more effective in reducing cell proliferation of CaSki and HeLa cells relative to the control cells and to cells treated with Lut or AsA alone, but C33A cells did not show a synergistic effect.

Luteolin and Asiatic Acid Upregulated Mitochondrial ROS (mitoROS) and Downregulated Glutathione (GSH) in Cervical Cancer Cell Lines
Excessive mitochondrial ROS has been shown to induce cell injury and death an induce apoptosis or necroptosis. Mitochondrial antioxidants have been reported to be e fective in cancer prevention and anticancer therapy. Therefore, we examined mitochon drial ROS and mitochondrial antioxidant expressions using the MitoSOX Red reagent an an enzyme-linked immunosorbent assay (ELISA). The results revealed that the MitoSO Red-derived fluorescence level was significantly increased at 50 μM Lut,= and 50 or 7

Luteolin and Asiatic Acid Upregulated Mitochondrial ROS (mitoROS) and Downregulated Glutathione (GSH) in Cervical Cancer Cell Lines
Excessive mitochondrial ROS has been shown to induce cell injury and death and induce apoptosis or necroptosis. Mitochondrial antioxidants have been reported to be effective in cancer prevention and anticancer therapy. Therefore, we examined mitochondrial ROS and mitochondrial antioxidant expressions using the MitoSOX Red reagent and an enzyme-linked immunosorbent assay (ELISA). The results revealed that the MitoSOX Red-derived fluorescence level was significantly increased at 50 µM Lut,= and 50 or 75 µM AsA in CaSki and HeLa cells (CaSki cells, 363.8 ± 49.4% vs. 334.5 ± 30.0% vs. 764.5 ± 123.6%; HeLa cells, 813.8 ± 286.1% vs. 1183.3 ± 201.6% vs. 1386.2 ± 363.2%, respectively) as compared with the non-treated control cells (100%, Figure 4A

Luteolin and Asiatic Acid Inhibited Cell Migration of Cervical Cancer Cells
To investigate the cell migration ability of cervical cancer cell lines treated with Lut and AsA, we performed a wound-healing assay. The cells in the untreated control group showed a faster cell migration ability than cells treated with Lut or AsA ( Figure 5A-C, p < 0.05). Furthermore, the combination of Lut and AsA markedly inhibited the migration

Luteolin and Asiatic Acid Inhibited Cell Migration of Cervical Cancer Cells
To investigate the cell migration ability of cervical cancer cell lines treated with Lut and AsA, we performed a wound-healing assay. The cells in the untreated control group showed a faster cell migration ability than cells treated with Lut or AsA ( Figure 5A-C, p < 0.05). Furthermore, the combination of Lut and AsA markedly inhibited the migration rate of CaSki and HeLa cells as compared with treatment with Lut or AsA alone at 12 and 24 h, although the combination with a higher dose of AsA in CaSki and HeLa cells did not. As integrin/FAK/paxillin has been reported to cause tumor migration and invasion, we next examined these modulated cell migration-related protein expressions by western blotting. Lut or AsA monotherapy significantly decreased the phospho-FAK (ratio of control: 0. 69

Luteolin and Asiatic Acid Activated the Mitochondrial-Related Intrinsic Signaling Pathway
To further confirm our prior results showing that Lut and AsA treatment induced cervical cancer cell death by apoptosis, apoptosis-related proteins expressions were determined by western blotting. The western blotting findings revealed that Lut or AsA treatment significantly increased the expressions of cleaved PARP-1, pro-apoptotic protein Bax and cleaved caspase-3, whereas anti-apoptotic protein Bcl-2 was significantly decreased in expression as compared with the control in CaSki and HeLa cervical cancer cells (CaSki cells, ratio of control: 1. 23   HeLa (C) cell semi-quantitative analysis of relative wound closure was performed by measuring the width of the wounds. Cell migration-related proteins, p-FAK, integrin β1 and paxillin, levels in CaSki (D) and HeLa (E) cells were analyzed through western blotting after 24 h of treatment. GAPDH served as the loading control. Quantitative results are shown in the lower plot. Values represent mean ± SD from three replicates. *, †, δ and γ p < 0.05 compared with CON, L50-, A50-or A75-treated group. CON, 0.1% DMSO; L50, 50 µM luteolin; A50, 50 µM asiatic acid; A75, 75 µM asiatic acid. L50 + A50, 50 µM luteolin + 50 µM asiatic acid; L50 + A75, 50 µM luteolin + 75 µM asiatic acid.

Luteolin and Asiatic Acid Activated the Mitochondrial-Related Intrinsic Signaling Pathway
To further confirm our prior results showing that Lut and AsA treatment induced cervical cancer cell death by apoptosis, apoptosis-related proteins expressions were determined by western blotting. The western blotting findings revealed that Lut or AsA treatment significantly increased the expressions of cleaved PARP-1, pro-apoptotic protein Bax and cleaved caspase-3, whereas anti-apoptotic protein Bcl-2 was significantly decreased in expression as compared with the control in CaSki and HeLa cervical cancer cells (CaSki cells, ratio of con  (Figure 6 and Figure S3, p < 0.05). Taken together, these findings showed that Lut and AsA induced pro-apoptotic effects in CaSki and HeLa cells through a mitochondria-related intrinsic signaling pathway. ratio of control: 4.0 ± 0.1 vs. 1.25 ± 0.03 vs. 2.16 ± 0.10 vs. 0.83 ± 0.04, respectively) as compared with Lut or AsA monotherapy, although no significant differences were observed in the Bax expression in CaSki cells and Bcl-2 expression in HeLa cells compared with AsA monotherapy (Figures 6 and S3, p < 0.05). Taken together, these findings showed that Lut and AsA induced pro-apoptotic effects in CaSki and HeLa cells through a mitochondriarelated intrinsic signaling pathway.

Luteolin and Asiatic Acid Suppressed Cervical Cancer Cell-Derived Xenograft Tumors
To further confirm the in vitro anti-proliferation effects of Lut and AsA, a CaSki xenograft tumor model was constructed in BALB/c nude mice. The schematic timeline of this experiment was as shown in Figure 9A. Lut or AsA markedly inhibited the in vivo tumor growth of CaSki xenografts, particularly Lut combined with AsA. After treatment for 16 days, the average tumor volume of the CaSki xenograft tumors was 349.4 ± 108.7 mm 3 vs. 274.1 ± 88.2 mm 3 vs. 230.3 ± 58.4 mm 3 vs. 165.7 ± 42.6 mm 3 in the vehicle control, AsA-, Lut-and Lut plus AsA-treated groups, while the tumor growth inhibition rate was 21.6% vs. 34.1% vs. 52.6%, respectively ( Figure 9B, p < 0.05). No significant difference in body weight between groups was observed during the experimental period (21.3 ± 1.4 g vs. 21.4 ± 1.2 g vs. 21.4 ± 0.9 g vs. 21.7 ± 1.5 g, respectively) and the experimental animal survival Values represent mean ± SD from three replicates. * and † p < 0.05 compared with CON or AsA + Lut-treated group. CON, 0.1% DMSO; Comb., 50 µM luteolin + 50 µM asiatic acid; Comb. + z-VAD, 50 µM luteolin + 50 µM asiatic acid + 2.5 µM z-VAD-fmk.

Luteolin and Asiatic Acid Suppressed Cervical Cancer Cell-Derived Xenograft Tumors
To further confirm the in vitro anti-proliferation effects of Lut and AsA, a CaSki xenograft tumor model was constructed in BALB/c nude mice. The schematic timeline of this experiment was as shown in Figure 9A. Lut or AsA markedly inhibited the in vivo tumor growth of CaSki xenografts, particularly Lut combined with AsA. After treatment for 16 days, the average tumor volume of the CaSki xenograft tumors was 349.4 ± 108.7 mm 3 vs. 274.1 ± 88.2 mm 3 vs. 230.3 ± 58.4 mm 3 vs. 165.7 ± 42.6 mm 3 in the vehicle control, AsA-, Lut-and Lut plus AsA-treated groups, while the tumor growth inhibition rate was 21.6% vs. 34.1% vs. 52.6%, respectively ( Figure 9B, p < 0.05). No sig-nificant difference in body weight between groups was observed during the experimental period (21.3 ± 1.4 g vs. 21.4 ± 1.2 g vs. 21.4 ± 0.9 g vs. 21.7 ± 1.5 g, respectively) and the experimental animal survival rate was 100%, suggesting that treatment with Lut and AsA alone or in combination did not cause host drug toxicity ( Figure 9C). The average tumor weight was significantly reduced in the AsA-, Lut-and Lut plus AsA-treated groups (0.13 ± 0.03 g vs. 0.11 ± 0.03 g vs. 0.05 ± 0.02 g, respectively) as compared with the control group (0.18 ± 0.07 g, Figure 9D, p < 0.05). Moreover, in performing H&E staining, we found that the Lut plus AsA treatment more effectively reduced the tumor cellularity and apoptosis in the CaSki xenografts than AsA or Lut monotherapy. Similarly, consistent with the in vitro findings in this study, IHC study demonstrated more significantly downregulated expressions of ki67 (4.1 ± 1.8 positive cells), integrin β1 (2.9 ± 1.2 positive cells), p-FAK (4.0 ± 1.5 positive cells) and paxillin (4.0 ± 2.0 positive cells) and an upregulated cleaved caspase-3 (13.0 ± 4.5 positive cells) protein expression in tumors treated with Lut plus AsA as compared with the AsA-, Lut-and control-groups (ki67, 7.5 ± 2.4 vs. 6.0 ± 1.6 vs. 16.2 ± 3.7 positive cells; integrin β1, 5.1 ± 1.8 vs. 6.4 ± 2.3 vs. 8.4 ± 2.6 positive cells; p-FAK, 6.7 ± 1.9 vs. 5.0 ± 1.3 vs. 9.0 ± 2.9 positive cells; paxillin, 9.1 ± 3.6 vs. 7.5 ± 2.9 vs. 17.3 ± 3.8 positive cells; cleaved caspase-3, 6.4 ± 2.6 vs. 8.0 ± 2.9 vs. 4.5 ± 1.8 positive cells, respectively; Figure 9E, p < 0.05). These results further suggested that Lut plus AsA combined treatment acted as an anti-proliferative, anti-migratory and pro-apoptotic agent in vivo. not cause host drug toxicity ( Figure 9C). The average tumor weight was significantly reduced in the AsA-, Lut-and Lut plus AsA-treated groups (0.13 ± 0.03 g vs. 0.11 ± 0.03 g vs. 0.05 ± 0.02 g, respectively) as compared with the control group (0.18 ± 0.07 g, Figure  9D, p < 0.05). Moreover, in performing H&E staining, we found that the Lut plus AsA treatment more effectively reduced the tumor cellularity and apoptosis in the CaSki xenografts than AsA or Lut monotherapy. Similarly, consistent with the in vitro findings in this study, IHC study demonstrated more significantly downregulated expressions of ki67  Values represent mean ± SD (n = 6/group). *, † and δ p < 0.05 compared with CON, Lut-or AsAtreated groups. CON, control; Lut, luteolin; AsA, asiatic acid; Comb., luteolin combined with asiatic acid.

Discussion
Cervical cancer remains a leading cause of high mortality and morbidity in women worldwide. Although chemotherapy can effectively to improve the survival of cervical cancer patients, it is also prone to increase drug resistance, toxic side effect and complications, causing great suffering to patients. Therefore, a new oncology therapy for cervical cancer by plant-derived natural products (such as alkaloid, flavonoids, phenols and terpenoids) has attracted attention. AsA and Lut themselves are rich in natural triterpenoids and flavonoids, and their therapeutic effects on multiple cancers' proliferation and metastasis have been studied. Based on this, this study aimed to assess the anticancer effect of AsA and Lut in human cervical cancer. Simultaneously, we also evaluated whether Lut combined with AsA more notably enhanced the anticancer effect than Lut or AsA alone in vitro and in vivo. Our study results showed that AsA or Lut alone can reduce the cell viability of human cervical cancer cells (CaSki, HeLa and C33A, Figure 1), which were consistent with previous studies showing that AsA significantly inhibited cell growth of Values represent mean ± SD (n = 6/group). *, † and δ p < 0.05 compared with CON, Lut-or AsA-treated groups. CON, control; Lut, luteolin; AsA, asiatic acid; Comb., luteolin combined with asiatic acid.

Discussion
Cervical cancer remains a leading cause of high mortality and morbidity in women worldwide. Although chemotherapy can effectively to improve the survival of cervical cancer patients, it is also prone to increase drug resistance, toxic side effect and complications, causing great suffering to patients. Therefore, a new oncology therapy for cervical cancer by plant-derived natural products (such as alkaloid, flavonoids, phenols and terpenoids) has attracted attention. AsA and Lut themselves are rich in natural triterpenoids and flavonoids, and their therapeutic effects on multiple cancers' proliferation and metastasis have been studied. Based on this, this study aimed to assess the anticancer effect of AsA and Lut in human cervical cancer. Simultaneously, we also evaluated whether Lut combined with AsA more notably enhanced the anticancer effect than Lut or AsA alone in vitro and in vivo. Our study results showed that AsA or Lut alone can reduce the cell viability of human cervical cancer cells (CaSki, HeLa and C33A, Figure 1), which were consistent with previous studies showing that AsA significantly inhibited cell growth of ovarian cancer [19], nasopharyngeal carcinoma [22] and breast cancer [23,24]; and Lut showed results against leukemia [48], prostate cancer [49] and breast cancer [41,42]. In addition, the Lut combined with AsA treatment also had a significantly more synergistic effect against cervical cancer growth than Lut or AsA alone in CaSki and HeLa cells, and synergistically inhibited xenograft animal tumor growth (Figure 9). These results were consistent with those previously reported by Jeon et al. and Johnson et al., who indicated that the combined use of Lut and paclitaxel or gemcitabine synergistically augmented anticancer activity in breast and pancreatic cancer treatment [50,51]. Lian et al. also showed that AsA combined with naringenin could have a greater effect to suppress melanoma and lung carcinoma growth [27]. In addition, Ham et al. revealed that Lut treatment resulted in a more significant cytotoxic effect in HPV-positive cervical cancer cells and had less effect on HPV-negative cervical cancer C33A cells [52], with as Kim et al. indicating that bee venom could significantly promote the suppression of cell growth in HPV16-infected cells (CaSki cells) and HPV18-infected cells (HeLa cells) compared to HPV-noninfected cells (C33A cells), via downregulation of HPV E6/E7 expression [53]. Similarly, AsA, Lut and combination treatment promoted cell apoptosis (as upregulation of Bax/Bcl-2 ratio, cleaved-PARP-1 and cleaved caspase-3) and inhibited cell migration (downregulation of integrin β1, FAK and paxillin) in both CaSki and HeLa cells and xenograft animal model (Figures 2, 5, 6 and 9), which were consistent results of previous studies. [22,39,54,55].
As consistent with results of previous studies, AsA induced cell death by causing cancer cells to undergo G0/G1-(SKOV3 and OVCAR-3) [19 Ren 2016] or S-G2/M-phases arrest (MCF-7, MDA-MB-231, SW480 and HCT116) [21,56] and Lut was involved in G0/G1-(HepG2, NCI-H1975 and NCI-H1650) [57,58] or G2/M-phases arrest (A549, HeLa, LoVo, HCT-116 and HT-29) [44,[59][60][61]. Lut combined with oxaliplatin treatment promoted SGC-7901 cell apoptosis by altering G0/G1 phase proportion [62]. In our study, AsA, Lut and combination treatment were causing sub-G1 phase arrest in CaSki and HeLa cells (Figure 3), which was consistent with Horninaka et al., who revealed that Lut combined with TRAIL treatment synergistically induced cell cycle sub-G1 arrest in HeLa cells [46]. These differences in results represented in the cell cycle phase arrest of AsA and Lut treatment may be associated with cancer cell type specificity and individual AsA and Lut bioavailability (dose-or time-stimuli manner). Payent et al., [63] and Lv et al. [15] revealed that mitochondrial ROS exerted dual roles in tumor-suppression or tumor-promotion, and in the presence of a high level of mitROS that will cause tissue damage and cell death, AsA treatment could remarkably inhibit ROS formation and increase the GSH, superoxide dismutase (SOD) and CAT levels to attenuate oxidative stress damage, whereas Park et al. [64] and Li et al. [28] indicated that AsA caused the induction of ROS generation to induce apoptosis in SK-MEL-2 and HepG2 cells. Our results showed that AsA, Lut and combination significantly increased mitROS production and promoted cancer cell apoptosis (Figures 2 and 4), and, as Imhoff et al. [65] revealed, high levels of ROS-induced mitROS will cause cell apoptosis and autophagy to reduce tumorigenesis, thereby affecting the GSH and CAT expressions, which have not increased. However, luteolin was be found to increase or attenuate productions of ROS in HepG2 and MCF-7 cells, respectively [66,67].
Previous research revealed that the Lut or AsA induced apoptosis through the inactivation of the PI3K/AKT/mTOR/p70S6K pathway in some common cancers, including glioblastoma [68], lung [69], ovarian [19], colon [21] and breast cancers [41,70], which is consistent with our study finding that AsA, Lut and combination treatment inhibited PI3K, p-AKT and p-p70S6K protein expressions in CaSki and HeLa cells (Figure 7). However, Zhou et al. [71] showed that Lut activity against prostate cancer PC3 cell invasion was through active AKT to inhibit mdm2 expression, resulting in attenuated E-cadherin expression. In addition, activation of p-p38 and p-JNK1/2 were inhibited with AsA, Lut and combination treatment in CaSki and HeLa cells, whereas p-ERK1/2 was not inhibited. Caspase inhibitor z-VAD-fmk significantly reversed CaSki and C33A cell growth and increased p-p38 activation in cells co-treated with Lut and AsA (Figure 8), which can be explained by the fact that p38 regulating Lut combined with AsA-induced cell apoptosis. These findings were consistent with those of previous reports that Lut and celecoxib combinatorial therapy protected against breast cancer cells (MDA-MB-231 and SkBr3) through activating ERK and inactivating AKT and through inactivating ERK and AKT in other breast cancer cells (MCF-7 and MCF7/HER18) [72]. However, Potočnjak et al. showed the Lut increased colon cancer SW620 cells' viability by upregulation of the p-ERK/p-JNK/p-p38 pathway [43]. Hsu et al. [56] indicated that AsA induced apoptosis and cell cycle arrest through activating the ERK/p38 MAPK pathway in breast cancers (MCF-7 and MDA-MB-231). Hish et al. and Chen et al. also showed that AsA suppressed human renal cancer cell migration and invasion via inhibition of the p-ERK/p-p38MAPK pathway [26], while AsA-induced apoptosis in cisplatin-resistant nasopharyngeal carcinoma cells was by activation of the p-JNK/p-p38MAPK pathway and inactivation of the p-ERK pathway [22]. These inconsistent results of the Lut-or AsA-mediated signaling pathway may be dependent on cell type and dose-or time-stimuli manner. The detailed mechanism will require further clarification.

Conclusions
Drug resistance and side effects are frequently developed in patients undergoing cervical cancer chemotherapy. A novel anticancer agent, especially derived from natural sources, is required to address these issues. Our study is the first to have demonstrated an anticancer effect of combined Lut plus AsA treatment on cervical cancer in vitro and in vivo. Lut combined with AsA effectively reduced cervical cancer cell viability by arresting the cell cycle in the sub-G1 phase and inducing caspase-mediated intrinsic apoptosis. Lut combined with AsA treatment downregulated PI3K/AKT signaling (PI3K, AKT and p70S6K), JNK/p38 MAPK signaling and FAK signaling (integrin β1, paxillin and FAK) and upregulated ERK signaling to induce apoptosis and inhibit cancer cell migration, resulting in an anticancer effect on cervical cancer ( Figure 10). Furthermore, our in vivo study revealed that Lut combined with AsA markedly inhibited cervical cancer cell-derived xenograft tumor growth, and the cytotoxic effect of Lut combined with AsA was minimal or nonexistent. Collectively, the present study showed that Lut combined with AsA may be used as an anticancer agent to improve the prognosis of cervical cancer. Indeed, with additional research to develop standardized dosages, Lut and AsA combination therapy could also be applied in clinical medicine. . Schematic diagram of luteolin combined with asiatic acid's anticancer molecular mechanism in cervical cancer. Lut combined with AsA mainly modulates FAK signaling (integrin β1, paxillin and FAK) and PI3K/AKT signaling (PI3K, AKT and p70S6K) and causes JNK/p38 downregulation and ERK upregulation, inactivating or activating various signaling targets, such as Bcl2, Bax, mitROS, caspase-3 and RARP1; this leads to the induction of caspase-mediated intrinsic apoptosis, sub-G1 phase arrest and inhibition of cancer cell migration, which increase the anticancer effect on cervical cancer.

Supplementary Materials:
The following are available online at www.mdpi.com/xxx/s1, Figure S1: Luteolin and asiatic acid inhibit cell proliferation in cervical cancer cell lines for 24, 48 and 72 h. Figure S2: Luteolin and asiatic acid inactivate the integrin β1/FAK/paxillin pathway. Figure S3: Luteolin and asiatic acid activate the mitochondrial-related intrinsic pathway (PARP-1, Bcl-2, Bax and Figure 10. Schematic diagram of luteolin combined with asiatic acid's anticancer molecular mechanism in cervical cancer. Lut combined with AsA mainly modulates FAK signaling (integrin β1, paxillin and FAK) and PI3K/AKT signaling (PI3K, AKT and p70S6K) and causes JNK/p38 downregulation and ERK upregulation, inactivating or activating various signaling targets, such as Bcl2, Bax, mitROS, caspase-3 and RARP1; this leads to the induction of caspase-mediated intrinsic apoptosis, sub-G1 phase arrest and inhibition of cancer cell migration, which increase the anticancer effect on cervical cancer.