The Synthetic Flavonoid Derivative GL-V9 Induces Apoptosis and Autophagy in Cutaneous Squamous Cell Carcinoma via Suppressing AKT-Regulated HK2 and mTOR Signals

Cutaneous squamous-cell carcinoma (cSCC) is one of most common type of non-black skin cancer. The malignancy degree and the death risk of cSCC patients are significantly higher than basal cell carcinoma patients. GL-V9 is a synthesized flavonoid derived from natural active ingredient wogonin and shows potent growth inhibitory effects in liver and breast cancer cells. In this study, we investigated the anti-cSCC effect and the underlying mechanism of GL-V9. The results showed that GL-V9 induced both apoptosis and autophagy in human cSCC cell line A431 cells, and prevented the growth progression of chemical induced primary skin cancer in mice. Metabolomics assay showed that GL-V9 potentially affected mitochondrial function, inhibiting glucose metabolism and Warburg effect. Further mechanism studies demonstrated that AKT played important roles in the anti-cSCC effect of GL-V9. On one hand, GL-V9 suppressed AKT-modulated mitochondrial localization of HK2 and promoted the protein degradation of HK2, resulting in cell apoptosis and glycolytic inhibition. On the other hand, GL-V9 induced autophagy via inhibiting Akt/mTOR pathway. Interestingly, though the autophagy induced by GL-V9 potentially antagonized its effect of apoptosis induction, the anti-cSCC effect of GL-V9 was not diluted. All above, our studies suggest that GL-V9 is a potent candidate for cSCC treatment.


Introduction
Cutaneous squamous cell carcinoma (cSCC), originating from the epidermis or adnexal keratinocytes, is one of the highest incidence of non-melanocytic skin carcinoma, only second to basal cell carcinoma. Because cSCC is a kind of highly metastatic malignant tumors, the death risk of cSCC patients is much higher than the patients with basal cell carcinoma [1]. Improvements in treatment of cSCC could be by means of novel agents targeting the signaling pathways that facilitate cancer cell growth and survival [2].
In recent years, a mushrooming of attention and expectations in cancer treatment have been paid to the studies of the active ingredients from natural products. Flavonoid wogonin is extracted from the traditional Chinese medicine Scutellaria baicalensis with multiple of pharmacological effects, The mitochondrial and cytoplasmic extracts were separated. Western blotting was carried out to analyze the expression of Cyt-c, and AIF. Data is shown as average ± SD from three independent experiments. * p < 0.05 and ** p < 0.01 (comparison to the control 100%).
The fate of cells succumbing to the intrinsic pathway is triggered by the loss of mitochondrial membrane potential (MMP, ΔΨm), which is an important function index for mitochondria and can be detected by JC-1 probe. As shown in Figure 1F, exposure to GL-V9 for 36 h resulted in the loss of MMP in a concentration dependent manner. We also isolated the mitochondrial and cytoplasmic protein of A431 cell, and investigated the levels of apoptosis-inducing factor (AIF) and cytochrome c in mitochondria and cytoplasm. Western blotting showed that A431 cells treated with GL-V9 for 36 h significantly decreased the level of cytochrome c (cyt-c) and AIF in mitochondria, and increased them in cytoplasm, suggesting GL-V9 promoted the release of cyt-c and AIF from mitochondria ( Figure 1G). The results demonstrated that GL-V9 caused mitochondrial dysfunction of A431 cells and induced apoptosis via mitochondrial-mediated pathway.

High Concentration of GL-V9 Induces Autophagy of A431 Cells, Which Potentially Antagonizes Its Apoptosis Inducing Effect
On the other hand, we tested the influence of GL-V9 in the autophagy of A431 cells. As shown in Figure 2A, the soluble form of LC3 (LC3-I) converted to the autophagosome-associated form (LC3-II) upon GL-V9 treatment. The polyubiquitin-binding protein p62 is also used as an autophagic marker, which binds directly to LC3 and is itself degraded by autophagy. Interestingly, low concentration of GL-V9 increased p62 protein level, but high concentration of GL-V9 significantly The mitochondrial and cytoplasmic extracts were separated. Western blotting was carried out to analyze the expression of Cyt-c, and AIF. Data is shown as average ± SD from three independent experiments. * p < 0.05 and ** p < 0.01 (comparison to the control 100%).
The fate of cells succumbing to the intrinsic pathway is triggered by the loss of mitochondrial membrane potential (MMP, ∆Ψ m ), which is an important function index for mitochondria and can be detected by JC-1 probe. As shown in Figure 1F, exposure to GL-V9 for 36 h resulted in the loss of MMP in a concentration dependent manner. We also isolated the mitochondrial and cytoplasmic protein of A431 cell, and investigated the levels of apoptosis-inducing factor (AIF) and cytochrome c in mitochondria and cytoplasm. Western blotting showed that A431 cells treated with GL-V9 for 36 h significantly decreased the level of cytochrome c (cyt-c) and AIF in mitochondria, and increased them in cytoplasm, suggesting GL-V9 promoted the release of cyt-c and AIF from mitochondria ( Figure 1G). The results demonstrated that GL-V9 caused mitochondrial dysfunction of A431 cells and induced apoptosis via mitochondrial-mediated pathway.

High Concentration of GL-V9 Induces Autophagy of A431 Cells, Which Potentially Antagonizes Its Apoptosis Inducing Effect
On the other hand, we tested the influence of GL-V9 in the autophagy of A431 cells. As shown in Figure 2A, the soluble form of LC3 (LC3-I) converted to the autophagosome-associated form (LC3-II) upon GL-V9 treatment. The polyubiquitin-binding protein p62 is also used as an autophagic marker, which binds directly to LC3 and is itself degraded by autophagy. Interestingly, low concentration of GL-V9 increased p62 protein level, but high concentration of GL-V9 significantly decreased p62. We thought that the upregulation of p62 by low concentration of GL-V9 might be related with the influence of GL-V9 in energy metabolism. To further confirm the change of autophagy, the GFP-LC3 distribution, extensively used as a biomarker for autophagy, was assayed. As shown Figure 2B, the distribution of GFP-LC3 was changed from largely diffuse to accumulated punctate structures upon GL-V9 treatment by 36 h. Besides, we used LysoTracker Red and found that GL-V9 could promote the aggregation of lysosome of A431 cells ( Figure 2C). All the indexes suggested that higher concentration of GL-V9 could induce autophagy of A431 cells.
Molecules 2020, 25, x 4 of 18 decreased p62. We thought that the upregulation of p62 by low concentration of GL-V9 might be related with the influence of GL-V9 in energy metabolism. To further confirm the change of autophagy, the GFP-LC3 distribution, extensively used as a biomarker for autophagy, was assayed. As shown Figure 2B, the distribution of GFP-LC3 was changed from largely diffuse to accumulated punctate structures upon GL-V9 treatment by 36 h. Besides, we used LysoTracker Red and found that GL-V9 could promote the aggregation of lysosome of A431 cells ( Figure 2C). All the indexes suggested that higher concentration of GL-V9 could induce autophagy of A431 cells. The connection between autophagy and apoptosis or other forms of cell death is a burgeoning area of research. To investigate whether GL-V9 induced apoptosis and autophagy were synergetic or antagonistic, we used caspase inhibitor Z-VAD-FMK and autophagy inhibitor 3-MA combined with GL-V9. As shown in Figure 2D, Z-VAD-FMK reversed the apoptosis inducing effect of GL-V9, but had little influence in the GL-V9-induced increase of autophagy marker LC3-II. Different with Z-VAD-FMK, after cell autophagy was inhibited by 3-MA, the cleaved of caspase 3 induced by GL-V9 was increased ( Figure 2E). These results suggested that the autophagy induced by high concentration of GL-V9 potentially antagonized the effect of apoptosis induction, which might be associated with the influence of cell metabolism. Instead, the overall anti-cSCC effects of GL-V9 were not influenced by the dual regulations of apoptosis and autophagy.

GL-V9 Induced Metabolome Changes in A431 Cells
Then we studied the metabolomics of A431 cells treated by GL-V9. The GC/MS analysis of the A431 cells extracts revealed a large number of peaks ( Figure 3). Deconvolution of the chromatograms The connection between autophagy and apoptosis or other forms of cell death is a burgeoning area of research. To investigate whether GL-V9 induced apoptosis and autophagy were synergetic or antagonistic, we used caspase inhibitor Z-VAD-FMK and autophagy inhibitor 3-MA combined with GL-V9. As shown in Figure 2D, Z-VAD-FMK reversed the apoptosis inducing effect of GL-V9, but had little influence in the GL-V9-induced increase of autophagy marker LC3-II. Different with Z-VAD-FMK, after cell autophagy was inhibited by 3-MA, the cleaved of caspase 3 induced by GL-V9 was increased ( Figure 2E). These results suggested that the autophagy induced by high concentration of GL-V9 potentially antagonized the effect of apoptosis induction, which might be associated with the influence of cell metabolism. Instead, the overall anti-cSCC effects of GL-V9 were not influenced by the dual regulations of apoptosis and autophagy.

GL-V9 Induced Metabolome Changes in A431 Cells
Then we studied the metabolomics of A431 cells treated by GL-V9. The GC/MS analysis of the A431 cells extracts revealed a large number of peaks ( Figure 3). Deconvolution of the chromatograms produced a total of 168 distinct peaks and 88 were authentically identified by comparing the mass spectrum of the peak with that available in the libraries and that of the reference metabolites. These included carbohydrates, fatty acids, amino acids, lipids, and amines. To obtain quantitative data, a characteristic mass (m/z) was selected for each peak, and the peak area was obtained for each inverse volume peak/molecule. produced a total of 168 distinct peaks and 88 were authentically identified by comparing the mass spectrum of the peak with that available in the libraries and that of the reference metabolites. These included carbohydrates, fatty acids, amino acids, lipids, and amines. To obtain quantitative data, a characteristic mass (m/z) was selected for each peak, and the peak area was obtained for each inverse volume peak/molecule.  Visual inspection of GC/MS profiles of metabolites in A431 cells of the control group at different GL-V9 concentrations revealed a difference between all administration groups and the control group. Based on the data matrix (with the two vectors of observations/samples and variables/molecules), an unsupervised PCA model was applied to overview the dataset. We found that there was a significant migration trend along with the increase of GL-V9 concentration ( Figure 4A). This result suggested that the metabolites and metabolic pathways were changed by GL-V9. In order to more effectively understand the metabolites and metabolic pathways most likely to be changed by GL-V9 in A431 cells, we assayed significant changes of 31 metabolites, involving carbohydrates, amino acids, small organic acids, fatty acids, lipids, and amines. Heatmap were generated for each sample species to visualize the intensities of differential metabolites in different groups ( Figure 4B). Further metabolic impact analysis of these selected metabolites showed that GL-V9 had main interference effect on carbohydrate metabolism, tricarboxylic acid cycle (TCA), and pyruvate metabolism ( Figure 4C). Enrichment analysis also showed that GL-V9 affected glucose metabolism, Warburg effect, amino acid metabolism, energy metabolism, and mitochondrial function ( Figure 4D). According to enrichment analysis, GL-V9 had the greatest influence on glucose metabolism. By investigation of two key substances of glycolysis pathway, glucose and glucose-6-phosphate, we found that intracellular glucose was accumulated but glucose-6-phosphate was decreased upon GL-V9 treatment ( Figure 4E). These results suggested that GL-V9 might affect glucose utilization and mitochondrial function of A431 cells. The influence of GL-V9 in the glycolysis resulted in the nutrient deficiency, which may be associated with the increased apoptosis by autophagy inhibitor 3-MA. Visual inspection of GC/MS profiles of metabolites in A431 cells of the control group at different GL-V9 concentrations revealed a difference between all administration groups and the control group. Based on the data matrix (with the two vectors of observations/samples and variables/molecules), an unsupervised PCA model was applied to overview the dataset. We found that there was a significant migration trend along with the increase of GL-V9 concentration ( Figure 4A). This result suggested that the metabolites and metabolic pathways were changed by GL-V9. In order to more effectively understand the metabolites and metabolic pathways most likely to be changed by GL-V9 in A431 cells, we assayed significant changes of 31 metabolites, involving carbohydrates, amino acids, small organic acids, fatty acids, lipids, and amines. Heatmap were generated for each sample species to visualize the intensities of differential metabolites in different groups ( Figure 4B). Further metabolic impact analysis of these selected metabolites showed that GL-V9 had main interference effect on carbohydrate metabolism, tricarboxylic acid cycle (TCA), and pyruvate metabolism ( Figure 4C). Enrichment analysis also showed that GL-V9 affected glucose metabolism, Warburg effect, amino acid metabolism, energy metabolism, and mitochondrial function ( Figure 4D). According to enrichment analysis, GL-V9 had the greatest influence on glucose metabolism. By investigation of two key substances of glycolysis pathway, glucose and glucose-6-phosphate, we found that intracellular glucose was accumulated but glucose-6-phosphate was decreased upon GL-V9 treatment ( Figure 4E). These results suggested that GL-V9 might affect glucose utilization and mitochondrial function of A431 cells. The influence of GL-V9 in the glycolysis resulted in the nutrient deficiency, which may be associated with the increased apoptosis by autophagy inhibitor 3-MA.

GL-V9 Induces Apoptosis by Inhibiting AKT-Regulated Mitochondrial Location of HK2, and Caused Autophagy via Suppression of AKT/mTOR Pathway
Hexokinase II (HK2) is dominant isoform of HK in cancer cells, which plays important roles in cell survival and glycolysis [26][27][28]. Because GL-V9 promoted the accumulation of intracellular glucose and decreased glucose-6-phosphate, GL-V9 might influence HK2. We found that GL-V9 downregulated the protein level of HK2, but had no significant influence in its mRNA expression ( Figure 5A,B). To investigate that whether GL-V9 influenced the protein stability, we co-treated A431 cells with GL-V9, proteasome inhibitor MG-132 and protein synthesis inhibitor cycloheximide (CHX). As shown in Figure 5C, after CHX blocked the synthesis of HK2, MG-132 reversed the downregulation of HK2 protein by GL-V9. Thus, GL-V9 promoted the protein degradation of HK2 to decrease its level. It is reported that HK2 can bind with VDAC, localize in mitochondria outer membrane and maintain the integrity of mitochondrial structure, leading to the inhibition of apoptosis and promotion of aerobic glycolysis in cancer cells [29]. GL-V9 decreased the level of HK2 in both mitochondria and cytoplasm ( Figure 5D). Meanwhile, GL-V9 promoted the dissociation of HK2 with VDCA ( Figure 5E). Immunofluorescence assay also showed that the mitochondrial location of HK2 was decreased ( Figure 5F).

GL-V9 Induces Apoptosis by Inhibiting AKT-Regulated Mitochondrial Location of HK2, and Caused Autophagy via Suppression of AKT/mTOR Pathway
Hexokinase II (HK2) is dominant isoform of HK in cancer cells, which plays important roles in cell survival and glycolysis [26][27][28]. Because GL-V9 promoted the accumulation of intracellular glucose and decreased glucose-6-phosphate, GL-V9 might influence HK2. We found that GL-V9 downregulated the protein level of HK2, but had no significant influence in its mRNA expression ( Figure 5A,B). To investigate that whether GL-V9 influenced the protein stability, we co-treated A431 cells with GL-V9, proteasome inhibitor MG-132 and protein synthesis inhibitor cycloheximide (CHX). As shown in Figure 5C, after CHX blocked the synthesis of HK2, MG-132 reversed the downregulation of HK2 protein by GL-V9. Thus, GL-V9 promoted the protein degradation of HK2 to decrease its level. It is reported that HK2 can bind with VDAC, localize in mitochondria outer membrane and maintain the integrity of mitochondrial structure, leading to the inhibition of apoptosis and promotion of aerobic glycolysis in cancer cells [29]. GL-V9 decreased the level of HK2 in both mitochondria and cytoplasm ( Figure 5D). Meanwhile, GL-V9 promoted the dissociation of HK2 with VDCA ( Figure 5E). Immunofluorescence assay also showed that the mitochondrial location of HK2 was decreased ( Figure 5F). Molecules 2020, 25, x 7 of 18 It is reported that activated AKT (phosphorylated AKT) phosphorylates HK2 and promotes the mitochondrial location of HK2 [30]. We found that GL-V9 downregulated the total level of the AKT and p-AKT as well as their level in mitochondria ( Figure 6A,B). The mitochondrial location of p-AKT was decreased as well ( Figure 6C). Moreover, AKT inhibitor MK-2206 inhibited the binding of HK2 with VDAC in mitochondria in A431 cells, instead AKT activator SC79 promoted their binding ( Figure 6D). These results suggested that GL-V9 inhibited AKT expression and activity, and affected mitochondrial localization of HK2, which was responsible for the mitochondria-mediated apoptosis induced by GL-V9.
AKT/mTOR signaling pathway is one of the classical pathways to regulate autophagy [31]. Therefore, we assayed the influences of GL-V9 in AKT/mTOR pathway in A431 cells. Western blot studies showed that the expression and activation of mTOR were both inhibited by GL-V9 ( Figure  6E), because mTOR specifically phosphorylates the p70S6 kinase at Thr-389, the phosphorylation of p70S6 kinase at this position is a routine and specific assay for monitoring mTOR activity [32]. Upon the treatment of GL-V9, the total protein level and the phosphorylation of p70S6K were also reduced ( Figure 6E). Thus, GL-V9 induced autophagy via inhibiting AKT/mTOR/p70S6 pathway.
Thus AKT played key roles in the anti-cSCC effect of GL-V9. When A431 cells were co-treated with GL-V9 and AKT activator SC79, the GL-V9-induced changes in caspase 3, p-mTOR, and LC3 were all reversed by SC79 ( Figure 6F). All above, GL-V9 induced apoptosis and autophagy of A431 cells by inhibiting the AKT-regulated HK2 and mTOR signals, respectively. It is reported that activated AKT (phosphorylated AKT) phosphorylates HK2 and promotes the mitochondrial location of HK2 [30]. We found that GL-V9 downregulated the total level of the AKT and p-AKT as well as their level in mitochondria ( Figure 6A,B). The mitochondrial location of p-AKT was decreased as well ( Figure 6C). Moreover, AKT inhibitor MK-2206 inhibited the binding of HK2 with VDAC in mitochondria in A431 cells, instead AKT activator SC79 promoted their binding ( Figure 6D). These results suggested that GL-V9 inhibited AKT expression and activity, and affected mitochondrial localization of HK2, which was responsible for the mitochondria-mediated apoptosis induced by GL-V9.
AKT/mTOR signaling pathway is one of the classical pathways to regulate autophagy [31]. Therefore, we assayed the influences of GL-V9 in AKT/mTOR pathway in A431 cells. Western blot studies showed that the expression and activation of mTOR were both inhibited by GL-V9 ( Figure 6E), because mTOR specifically phosphorylates the p70S6 kinase at Thr-389, the phosphorylation of p70S6 kinase at this position is a routine and specific assay for monitoring mTOR activity [32]. Upon the treatment of GL-V9, the total protein level and the phosphorylation of p70S6K were also reduced ( Figure 6E). Thus, GL-V9 induced autophagy via inhibiting AKT/mTOR/p70S6 pathway.
Thus AKT played key roles in the anti-cSCC effect of GL-V9. When A431 cells were co-treated with GL-V9 and AKT activator SC79, the GL-V9-induced changes in caspase 3, p-mTOR, and LC3 were all reversed by SC79 ( Figure 6F). All above, GL-V9 induced apoptosis and autophagy of A431 cells by inhibiting the AKT-regulated HK2 and mTOR signals, respectively. Molecules 2020, 25, x 8 of 18

GL-V9 Suppresses the Development of Primary Skin Cancer in Mice
To investigate the in vivo anticancer effect of GL-V9, we performed a mouse two-stage chemicalinducible primary skin cancer model ( Figure 7A). In the 18th week of the two-stage chemicalinducible skin cancer model, we can observed cutaneous lesions of papilloma in the back skin of mice. As shown in Figure 7B  (E) Western blot assays were used to examine the protein expression of mTOR, p-mTOR, p70s6k, and p-p70s6k. (F) A431 cells were co-treated with SC79 and GL-V9 for 36 h. The expression of key proteins involved with apoptosis and autophagy were assayed.

GL-V9 Suppresses the Development of Primary Skin Cancer in Mice
To investigate the in vivo anticancer effect of GL-V9, we performed a mouse two-stage chemical-inducible primary skin cancer model ( Figure 7A). In the 18th week of the two-stage chemical-inducible skin cancer model, we can observed cutaneous lesions of papilloma in the back skin of mice. As shown in Figure 7B-E, though the treatment group of low-dose GL-V9 (200 µL of 0.1M GL-V9 solution containing 8.2 mg GL-V9) did not shows a decent anticancer effect, high dose of GL-V9 (200 µL of 0.25M GL-V9 solution containing 20.5 mg GL-V9) had equal effect with fluorouracil (0.2 g of 5% fluorouracil cream containing 10 mg 5-Fu), and could effectively reduce the number and size of cutaneous lesions. The lesions areas skin were cut and used for histological hematoxylin-eosin (HE) staining analysis and immunohistochemistry (IHC) assay. HE staining analysis further indicated that GL-V9 significantly alleviate the increase in hyperplasia caused by two-stage chemical-inducible ( Figure 7F). We can observed a large amount of necrotic tissue in the groups with fluorouracil and high-dose of GL-V9 treatment. All these results showed that GL-V9 had potent capacity to suppress the progress of primary skin cancer in mice. the progress of primary skin cancer in mice.
IHC assay was done to further ascertain the effect of GL-V9 on key proteins in vivo ( Figure 7G). Ki67, as an important indicator of tumor growth [33]. In the primary skin cancer model group, compared with control group, protein expressions of ki67, AKT, p-AKT, mTOR, p-mTOR, and HKII were all significantly increased. After treatment of high-dose GL-V9 and the fluorouracil, the expression of Ki67, p-AKT, mTOR, p-mTOR, and HK2 were all decreased compared with model group, instead cleaved-caspase 3 were relative increased. These results showed that GL-V9 inhibited the growth of primary skin cancer in mice via suppressing AKT-regulated HK2 and mTOR signals in vivo.

Discussion
Cutaneous squamous cell carcinoma is an adverse outcome caused by the interaction of many factors, including environmental factors and self-factors [34]. As the elderly population grows and skin cancer screening improving, the incidence rates of cSCC are rising at a rate of 2.5-10% per year, which is becoming a public health problem [35,36]. It is generally accepted that most cSCC can be successfully treated with standard treatments, such as surgical excision, external drug application and local injection therapy. However, a subset of patients are still at risk for local recurrence, peripheral spread, and even metastasis to lymph node or distant organs, especially in individuals with compromised immune function [37]. Imiquimod and 5-Fluorouracil, clinically used for local IHC assay was done to further ascertain the effect of GL-V9 on key proteins in vivo ( Figure 7G). Ki67, as an important indicator of tumor growth [33]. In the primary skin cancer model group, compared with control group, protein expressions of ki67, AKT, p-AKT, mTOR, p-mTOR, and HKII were all significantly increased. After treatment of high-dose GL-V9 and the fluorouracil, the expression of Ki67, p-AKT, mTOR, p-mTOR, and HK2 were all decreased compared with model group, instead cleaved-caspase 3 were relative increased. These results showed that GL-V9 inhibited the growth of primary skin cancer in mice via suppressing AKT-regulated HK2 and mTOR signals in vivo.

Discussion
Cutaneous squamous cell carcinoma is an adverse outcome caused by the interaction of many factors, including environmental factors and self-factors [34]. As the elderly population grows and skin cancer screening improving, the incidence rates of cSCC are rising at a rate of 2.5-10% per year, which is becoming a public health problem [35,36]. It is generally accepted that most cSCC can be successfully treated with standard treatments, such as surgical excision, external drug application and local injection therapy. However, a subset of patients are still at risk for local recurrence, peripheral spread, and even metastasis to lymph node or distant organs, especially in individuals with compromised immune function [37]. Imiquimod and 5-Fluorouracil, clinically used for local lesion treatment, usually cause erythema, erosion, and scab, and patients' compliance [38]. Therefore, it is of great clinical value to develop a novel candidate for cSCC treatment.
In this study, we found a synthesized flavonoid GL-V9 derived from natural product wogonin, which had a potent therapeutic effect on cSCC. GL-V9 not only induced apoptosis and autophagy of human cSCC cell line A431 cells, but also suppressed the development of chemical induced primary skin cancer in mice. Besides, GL-V9 caused mitochondrial dysfunction and suppressed glycolysis to reprogram cell metabolism. Mechanism studies showed that AKT played important roles in the anti-cSCC effects of GL-V9, which inhibited AKT-regulated mitochondrial location of HK2 to induce apoptosis, and suppressed AKT/mTOR pathway to activate autophagy (Figure 8).
Molecules 2020, 25, x 10 of 18 lesion treatment, usually cause erythema, erosion, and scab, and patients' compliance [38]. Therefore, it is of great clinical value to develop a novel candidate for cSCC treatment.
In this study, we found a synthesized flavonoid GL-V9 derived from natural product wogonin, which had a potent therapeutic effect on cSCC. GL-V9 not only induced apoptosis and autophagy of human cSCC cell line A431 cells, but also suppressed the development of chemical induced primary skin cancer in mice. Besides, GL-V9 caused mitochondrial dysfunction and suppressed glycolysis to reprogram cell metabolism. Mechanism studies showed that AKT played important roles in the anti-cSCC effects of GL-V9, which inhibited AKT-regulated mitochondrial location of HK2 to induce apoptosis, and suppressed AKT/mTOR pathway to activate autophagy (Figure 8). In recent years, many studies focus on the anticancer effects of the active ingredients of natural products, such as flavonoids, polyphenols, polysaccharide, and so on [39][40][41]. Wogonin, as the precursor of GL-V9, showed obvious pharmacological effects, including anticancer effect. However, the low water-solubility and bioavailability of wogonin limited its development. GL-V9 not only inhibited the cell growth of liver cancer and breast cancer [42,43], also had better solubility and druggability than wogonin [44]. Whereas the effects of GL-V9 in cSCC has not been investigated before. We found that GL-V9 had potent anticancer activity in primary cSCC of mice. High dose of GL-V9 had equal effect with 5-Fu, instead had much lower toxicity than 5-Fu. The toxicity and side effects of 5-Fu greatly limit its therapeutic effect and clinical usage. In addition to some common side effects-including application site reactions (such as redness, burning, erosion pain), alopecia, sinus infection, and so on-fluorouracil cream has severe bone marrow suppression, which will reduce the body's immune system and greatly increase the death risk. In vivo studies, compared with 5-Fu, none of the mice treated with GL-V9 exhibited any physical discomfiture. Our previous reports also showed the low toxicity of GL-V9 in vivo [45]. No abnormal hematological parameters and morphological changes were observed in the organs of the tumor-bearing mice that were treated with GL-V9. Thus, the development of GL-V9 might provide a potential candidate with less toxicity for cSCC treatment.
AKT promotes the glycolysis and apoptosis via the regulation of mitochondrial HK2 [14,46]. Here, GL-V9 downregulated the protein level and activity of AKT in A431 cells. A previous research have found that GL-V9 induces the lysosome-dependent degradation of AKT1 [20]. Our recent In recent years, many studies focus on the anticancer effects of the active ingredients of natural products, such as flavonoids, polyphenols, polysaccharide, and so on [39][40][41]. Wogonin, as the precursor of GL-V9, showed obvious pharmacological effects, including anticancer effect. However, the low water-solubility and bioavailability of wogonin limited its development. GL-V9 not only inhibited the cell growth of liver cancer and breast cancer [42,43], also had better solubility and druggability than wogonin [44]. Whereas the effects of GL-V9 in cSCC has not been investigated before. We found that GL-V9 had potent anticancer activity in primary cSCC of mice. High dose of GL-V9 had equal effect with 5-Fu, instead had much lower toxicity than 5-Fu. The toxicity and side effects of 5-Fu greatly limit its therapeutic effect and clinical usage. In addition to some common side effects-including application site reactions (such as redness, burning, erosion pain), alopecia, sinus infection, and so on-fluorouracil cream has severe bone marrow suppression, which will reduce the body's immune system and greatly increase the death risk. In vivo studies, compared with 5-Fu, none of the mice treated with GL-V9 exhibited any physical discomfiture. Our previous reports also showed the low toxicity of GL-V9 in vivo [45]. No abnormal hematological parameters and morphological changes were observed in the organs of the tumor-bearing mice that were treated with GL-V9. Thus, the development of GL-V9 might provide a potential candidate with less toxicity for cSCC treatment.
AKT promotes the glycolysis and apoptosis via the regulation of mitochondrial HK2 [14,46]. Here, GL-V9 downregulated the protein level and activity of AKT in A431 cells. A previous research have found that GL-V9 induces the lysosome-dependent degradation of AKT1 [20]. Our recent studies also show that GL-V9 may influence the protein stability of AKT by disturbing the binding between Hsp90 and its client proteins (data not shown). The specific mechanism still need further investigation. The suppression of AKT by GL-V9 results in the decrease of mitochondrial HK2. HK2 is highly expressed in various cancers, which binds with the VDAC on the outer mitochondria, enhances aerobic glycolysis and resists apoptosis of cancer cells [27,47]. Differently from normal cells, cancer cells usually tend to choose the aerobic glycolysis to produce energy for cell growth. Metabolism reprograming is considered as a unique characteristic of cancer cells and promising target for cancer treatment [48]. In this study, we analyzed the influences of GL-V9 in the metabolism of A431 cells via metabolomics, and found that GL-V9 reprogrammed the metabolism of cancer cells, mainly influencing glycometabolism, amino acid metabolism, and mitochondrial metabolism. GL-V9 tended to inhibit glycolysis, which might be aroused by suppressing mitochondrial HK2. We could conclude that GL-V9 has significant influence in mitochondrial function of cancer cells. The damage of mitochondria is always accompanied by the changes of intracellular oxidative stress. It has been reported that GL-V9 suppresses thioredoxin-1 and increases reactive oxygen species (ROS) level of hepatoma cells by significantly promoting intracellular O 2 •− level, but not affecting H 2 O 2 production [7,45]. Although the anticancer efficacy of GL-V9 is better than its precursor wogonin, some reports show that the anticancer effects of GL-V9 is similar to wogonin, such as induction of apoptosis and ROS production [49], and inhibition of PI3K/AKT pathway [50]. However, whether the anti-cancer target of GL-V9 and wogonin are same and has cancer type specificity, still need more investigations. In addition to cell apoptosis, HK2 and AKT are also involved with the regulation of autophagy. It is reported that HK2 prevented autophagy-driven monocyte differentiation [51]. HK2 promotes starvation-induced autophagy by binding to, and inhibiting, TORC1 [52]. Studies have shown that the AKT/mTOR signaling pathway inhibits autophagy death of cancer cells by regulating the transcription and translation of growth promoting oncogenes or proteins [53]. High concentration of GL-V9 inhibits AKT/mTOR pathway, and induces apoptosis as well as autophagy. Recently, the relationship between apoptosis and autophagy are still complicated and confused. A flavonoid named oroxylin A is the isomer of wogonin. It was reported that pharmacological inhibition of autophagy suppressed apoptotic death of HepG2 cells treated with oroxylin A. In this situation, cell apoptosis cooperates with autophagy to promote the death of cancer cells. Instead, inhibition of autophagy by 3-MA promoted the apoptosis induced by GL-V9, which may be explained by the influences of autophagy in apoptotic signals. Some studies report that the clear of apoptosis cell by autophagy is a selective mechanism for cancer cells, which can help cancer cells overcoming the death stress caused by chemotherapy and result in drug resistance [54,55]. Therefore, the further studies of GL-V9 in cancer treatment should take into account of the double effects of apoptosis and autophagy. Co-treatment of autophagy inhibitors or metabolic regulators with GL-V9 will be explored, which may be significantly enhance the therapeutic efficacy.

Cell Culture
Then the normal mice were as control, and the tumor-bearing mice were randomly divided into four groups, including model, GL-V9 0.1M, GL-V9 0.25M, and fluorouracil groups (n = 10 per group). GL-V9 (dissolved in acetone solvent, 0.25 M and 0.1 M, 200 µL/ per time containing 20.5 mg and 8.2 mg GL-V9, respectively) and fluorouracil Ointment (5%: 10 g, 0.2 g cream/ per time containing 10 mg 5-Fu) were applied to the lesion area three times a week. The choice of GL-V9 dose was referred to the compared growth inhibitory effects of GL-V9 and 5-Fu in mice hepatoma reported previous [57]. During the period of model establishment and drug treatment, the mice of control group were applied with same volume of acetone on the back skin, parallelly. In 18 weeks, all mice were killed and their back skin was taken for hematoxylin-eosin and immunohistochemical staining testing.

MTT Assay
A431 cells were treated with 1.0-100 µM GL-V9 for 24, 36, and 48 h, respectively. After incubation, 20 µL of 5 mg/mL MTT was added. Four hours later, the culture medium was removed and the formed formazan was dissolved by 100 µL DMSO per well. The absorbance at 570 nm was measured by Universal Microplate Reader (EL800, BIO-TEK Instruments Inc., Winooski, VT, USA).
Cell Inhibition ratio (I%) = (A control − A treated )/A control × 100% (1) where A control and A treated are the average absorbance of three parallel experiments from treated and control groups, respectively. The IC 50 was taken as the concentration that caused 50% inhibition of cell proliferation. To determine the IC 50 value, we fit the data using an equation for a sigmoidal dose-response provided by GraphPad™ Prism ® 4.0.

Apoptosis Assay
A431 cells were treated with GL-V9 for 36 h and harvested. Apoptotic cells were identified by double supravital staining with recombinant FITC (fluorescein isothiocyanate)-conjugated Annexin-V and PI, using the Annexin V-FITC Apoptosis Detection kit (BioVision, Inc., Milpitas, CA, USA) according to the manufacturer's instructions. Flow cytometric analysis was performed immediately after supravital staining. Data acquisition and analysis were performed in a Becton-Dickinson FACSCalibur flow cytometer using Cell Quest software Pro V5.2.

DAPI Staining
Administrated A431 cells were seeded on to glass coverslips processed for immunofluorescence. The glass coverslips were washed twice with cold PBS for 5 min, fixed with 4% paraformaldehyde (PFA) for 30 min, and incubated with Triton X-100 for 10 min. After incubation, the A431 cells were blocked with PBS containing 3% BSA for 1 h, and then the coverslips were stained with diamidino-phenyl-indole (DAPI, purchased from Beyotime Biotechnology Co., Hangzhou, China) for 30 min. The images were captured with an Olympus FV1000 confocal microscope (Olympus Corp., Tokyo, Japan).

Mitochondrial Membrane Potential (MMP) Assay
The loss of MMP was measured by the Mitochondrial Membrane Potential Detection kit (KeyGen Biotech Co., Ltd., Nanjing, China). After being incubated with GL-V9 for 36 h, all floating and attached cells were harvested and resuspended with ice-cold PBS (2000 rpm × 5 min). Then the cell suspensions were incubated in JC-1 prepared with 1× Incubation Buffer for 20 min at 37 • C, and detected by FACSCalibur flow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA).

Extraction of Mitochondrial and Cytosolic Fractions
The mitochondrial and cytosolic fractions of A431 cells were performed using mitochondria/cytosol fractionation kit (KeyGen Biotech, Nanjing, China) according to the following protocol. The A431 cells were administrated with different concentrations of GL-V9 for 36 h and were collected and keep warm in 100 mL ice-cold mitochondrial lyses buffer for 10 min. The suspension of A431 cell was homogenized for strike with a tight pestle. The homogenate was subjected to centrifuging at 600× g for 10 min at 4 • C to remove nuclei and unbroken cells. Then the collection was harvested and centrifuged again at 12,000× g for 30 min at 4 • C to obtain the supernatant of cytosol and deposition of mitochondria fraction. Samples of cytosol and mitochondria were dissolved in lyses buffer at −20 • C.

Immunofluorescence
Cells were collected and seeded onto glass coverslips processed for immunofluorescence. The glass coverslips were washed twice with cold PBS for 5 min, fixed with 4% paraformaldehyde for 20 min and incubated with 0.2% Triton X-100 for 5 min at 4 • C. After incubation, the cells were blocked with PBS containing 3% BSA for 1 h and incubated with anti-HK2 antibody (1:100, abclonal), anti-p-AKT antibody (1:100, abclonal) as well as anti-TOM20 antibody (1:400, abcam) overnight. After being washed twice with cold PBS for 10 min, the cells were stained with FITC-conjugated Goat Anti-Rabbit and PE-conjugated Goat Anti-Mouse IgG second antibody (1:500, abcam) for 1 h. The images were captured with a confocal microscope (Olympus FV1000, Olympus Corp., Tokyo, Japan).

Real-Time PCR Analysis
Total RNA was extracted using TriPure Isolation Reagent (Roche Diagnostics, Mannheim, Germany). One microgram of total RNA was used to transcribe the first strand cDNA with SuperScript II reverse transcriptase (Invitrogen). Real-time PCR was completed on an ABI PRISM Sequence Detector 7500 (PerkinElmer, Branchburg, NJ, USA) using Sequence Detector version 1.7 software (Applied Biosystems, Foster City, CA, USA). SYBR Green PCR Master Mix was purchased from Applied Biosystems. Forward and reverse primers for targeted mRNA were designed and purchased from TAKARA BiotechnologyCo., Ltd. (Dalian, China). The primer sets used in the PCR assay were as follows: HK2-sense: 5 -GGCTCTGGACAGGTGGTAAAGA-3 ; HK2-antisense: 5 -CGGTAATGCACCACCTTGGTGT-3 ; GAPDH -sense: 5 -TAGTGGAAGGACTCATGACC-3 ; GAPDH-antisense: 5 -TCCACCACCCTGTTGCTGTA-3 Fold change of mRNA level was calculated. After completion of the PCR, the baselines and thresholds were set for both samples and internal GAPDH control.

Immunohistochemical Staining
For IHC analysis, mice skin was collected and paraformaldehyde fixed, paraffin-embedded sections of skin tissues (4 µm thick) were mounted on slides coated with 2-aminopropyltriethoxysilane, which then were baked, deparaffinized, rinsed with 3% hydrogen peroxide, and incubated with proteinase K (0.5 mg/mL). After that, these sections were washed and then blocked with StartingBlockTM blocking buffers (Pierce, Rockford, IL, USA) for 5 min and subsequently incubated with corresponding polyclonal antibodies for 30 min. Finally, the sections were incubated with Strept Avidin-Biotin Complex (Solarbio) for 30 min at room temperature, followed by detection with a 3,3-diaminobenzidine tetrahydrochoride solution (chromogen) (ZSGB-BIO) and hematoxylin (counterstain). Sections were further mounted with neutral gums. IHC sections were photographed by Mantra 1.01 (PerkinElmer, Waltham, MA, USA).

Metabolomics
For metabolomics analysis, cells and medium samples were pretreated, extracted, and derivatized as previously reported [57]. Chromatographic separation of the analysts was achieved with in Shimadzu GCMSQP2010 (Shimadzu Corp., Tokyo, Japan) equipped with a RTx-5MS column (30 mm × 0.25 mm i.d. fused-silica capillary column chemically bond with a 0.25 µm cross bond, 5% diphenyl/ 95% dimethyl polysiloxane, Restek Corporation, PA, USA). The raw data acquired with GC/MS were processed, and the metabolites were identified. After identification, to selected one feature ion as the quant mass, and the peak area was acquired for each peak, respectively. In addition, metabolic pathway and enrichment analysis were performed by inputting discriminant molecules into Metaboanalyst 4.0.

Multivariate Statistical Analysis
The relative quantitative data for each peak (peak area) was first normalized against the IS, and then the data matrix was constructed by the normalized peak areas with the sample names as observations in the first column, and retention times/peaks as the response variables in the first row. The heatmap was assessed using the R language with own coded. The multivariate data was evaluated using the SIMCA P 13.0 software (Umetrics, Umeå, Sweden).

Statistical Analysis
Values were expressed as the mean ± SD of at least three independent experiments. One way analysis of variance (ANOVA) was used to compare in groups.

Conclusions
Collectively, our findings demonstrate that the GL-V9 is a potent candidate for the treatment of local cutaneous squamous cell carcinoma via induction of apoptosis and autophagy. Inhibition of AKT plays important roles in the anti-cSCC effects of GL-V9, which leads to the suppression of AKT-mediated HK2 location in mitochondrial and AKT/mTOR pathway. The findings that GL-V9 has double regulation in apoptosis and autophagy will provide some insights for the therapeutic strategies of cSCC.

Conflicts of Interest:
The authors declare no conflict of interest.