**About the Editor**

**Claudio Luparello** https://www.unipa.it/persone/docenti/l/claudio.luparello/en/.

### *Editorial* **Role of Natural Bioactive Compounds in the Rise and Fall of Cancers**

#### **Claudio Luparello**

Department of Biological, Chemical and Pharmaceutical Sciences and Technologies, University of Palermo, Viale delle Scienze, Edificio 16, 90128 Palermo, Italy; claudio.luparello@unipa.it; Tel.: +39-091-238-97405

Received: 31 August 2020; Accepted: 2 September 2020; Published: 3 September 2020

Recent years have seen the idea of a close association between nutrition and the modulation of cancer development/progression reinforced. In fact, an increasing number of experimental and epidemiological evidence has been produced, supporting the concept that many different bioactive components of food (e.g., polyphenols, mono- and polyunsaturated fatty acids, methyl-group donors ... ) may be implicated in either the promotion of or the protection against carcinogenesis. At the cellular level, such compounds can have an impact on different but sometimes intertwined processes, such as growth and differentiation, DNA repair, programmed cell death, and oxidative stress. In addition, compelling evidence is starting to build up of the existence of primary epigenetic targets of dietary compounds, such as oncogenic/oncosuppressor miRNAs or DNA-modifying enzymes, which in turn impair gene expression and function. This editorial aims to summarize the themes of the 31 papers (20 original articles and 11 reviews) published in the Special Issue "Role of Natural Bioactive Compounds in the Rise and Fall of Cancers" presenting the latest findings on the intracellular pathways and mechanisms affected by selected natural molecules influencing the fine-tuning of cancer phenotype.

Plant polyphenols have been among the most studied natural compounds by the contributors to this issue.

In the original article group, Polonio-Alcalà et al. [1] showed the additive and synergistic effects of the flavonoids (−)-epigallocatechin-3-gallate (EGCG) from green tea and its naphthalene derivative G28, which are fatty acid synthase inhibitors, in combination with epidermal growth factor receptor tyrosine kinase inhibitors on gefitinib-resistant lung adenocarcinoma models. Moreover, Wei et al. [2] examined the effect and mechanism of action of EGCG alone and in combination with current chemotherapeutics on pancreatic cancer cell growth, demonstrated the impairment of cell proliferation via the phosphofructokinase inhibition-mediated suppression of glycolysis in a ROS-dependent manner, and the additive enhancement of the anticancer effect of gemcitabine both in vitro and in pancreatic xenografts by the further inhibition of glycolysis and the impairment of cell kinetics. In the Review section, Farooqi et al. [3] analyzed the pleiotropic abilities of EGCG to regulate intracellular signalizations such as those related to JAK/STAT, Wnt/β-catenin, TGF/SMAD, SHH/GLI and NOTCH pathways, also commenting on the ability of EGCG to modulate non-coding RNAs and the methylation-associated machinery in different cancers.

Other natural phenolic compounds whose activity is discussed in the original articles of this issue are:


pathway, GLUT-1, PKM2 and MCT4, likely resulting in a decreased glucose entrance and biomass production [5];


In the Review section, Perrone et al. [13] discussed the effects of polyphenols in preventing the progression of central and peripheral nervous system tumors underlining the beneficial effect of dietary compounds on the microbioma–intestine–brain axis. Barbosa and Martel [14] examined the role played by a wide variety of synthetic and natural substances, including polyphenols, on the impairment of glucose uptake by neoplastic breast cells thereby resulting in a tumor-restraining effect. Ong et al. [15] reported the broad-range in vitro/in vivo anticancer properties of the *Magnolia*-derived polyphenol honokiol based upon its ability to impair cell cycle progression, inhibit epithelial–mesenchymal transition, and suppress cell motility, invasion, metastasis and angiogenesis. Zhou et al. [16] summarized the late preclinical studies on the applications of bioactive polyphenols in lung cancer therapy, focusing on the molecular mechanisms at the basis of their therapeutic effects and also discussing the potential of the polyphenol-based combination therapy. Goh et al. [17] reviewed data on the anti-colon cancer properties of nobiletin, a polymethoxyflavone extracted from citrus peel, and its derivatives which are able to arrest the cell cycle, inhibit cell proliferation, induce apoptosis, prevent tumor formation, reduce inflammatory effects and limit angiogenesis, also exploring better drug delivery strategies due to the low oral bioavailability of the compounds. Ong et al. [18] focused their review on the pharmacological properties and therapeutic potential of formononetin [7-hydroxy-3-(4-methoxyphenyl)-4H-1-benzopyran-4-one], one of the main bioactive components of red clover, which regulates various transcription factor- and growth factor-mediated oncogenic pathways attenuating metastasis and tumor growth in vivo in multiple cancer cell models and also alleviating the possible causes of chronic inflammation that are linked to the cancer survival of neoplastic cells and their resistance against chemotherapy.

The other articles and reviews addressed further cancer-related issues relevant to types of compounds of a different nature, specifically:


In the Review section, Del Cornò et al. [28] discussed the modulatory effects of dietary β-glucans, present in diverse edible mushrooms, baker's yeast, cereals and seaweeds, on human innate immunity cells and their potential role in cancer control. Lee et al. [29] reviewed a large number of data on the role played by different cytokines, lipids and other natural molecules on the suppression of epithelial–mesenchymal transition in cancer progression. Ennour-Idrissi et al. [30] focused their attention on the bioaccumulation of persistent organic pollutants in the food chain and the association of exposure with breast cancer risk. Farooqi et al. [31] presented the current views about the ability of berberine, a natural alkaloid compound found in several medicinal plants, to target different signaling cascades in various cancers, also discussing the nanocarrier strategies developed to improve the delivery of the compound.

The number of manuscripts published in this Special Issue indicates an active interest in research about the molecular/pharmacological mechanisms used by natural products exerting anti-tumoral effects which deserve further and deeper studies. I wish to thank all the contributors of this issue for sharing with us their experimental or reviewed data which will surely attract readers' attention and encourage the publication of other high-quality papers in this field.

**Funding:** This research received no external funding.

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

#### **References**


© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

#### *Article*

### **Fatty Acid Synthase Inhibitor G28 Shows Anticancer Activity in EGFR Tyrosine Kinase Inhibitor Resistant Lung Adenocarcinoma Models**

**Emma Polonio-Alcalá 1,2,**†**, Sònia Palomeras 1,**†**, Daniel Torres-Oteros 3, Joana Relat 3,4,5, Marta Planas 6, Lidia Feliu 6, Joaquim Ciurana 2, Santiago Ruiz-Martínez 1,\* and Teresa Puig 1,\***


Received: 26 March 2020; Accepted: 16 May 2020; Published: 19 May 2020

**Abstract:** Epidermal growth factor receptor (EGFR) tyrosine kinases inhibitors (TKIs) are effective therapies for non-small cell lung cancer (NSCLC) patients whose tumors harbor an EGFR activating mutation. However, this treatment is not curative due to primary and secondary resistance such as T790M mutation in exon 20. Recently, activation of transducer and activator of transcription 3 (STAT3) in NSCLC appeared as an alternative resistance mechanism allowing cancer cells to elude the EGFR signaling. Overexpression of fatty acid synthase (FASN), a multifunctional enzyme essential for endogenous lipogenesis, has been related to resistance and the regulation of the EGFR/Jak2/STAT signaling pathways. Using EGFR mutated (EGFRm) NSCLC sensitive and EGFR TKIs' resistant models (Gefitinib Resistant, GR) we studied the role of the natural polyphenolic anti-FASN compound (−)-epigallocatechin-3-gallate (EGCG), and its derivative G28 to overcome EGFR TKIs' resistance. We show that G28's cytotoxicity is independent of TKIs' resistance mechanisms displaying synergistic effects in combination with gefitinib and osimertinib in the resistant T790M negative (T790M−) model and showing a reduction of activated EGFR and STAT3 in T790M positive (T790M+) models. Our results provide the bases for further investigation of G28 in combination with TKIs to overcome the EGFR TKI resistance in NSCLC.

**Keywords:** NSCLC; EGFR TKI; FASN inhibitors; resistance; STAT3; EGCG

#### **1. Introduction**

Lung cancer is the most incident and the leading cause of cancer death worldwide. Non-small cell lung cancer (NSCLC) subtype is the most common and it represents 80–85% of lung cancers diagnosed.

Early-stage NSCLC patients have long-term survival after surgery. However, approximately 75% of patients are diagnosed in advanced stages [1,2].

Gefitinib is a first generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI). It was approved in 2003 by the Food and Drug Administration (FDA) for the treatment of patients whose tumors harbor an EGFR sensitizing/activating mutation (EGFRm) i.e., exon 19 deletion (ΔE746-A750) or the point mutation in exon 21 that leads to the amino acid substitution L858R [3–5]. Despite this therapy represents a breakthrough in the treatment of EGFRm NSCLC patients, in a median time of 9–16 months nearly all patients do not achieve a complete response. One of the most common resistance mechanisms described is the EGFR point mutation in exon 20 that leads to the replacement of threonine 790 by methionine (T790M), which normally derives to lethal disease progression [6]. Osimertinib is an irreversible third generation TKI effective in EGFRm T790M positive (T790M+) patients. However, the point mutation C797S in exon 20 has appeared as the main resistance mechanism to the latest FDA-approved TKI [6]. Other mechanisms for EGFR TKI resistance include Met amplification, phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PI3KCA) mutations, appearance of stem-like properties as evidenced by increase in epithelial–mesenchymal transition (EMT) and histological transformation, epidermal growth factor receptor type 2 (ErbB2) gene amplification, increase of insulin-like growth factor 1 receptor (IGF-1R) signaling pathway through the loss of inhibitory insulin-like growth factor-binding protein (IGF-BP) and loss or reduction of phosphatase and tensin homolog (PTEN), activation of AXL tyrosine kinase receptor or B-Raf proto-oncogene, and serine/threonine kinase (BRAF) mutations [6–12].

Recently, the activation of a signal transducer and activator of transcription 3 (STAT3) has been described as an alternative resistance mechanism allowing cancer cells to escape the EGFR signaling or the TKI suppression [13]. STAT3 is involved in the transcription of many genes related to cell differentiation, proliferation, resistance to apoptosis, angiogenesis, metastasis, and immune response [14–16]. Besides being phosphorylated by EGFR [17], STAT3 can also be activated in response to different cytokines and growth factors such as interleukin 6 (IL-6), interferon-alpha (IFN-α) or epidermal growth factor (EGF), among others [18]. Approximately 60% of patients show STAT3 activation, which correlates with poorly differentiated tumors, the presence of metastasis, and the late clinical stage [19,20]. STAT3 phosphorylation has been related to disease progression in a small cohort of patients after EGFR TKI treatment [21]. Additionally, neither gefitinib nor osimertinib are able to inhibit STAT3 activation [22,23].

Energy metabolism deregulation has been described as a hallmark of cancer, allowing cell growth and proliferation [24]. Fatty acid synthase (FASN) is an essential enzyme for the de novo synthesis of long-chain fatty acids from acetyl-CoA, malonyl-CoA, and NADPH [25]. Unlike most normal cells, highly-proliferative cancer cells overexpress this lipogenic enzyme, being an interesting target in cancer therapy [26,27]. FASN is strongly associated to poor prognosis and resistance to treatment in different human tumors such as breast [28], bladder [29], pancreatic [30], or lung cancer [31]. Moreover, FASN overexpression has also been proposed as a multidrug resistance mechanism, protecting cells from drug-induced apoptosis through the overproduction of palmitic acid [32]. The natural compound (−)-epigallocatechin-3-gallate (EGCG) is a polyphenolic flavonoid from green tea that has been broadly studied for its cardiovascular, neuroprotective, anticancer, and antimicrobial properties [33,34]. EGCG has been reported to compete with NADPH to bind the β-ketoacyl reductase domain of FASN [35]. The ability of several FASN inhibitors to regulate the canonical EGFR/Jak2/STAT pathway has also been stated in the literature [36,37]. We and others have shown that FASN inhibition is mainly related to EGFR/HER2 signaling pathways, leading to cytotoxic effects in vitro and in vivo in a wide range of carcinomas, including breast and lung [38–42]. To date, many EGCG derivatives have been developed to improve efficacy and increase stability in physiological conditions. Among them, the naphthalene derivative G28 has shown interesting antiproliferative features against sensitive and resistant breast cancer cells [38,43,44].

The purpose of this work was to study the role of FASN inhibitors (EGCG and G28) to overcome TKI resistance in NSCLC. FASN inhibitors were tested alone and in combination with EGFR TKIs (gefitinib and osimertinib) in EGFRm NSCLC models resistant to EGFR TKIs (Gefitinib Resistant, GR). In addition, we also evaluated gene and protein expression changes of FASN, EGFR, and STAT3 resulting from treatments with FASN inhibitors and EGFR TKIs alone or in combination. We show that FASN inhibitor G28 cytotoxicity is independent of EGFR TKI resistance mechanisms. Interestingly, G28 compound exhibited a cytotoxic effect in combination with gefitinib showing changes in EGFR/STAT3 pathway in T790M positive (T790M+) GR models and strong synergism in combination with gefitinib or osimertinib in T790M negative (T790M−) GR model.

#### **2. Results**

#### *2.1. EGFRm GR NSCLC Models Are Sensitive to FASN Inhibition*

In order to study the role of FASN in the acquisition of EGFR TKI resistance in NSCLC, we used the sensitive PC9 cell line carrying the EGFR exon 19 deletion (ELREA) and three GR models, two T790M+ models (PC9-GR1 and PC9-GR4), and one T790M− model (PC9-GR3) [45].

#### 2.1.1. EGFRm NSCLC Models Overexpress FASN

Firstly, we determined whether EGFRm NSCLC models express FASN enzyme. Hence, FASN protein (Figure 1a) and mRNA expression levels (Figure 1b) were analyzed by immunoblotting and quantitative real time PCR (qRT-PCR), respectively.

**Figure 1.** FASN protein and mRNA expression levels in sensitive (PC9) and Gefitinib Resistant (GR) models (PC9-GR1, PC9-GR3, and PC9-GR4). (**a**) Western blot analysis (quantification in upper panel and bands in lower panel) of FASN protein expression. GAPDH was used as a loading control. Results shown are representative from at least three independent experiments. (**b**) FASN endogenous mRNA levels were obtained by qRT-PCR and normalized against the GAPDH gene. FASN expression in the sensitive cells was normalized to 1 an expressed as a fold change, to which all other conditions were compared. Results shown are mean ± SE from three independent experiments. \* *p* < 0.050, \*\*\* *p* < 0.001 indicate levels of statistically significance.

All models showed FASN protein and mRNA expression. Despite no differences in mRNA, GR models presented significantly higher protein expression levels (PC9-GR1 *<sup>p</sup>* <sup>=</sup> 8.710 <sup>×</sup> <sup>10</sup><sup>−</sup>4; PC9-GR3 *<sup>p</sup>* <sup>=</sup> 3.160 <sup>×</sup> <sup>10</sup><sup>−</sup>4, and PC9-GR4 *<sup>p</sup>* <sup>=</sup> 0.049) in comparison to PC9.

#### 2.1.2. PC9-GR3 Model Is Resistant to Gefitinib and Osimertinib

We confirmed the resistance to EGFR TKIs in PC9 and GR models. For that, we measured the cytotoxic effect of gefitinib and osimertinib on all models by determining the half-maximal inhibitory concentration (IC50) using the MTT assay (Figure 2).

**Figure 2.** Cell proliferation inhibition of EGFR TKIs (gefitinib and osimertinib) in parental and Gefitinib Resistant (GR) models. Sensitive (PC9) and GR models (PC9-GR1, PC9-GR3, and PC9-GR4) were treated with increasing concentrations of (**a**) gefitinib (from 2.5 <sup>×</sup> <sup>10</sup>−<sup>3</sup> to 1 <sup>μ</sup>M for PC9 and 1–40 <sup>μ</sup>M for GR models) and (**b**) osimertinib (0.02–2000 nM for PC9, PC9-GR1, and PC9-GR4 and 500–7500 nM for PC9-GR3) for 72 h. Results shown are expressed as percentage of surviving cells after drug treatment (mean ± SE) and are representative from at least three independent experiments.

As expected, GR models were significantly more resistant to gefitinib with IC50 values in the micromolar range compared to the nanomolar IC50 found in the PC9 cell line (PC9-GR1 *<sup>p</sup>* <sup>=</sup> 2.793 <sup>×</sup> <sup>10</sup><sup>−</sup>7; PC9-GR3 *p* = 1.631 <sup>×</sup> 10−10, and PC9-GR4 *p* = 1.000 <sup>×</sup> 10−6). Although no significant differences were found in the IC50 value for gefitinib between the two T790M+ GR models, the IC50 value of the PC9-GR3 model for gefitinib was significantly greater than PC9-GR1 (*<sup>p</sup>* <sup>=</sup> 7.953 <sup>×</sup> <sup>10</sup><sup>−</sup>7) and PC9-GR4 (*p* = 1.659 <sup>×</sup> 10<sup>−</sup>7). PC9-GR3 model was also resistant to osimertinib compared to other models (PC9 *<sup>p</sup>* <sup>=</sup> 2.799 <sup>×</sup> <sup>10</sup><sup>−</sup>9; PC9-GR1 *<sup>p</sup>* <sup>=</sup> 3.749 <sup>×</sup> <sup>10</sup><sup>−</sup>8, and PC9-GR4 *<sup>p</sup>* <sup>=</sup> 5.200 <sup>×</sup> <sup>10</sup><sup>−</sup>9).

#### 2.1.3. FASN Inhibitors Present Cytotoxic Effects in NSCLC Models

Cancer cells have been described to increase the de novo lipogenesis through the activation of FASN and its inhibition has proven to cause cell death. Therefore, this enzyme has become a promising candidate for the development of new anticancer therapies. Here we tested the cytotoxic activity of the two FASN inhibitors, EGCG and its derivative G28. MTT cell viability assays showed that the natural polyphenolic compound EGCG was cytotoxic for PC9 (IC50 = 77.9 ± 1.9 μM), PC9-GR1 (IC50 = 74.3 ± 4.3 μM), PC9-GR3 (IC50 = 91.0 ± 5.5 μM), and PC9-GR4 (IC50 = 75.6 ± 2.4 μM) NSCLC models with no significant differences (*p* = 0.358; Figure 3a).

The synthetic EGCG derivative G28 showed higher cytotoxicity in all tested models with IC50 of 12.8 ± 1.3 μM for PC9, 12.0 ± 0.8 μM for PC9-GR1, 17.8 ± 1.3 μM for PC9-GR3, and 11.2 ± 1.2 μM for PC9-GR4 (Figure 3b). Besides, only PC9-GR3 showed a significantly higher IC50 value compared to PC9 (*p* = 0.030), PC9-GR1 (*p* = 0.005), and PC9-GR4 (*p* = 0.002).

**Figure 3.** Cell proliferation inhibition of FASN inhibitors in parental and Gefitinib Resistant (GR) models. Sensitive (PC9) and GR models (PC9-GR1, PC9-GR3, and PC9-GR4) were treated with increasing concentrations of (**a**) EGCG (5–150 μM) and (**b**) G28 (2–40 μM) for 72 h. Results shown are expressed as the percentage of surviving cells after drug treatment (mean ± SE) and are representative from at least three independent experiments.

#### 2.1.4. G28 Inhibits FASN in EGFRm NSCLC Models

The ability to internalize and inhibit FASN activity of EGCG and G28 after being exogenously added to the media was tested in sensitive and GR models (Figure 4). EGCG and G28 inhibited FASN in PC9 cells, resulting in a similar FASN activity reduction of roughly 80% (*p* = 0.265). Moreover, G28 significantly reduced FASN activity in all GR models in comparison with EGCG (PC9-GR1 *<sup>p</sup>* <sup>=</sup> 3.000 <sup>×</sup> <sup>10</sup><sup>−</sup>5; PC9-GR3 *p* = 0.001; PC9-GR4 *p* = 0.008) while the EGCG compound was not able to diminish FASN activity.

**Figure 4.** G28 compound inhibits FASN activity in Gefitinib Resistant (GR) models. Cells were treated for 72 h with EGCG or G28 at a concentration equal to their IC50 and with DMSO as control. FASN activity was assayed by counting radiolabeled fatty acids synthesized de novo. Bars represent the remaining activity as a percentage in treated versus untreated cells (control, Ctrl). Data are mean ± SE from at least three independent experiments. \*\* *p* < 0.010, \*\*\* *p* < 0.001 indicate levels of statistically significance.

#### 2.1.5. Apoptosis Induction of FASN Inhibitors and EGFR TKIs Treatments

We also investigated whether the cell death caused by treatment with EGFR TKIs or FASN inhibitors was the result of apoptosis induction in both sensitive and GR models. Poly(ADP-ribose) polymerase (PARP) terminal proapoptotic protein activated after cleavage was used as an apoptosis

marker. The effects of all compounds on PARP was determined by Western blot analysis in all models (Figure 5). The uncropped Western blots can be found in Figure S2a.

**Figure 5.** Effects of FASN inhibitors and EGFR TKIs on apoptosis determined by PARP cleavage. Sensitive (PC9) and Gefitinib Resistant (GR) models (PC9-GR1, PC9-GR3, and PC9-GR4) were treated for 72 h with a concentration equivalent to IC50 of each drug. Untreated cells were used as an internal control (Ctrl) and GAPDH as a loading control. Results shown are representative from at least three independent experiments.

The cleaved form of PARP (89 kDa) appeared in either sensitive or resistant T790M− models after treatment with the IC50 concentration of both EGFR TKIs (gefitinib and osimertinib) and FASN inhibitors (EGCG and G28), indicating the induction of apoptosis. Gefitinib treatment led to PARP cleavage in T790M+ GR models. Additionally, EGCG and osimertinib led to the formation of cleaved PARP in PC9-GR1 model.

#### *2.2. G28 Increases EGFR Activation in EGFRm NSCLC Models*

FASN has been previously related to AKT/ERK/EGFR signaling pathways [46] and the inhibition of the transcription factor STAT3 [36] in lung adenocarcinomas. Thus, differences in FASN, EGFR, and STAT3 protein and mRNA expression levels after FASN inhibitors or EGFR TKIs treatment were analyzed through immunoblotting (Figure 6a) and qRT-PCR (Figure 6b) in PC9 and GR models. The uncropped Western blots can be found in Figure S2b.

A reduction of FASN mRNA expression levels was observed in sensitive and GR models treated with FASN inhibitors, being statistically significant in the PC9-GR4 treated with G28 (*p* = 0.004) and PC9 cells treated with G28 (*<sup>p</sup>* <sup>=</sup> 7.370 <sup>×</sup> <sup>10</sup><sup>−</sup>4) and gefitinib (*<sup>p</sup>* <sup>=</sup> 3.210 <sup>×</sup> <sup>10</sup><sup>−</sup>4). T790M<sup>+</sup> GR models presented a basal hyperactivation of EGFR that was inhibited after treatment with EGFR TKIs. Regarding FASN inhibitors, EGCG and G28 increased its activation in PC9 cells contrary to the PC9-GR1 model, while no changes were observed in PC9-GR3 and PC9-GR4 models.

EGCG reduced EGFR protein levels in sensitive and T790M+ GR models that did not correlate with mRNA expression levels. Contrary, G28 significantly increased EGFR mRNA expression in all models, without protein level modification. No changes in EGFR mRNA levels were observed after gefitinib treatment, however higher protein levels were detected in T790M− GR models. Osimertinib treatment did not lead to EGFR protein nor mRNA expression alteration. EGFR TKIs treatment increased STAT3 activation in all GR models. Gefitinib increased p-STAT3 levels in sensitive cells, while the highest STAT3 activation in GR models was found with osimertinib treatment. No STAT3 protein or mRNA expression differences were found in any of the models and treatments assayed.

Lung carcinomas are highly proliferative and resistance acquisition after EGFR TKI-based therapy is a major problem. Overproduction of palmitic acid by FASN emerges as a resistance mechanism, protecting cells from drug-induced apoptosis [47]. The combination of these drugs with a different mechanism of action may decrease resistance development and improve treatment response.

**Figure 6.** FASN, EGFR, and STAT3 protein and mRNA expression levels after FASN inhibitors (EGCG and G28) and EGFR TKIs (gefitinib and osimertinib) treatment in sensitive (PC9) and Gefitinib Resistant (GR) models (PC9-GR1, PC9-GR3, and PC9-GR4). (**a**) Western blot analysis of FASN, EGFR, and STAT3 protein expression after 72 h of FASN inhibitors (IC50) and EGFR TKIs (IC50) treatment in EGFR TKI sensitive and GR models. Untreated cells were used as an internal control (Ctrl) and GAPDH as a loading control. Results shown are representative from at least three independent experiments. (**b**) FASN, EGFR, and STAT3 mRNA levels after treatment with FASN inhibitors and EGFR TKIs in sensitive and GR models. mRNA levels were obtained by qRT-PCR and normalized against the GAPDH gene. All conditions were compared to control (untreated cells), which was normalized to 1 (indicated by the dotted line) and expressed as a fold change. Experiments were performed at least three times. \* *p* < 0.050, \*\* *p* < 0.010, \*\*\* *p* < 0.001, indicate levels of statistically significance.2.3. G28 Combined with EGFR TKIs Outcomes in Synergistic Effects.

Once the effects of FASN inhibitors and EGFR TKIs alone were analyzed in all models, the combinatorial treatment between FASN inhibitors and EGFR TKIs was also studied in order to evaluate the possible synergistic interactions (Figure 7 and Figure S1). Despite the fact that EGCG did not reduce FASN activity in GR models, no IC50 differences were found in comparison with sensitive cells. Therefore, GR models were treated with EGCG or G28 in combination with the EGFR TKI to which they were resistant. Three EGFR TKIs concentrations were chosen based on MTT assays (vide supra). The combination of EGCG with either gefitinib or osimertinib resulted in mostly additive effect as shown by the combination index (CI; Figure 7a). The combination of G28 with gefitinib generally led to additivism in T790M+ GR models, with some synergism found in G28 concentrations ranging from 10 to 20 μM. Remarkably, T790M− GR model treated with G28 combined with gefitinib or osimertinib showed greater synergistic effects (Figure 7b).

#### *2.3. G28 Reduces STAT3 Activation in T790M*+ *GR Models Alone or in Combination with Gefitinib*

In order to discern whether G28 is able to reduce the STAT3 activation produced by EGFR TKIs, a combinatorial analysis was performed (Figure 8). As before, FASN, EGFR, and STAT3 protein and mRNA levels were analyzed using Western blot (Figure 8a) and qRT-PCR (Figure 8b) in GR models treated with synergistic drug concentrations (all of them under the IC50 value). The uncropped Western blots can be found in Figure S2c. Therefore, GR models were treated with G28 at 15 μM in combination with 1 μM gefitinib, which is the highest clinically achievable plasma concentration [5]. Osimertinib-resistant PC9-GR3 model was co-treated with 15 μM G28 and 0.5 μM osimertinib. All concentrations were also used in single-treatment to elucidate the effects produced by the combination.

T790M+ GR models treated with G28 alone and in combination with gefitinib showed both FASN protein and mRNA expression decrease. In the PC9-GR3 model, the FASN protein was slightly diminished in mono- and co-treatments. This decrease is in accordance to FASN mRNA expression, which was significantly reduced in combination with osimertinib. Although G28 alone showed more activated EGFR compared to control in all GR models, the combination of G28 with both EGFR TKIs decreased p-EGFR levels. Regarding the total EGFR, co-treatment resulted in higher protein levels compared to monotreatment in all GR models. However, only T790M+ GR models treated with G28 in combination with gefitinib presented significantly higher EGFR mRNA expression (PC9-GR1 *p* = 0.036 and PC9-GR4 *p* = 0.040). Moreover, G28 both alone and in combination with gefitinib reduced STAT3 activation in T790M+ GR models. No changes in STAT3 activation were seen in PC9-GR3 in any treatment. None of the models analyzed showed alterations in the STAT3 protein or mRNA expression levels.

**Figure 7.** Combination index (CI) of FASN inhibitors (EGCG and G28) and EGFR TKIs (gefitinib and osimertinib) treatment in Gefitinib Resistant (GR) models (PC9-GR1, PC9-GR3, and PC9-GR4). (**a**) CI of EGCG and gefitinib in PC9-GR1, PC9-GR3, and PC9-GR4 models or osimertinib in PC9-GR3. (**b**) CI of G28 and gefitinib in PC9-GR1, PC9-GR3, and PC9-GR4 models or osimertinib in PC9-GR3. PC9-GR1, PC9-GR3, and PC9-GR4 models were treated with EGCG (10–150 μM) or G28 (2-30 μM) in combination with gefitinib (1, 2.5, and 5 μM) for 72 h. PC9-GR3 cells were also treated with EGCG (10–150 μM) or G28 (2–30 μM) in combination with osimertinib (0.5, 1, and 2 μM) for 72 h. Results were determined using the MTT assay and are expressed as the CI based on the Chou and Talalay method [48]. The dotted line indicates additivism (CI approximately equal to 1). CI > 1 designates antagonistic effects and CI < 1 synergistic effects. Experiments were performed at least three times. Results shown are mean ± SE.

**Figure 8.** FASN, EGFR, and STAT3 protein and mRNA expression in sensitive (PC9) and Gefitinib Resistant (GR) models (PC9-GR1, PC9-GR3, and PC9-GR4) treated with FASN inhibitors (EGCG and G28) in combination with EGFR TKIs (gefitinib and osimertinib). (**a**) Western blot analysis of FASN, EGFR, and STAT3 in PC9-GR1, PC9-GR3, and PC9-GR4 models treated with G28 alone and in combination with gefitinib, and PC9-GR3 treated with G28 and osimertinib for 72 h. Results shown are representative from three independent experiments. Untreated cells are used as internal control (Ctrl) and GAPDH as a loading control. (**b**) FASN, EGFR, and STAT3 mRNA levels analysis in PC9-GR1, PC9-GR3, and PC9-GR4 models treated with G28 in combination with gefitinib, PC9-GR3 model treated with the combination G28 and osimertinib for 72 h. mRNA levels were obtained by qRT-PCR and normalized against the GAPDH gene. All conditions were compared to the control (untreated cells), which was normalized to 1 (indicated by the dotted line) and expressed as a fold change. Experiments were performed at least three times. \* *p* < 0.050, \*\* *p* < 0.010, \*\*\* *p* < 0.001 indicate levels of statistical significance.

#### **3. Discussion**

Despite significant advances in EGFRm NSCLC treatment, current therapy is still ineffective to many patients due to the late stage of diagnosis and acquisition of resistance to EGFR TKIs [2,6]. Hence, several efforts have been made on the identification of EGFR TKIs resistance mechanisms to develop an effective treatment for these patients. Some authors pointed out the important role of FASN in drug resistance due to its capacity to allow fast synthesis of new phospholipids for membrane remodeling and plasticity [49,50]. Although the relationship between EGFR and FASN remains unclear, it has recently been described that EGFR upregulates FASN in TKI-resistant EGFRm NSCLC [41]. In addition, FASN inhibition showed cytotoxic effects in lung cancer [40] and resensitized cells to chemotherapy, anti-EGFR and anti-HER2 therapies in breast cancer [38,51]. Therefore, the FASN enzyme may become a promising target for anticancer therapy in EGFRm NSCLC.

Here we studied the effects of natural polyphenolic compound EGCG and its derivative G28 to overcome EGFR TKI resistance in sensitive and GR models. The higher FASN protein levels observed in EGFR TKI resistant models (Figure 1) demonstrated its potential involvement in EGFR TKI resistance acquisition. FASN inhibitors showed similar cytotoxic effects between sensitive and resistant models with IC50 values ranging from 75 to 90 μM for EGCG and from 12 to 18 μM for G28 (Figure 3). As determined by K. Jacobsen and coworkers, PC9-GR1 is a T790M+ GR model that also presents MET and EphA2 activation, the PC9-GR3 model exhibits AXL overexpression, and the T790M+ PC9-GR4 model shows EphA2 activation and AXL overexpression [45]. In correlation with this, the higher G28 IC50 found in PC9-GR3 model and the significantly similar IC50 values in the two T790M+ GR models indicate that none of the known resistance mechanisms described are related to FASN inhibition.

Both compounds have the ability to internalize and inhibit FASN activity as observed in parental PC9 cells (Figure 4), however G28 was of average 5.5 times more effective than EGCG. We have previously shown the ability of the natural compound EGCG to inhibit FASN activity in wild type EGFR NSCLC cells and different breast cancer subtypes [40,43]. Despite their cytotoxicity, only the synthetic compound significantly reduced FASN activity in GR models (Figure 4). Other studies proved the ability of G28 to inhibit FASN activity in triple-negative breast cancer (TNBC) cells [43]. Nevertheless, our study demonstrated that EGCG anticancer activity was independent of FASN inhibition in GR models. It has been extensively described that EGCG has multiple targets and it is involved in multiple signaling pathways and transcription factors, membrane-associated receptors tyrosine kinase (RTKs), or DNA methylation [33]. Some authors exposed that the mechanism underlying EGCG antitumor potency is due to the suppression of the EGFR signaling pathway in NSCLC [52]. Others observed a very stable complex between EGCG and EGFRm (exon 19 ELREA deletion) that was lost with the T790M substitution [53]. Kozue and coworkers showed that EGCG reduced stemness and immunogenicity in EGFR wild type NSCLC cells in vitro and in vivo through the inhibition of AXL [54]. AXL is a tyrosine kinase receptor that has been related to drug resistance and the induction of malignant properties [55] and is overexpressed in GR models [45].

The apoptosis induction was verified by PARP cleavage for all treatments in sensitive and T790M− GR models (Figure 5). However, no PARP cleavage was observed through Western blot analysis in T790M+ GR models after G28 treatment. These results suggest that G28 might cause an apoptosis-independent proliferation reduction in T790M+ GR models as found by others in NSCLC wild-type EGFR models treated with the natural plant polyphenol resveratrol (3,5,40-trihydroxystilben) [56]. The anticancer activity of polyphenols has been shown to be mediated by numerous mechanisms including cell cycle arrest. EGCG, among other natural compounds, showed down-regulation of cyclin-dependent kinases (CDK) and modulation in CDK inhibitors in different human carcinomas [33,57]. The lack of PARP cleavage in some of the treatments and models could be due to the activation of other programmed cell death mechanisms such as autophagy [58].

Alteration in EGFR expression was observed after treatment with FASN inhibitors. EGCG produced a decrease in EGFR protein levels in T790M+ GR models, which is in agreement with Ali et al., who observed an EGFR decrease in sensitive and resistant PC9 cells treated with Orlistat, a FDA-approved FASN inhibitor for obesity management [41]. The synthetic G28 compound increased EGFR mRNA levels in all models, indicating that EGFR overexpression could be an alternative pathway to FASN inhibition. The EGFR activation after FASN gene overexpression in epithelial breast cells has been previously shown [42]. Furthermore, our results suggest that the EGFR pathway could be implicated in FASN regulation at a transcriptional or translational level as exposed before [40]. Despite all models increased EGFR mRNA expression after G28 treatment, only PC9 cells seemed to compensate the effect of G28 by increasing EGFR activation. Therefore, cell proliferation reduction found in all models after FASN inhibition (Figure 3) could be due to the lack of post-translational palmitoylation substrate [41].

It seems increasingly clear that persistent STAT3 activation is related to EGFR-based therapies resistance [13]. STAT3 is under the control of different cytokines and growth factors playing an important role in metastasis, proliferation, survival, invasion, migration, and angiogenesis [14–16]. Natural polyphenols, normally multitarget inhibitors, are now emerging as promising STAT3 inhibitors or its upstream signaling molecules Src, gp130, or NFкB [18]. Among them, EGCG has been reported to reduce STAT3 phosphorylation in head and neck carcinomas [59] and pancreatic cancer [60]. Based on the above, we hypothesized that EGCG and its derivative G28 would diminish the STAT3 activation produced by EGFR TKIs in NSCLC cells. Both compounds reduced p-STAT3 levels in a dose-dependent manner when used alone in comparison to control samples. In agreement with previous studies, STAT3 was activated by EGFR TKIs (Figure 6) [22].

Combinatorial treatments of EGCG and G28 with EGFR TKIs were performed in order to study their effects on GR models (Figure S1). The combination of gefitinib and osimertinib with EGCG showed additivism whereas synergistic effects were identified in combination with G28 (Figure 7). Other authors have previously reported synergistic outcomes after co-treatment with a STAT3 inhibitor (TPCA-1) and EGFR TKI in EGFRm NSCLC models [61]. Previous results from our group demonstrated that G28 compound had a synergistic interaction with pertuzumab and temsirolimus in HER2+ breast cancer cells [38] and EGCG with cetuximab in TNBC [51]. Interestingly, G28 in combination with gefitinib decreased the activation of STAT3 to the same extent as when used alone in T790M+ GR models. Thus, other mechanisms must be involved in the synergistic effects found. On the other hand, no differences in p-STAT3 were observed in PC9-GR3. This could be, in part, due to the multiple pathways that can be altered in the acquisition of gefitinib and osimertinib double-resistance in PC9-GR3 model such as mutations in EGFR/STAT3 or other related up- or downstream signaling molecules, leading to the constitutive activation of STAT3. The analysis of the main genes regulated by the STAT3 transcription factor could provide relevant information to identify some pathways alterations after FASN inhibitors treatment. The synergistic effects found in GR models co-treated with FASN inhibitor G28 and EGFR TKIs supports the idea that EGFR palmitoylation mediated by FASN leads to TKI resistance acquisition in EGFRm NSCLC [41]. However, the specific G28 mechanism of action and possible targets are still unknown and further studies are needed. Wu et al. suggested that FASN inhibition may cause an imbalance in the membrane lipids levels, which may produce a membrane localization decrease of IGF-1R, and the inactivation of the downstream STAT3 signaling pathway [36]. Furthermore, the IGF-1R is a transmembrane tyrosine kinase linked to MAPK and PI3K/AKT pathways, shared with EGFR, which could explain the effects found only in T790M+ GR models [62].

Taken together, our observations suggest that FASN has a key role in acquired TKI-resistant EGFRm NSCLC. The inhibition of this enzyme resensitizes cells to EGFR TKIs treatments. These results encourage for further studies to analyze the combinatorial treatment of FASN inhibitors and EGFR TKIs to overcome the EGFR TKI resistance in NSCLC.

#### **4. Materials and Methods**

#### *4.1. Cell Lines and Culture Conditions*

Human adenocarcinoma PC9 cells and its gefitinib resistant derivatives PC9-GR1, PC9-GR3, and PC9-GR4 models [45] were kindly provided by Dr. R. Rosell and Dr. M. A. Molina (Barcelona, Spain). All cells were routinely grown in RPMI-1640 medium (Lonza, Basel, Switzerland), supplemented with 10% fetal bovine serum (FBS; HyClone Laboratories, GE Healthcare, Chicago, IL, USA), 50 U/mL penicillin, and 50 μg/mL streptomycin (Lonza). In all cases, the cells were used immediately after resuscitation and were maintained at 37 ◦C in a humidified atmosphere with 5% CO2, propagated following established protocols, remaining free of mycoplasma throughout the experiments.

#### *4.2. Cell Proliferation Assays*

To investigate the effect of EGFR TKIs or FASN inhibitors cells were seeded into 96-well plates at the appropriate density in their growth medium. Gefitinib and osimertinib were kindly provided by AstraZeneca. EGCG was purchased from Sigma (USA) and G28 was synthesized as described elsewhere [43]. For monotreatment assays, after 24 h cells were treated with different concentrations of each drug for 72 h. Cell viability was determined by the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-tetrazolium bromide (MTT) assay (Sigma) as described elsewhere [44]. For drug combination experiments, cells were treated with three fixed concentrations of gefitinib (1, 2.5, and 5 μM) or osimertinib (0.5, 1, and 2 μM) in combination of a series of increasing concentrations of EGCG or G28 for 72 h. Following treatment, cell proliferation was measured using the standard colorimetric MTT assay. Using the multi-well plate reader Benchmark Plus (Bio-Rad Laboratories, Inc., CA, USA), absorbance was determined at 570 nm. Combinatorial effects were evaluated using the combination index (CI) based on the Chou and Talalay method [48] using the CompuSynTM software (Biosoft, MO, USA). CompuSynTM calculates the CI values; if the value equals 1 the effect is considered additive, if above 1 antagonistic and below 1 synergistic. Data presented are from three separate wells per assay, and the assay was performed at least three times.

#### *4.3. Western Blot Analysis of Cell Lysates*

Gefitinib-sensitive and -resistant models were treated with EGFR TKI, FASN inhibitor, or the combination of both drugs for 72 h. Afterwards, attached and floating cells were harvested and lysed in ice-cold lysis buffer (Cell Signaling Technology Inc., MA, USA) containing 100 μg/mL phenylmethylsulfonylfluoride (PMSF; Sigma) by vortexing every 5 min for 30 min. Protein concentration was determined by the Lowry method (DC Protein Assay, Bio-Rad Laboratories, Inc.). Equal amounts of protein were heated in lithium dodecyl sulfate (LDS) sample buffer with a sample reducing agent (Invitrogen, CA, USA) for 10 min at 70 ◦C, separated by SDS-PAGE and transferred to nitrocellulose membranes (ThermoFisher Scientific Inc., MA, USA). Membranes were incubated for 1 h at room temperature in blocking buffer (5% skim milk powder in Tris-buffered saline (TBS)) 0.05% Tween (TBS-T) and overnight at 4 ◦C with the following primary antibodies diluted in blocking buffer: FASN (Cell Signaling Technology Inc.; #3180), p-STAT3Tyr705 (Cell Signaling Technology Inc.; #9131), p-EGFRTyr1068 (Cell Signaling Technology Inc.; #2234), PARP (Cell Signaling Technology Inc.; #9542), EGFR (Cell Signaling Technology Inc.; #2232), STAT3 (Cell Signaling Technology Inc.; #4904), and GAPDH (Proteintech, Manchester, UK; #60004-1-IG). Specific horseradish peroxidase (HRP)-conjugated secondary antibodies were incubated for 1 h at room temperature. The immune complexes were detected using chemiluminescent HRP substrate ClarityTM Western ECL Substrate (Bio-Rad Laboratories, Inc.) or West Femto Maximum Sensitivity Substrate (ThermoFisher Scientific Inc.) in a Bio-Rad ChemiDocTM MP Imaging System (Bio-Rad Laboratories, Inc.). Western blot analyses were repeated at least three times and representative results are shown.

#### *4.4. Quantitative Real-Time PCR (qRT-PCR) Analysis*

Cells were treated with EGFR TKIs or FASN inhibitors as a single agent or in combination for 72 h. Then, cells were PBS washed and resuspended in 1 mL of Qiazol (Qiagen, Hilden, Germany). GeneJET RNA Purification Kit (ThermoFisher Scientific Inc.) was used to isolate total RNA following the manufacturer's instructions. RNA samples were quantified using a Nano-Drop 2000 Spectophotometer (ThermoFisher Scientific). Total RNA was reverse-transcribed into complementary DNA (cDNA) using a High Capacity cDNA Archive Kit (Applied Biosystems, CA, USA). Different gene expression levels were determined using QuantStudio3 Real-time PCR System (ThermoFisher Scientific Inc.) with qPCRBIO SyGreen Mix Lo-Rox real-time PCR (PCR Biosystems Inc., PA, USA), following manufacture instructions. Primers used are shown in Table 1. qRT-PCR analyses were performed at least four times and each gene was run in triplicate. Gene expression levels were quantified using the standard formula 2ˆdCT and normalized to the housekeeping GAPDH (2ˆdCT).


**Table 1.** Primer design.

#### *4.5. Inhibition of Fatty Acid Synthase Activity*

Cells were seeded in 24-well plates at a density of 3 <sup>×</sup> 104 cells/well in RPMI supplemented with 10% fetal bovine serum (FBS). 24 h later, the maintenance medium was replaced by the treatment medium (RPMI-1640 supplemented with 1% lipoprotein-deficient FBS (Sigma-Aldrich)) along with the assayed compounds or vehicle (dimethyl sulfoxide, DMSO, Sigma-Aldrich). Cells were treated with a concentration corresponding to the previously determined IC50 for 72 h. For the last 6 h (1,2-14C) acetic acid sodium salt (53.9 mCi/mmol, Perkin Elmer Biosciences, Waltham, MA, USA) was added to the medium (0.5 μCi/mL). The lipid extraction was performed as previously described [43]. Cells were washed twice with phosphate-buffered saline (PBS, HyClone Laboratories, Logan, UT, USA) and once with MeOH/PBS (2:3). Cell pellets were resuspended in 0.2 M NaCl and lysed with freeze (liquid N2)—thaw (37 ◦C) cycles. Then, lipids were extracted with CHCl3/MeOH (2:1) and 0.1 M KOH. The organic phase was washed with CHCl3/MeOH/H2O (3:48:47) and the solvents were evaporated under vacuum conditions. Finally, pellets were resuspended in EtOH and counted by scintillation. The total protein content was quantified by the Bradford assay (Sigma-Aldrich).

#### *4.6. Statistical Analysis*

Parametric data were analyzed by the Student's *t* test when comparing two groups or the one-way analysis of variance (ANOVA) followed by Bonferroni or Tamhane's T2 post hoc test for multiple comparisons. The non-parametric data were analyzed with the Mann–Whitney U tests for non-normally independent variables; otherwise, the Kruskal–Wallis test was used for more than two groups. All data are expressed as mean ± SE. Levels of significance were set at *p* < 0.050 and are represented by asterisks, as follows: *p* < 0.050 (denoted as \*), *p* < 0.010 (denoted as \*\*), and *p* < 0.001 (denoted as \*\*\*). The statistical analysis was performed using the IBM SPSS software (Version 21.0; SPSS Inc., IL, USA).

#### **5. Conclusions**

The need to find more effective and less toxic therapeutic treatments for EGFRm NSCLC patients is one of the major challenges in lung cancer research. Natural compounds are emerging as potential anticancer candidates for their safety and multitarget intrinsic features. Improvement of the properties of natural compounds through the design of synthetic derivatives is aimed at maintaining these features while increasing, efficacy, bioavailability, and stability in physiological conditions.

Here we show additive and synergistic effects of the polyphenolic plant-derived EGCG compound and its derivative G28, respectively, in combination with EGFR TKIs in GR models. Despite the exact mechanism by which these compounds are cytotoxic is still unknown, our results shed light on their ability to modulate FASN/EGFR/STAT3 pathways. The capacity to affect multiple pathways might prove useful in overcoming other drug resistances. Further analyses are required to completely understand the mechanism of action.

Taken together, this paper supports the inhibition of the metabolic enzyme FASN by G28 compound in combination with EGFR TKIs as a new potential strategy for resistant EGFRm NSCLC.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6694/12/5/1283/s1, Figure S1: Proliferative curves of FASN inhibitors (EGCG and G28) alone and in combination with EGFR TKIs (gefitinib and osimertinib) in EGFR TKIs resistant models, Figure S2. Whole Western blot figures showing PARP, FASN, EGFR, STAT3 and GAPDH protein bands with molecular weight markers (merge of colorimetric and chemiluminescence). (a) Western blot images from Figure 5. (b) Western blot images from Figure 6. (c) Western blot images from Figure 8.

**Author Contributions:** Conceptualization, T.P. and S.R.-M.; methodology, E.P.-A., S.P., S.R.-M., J.R., D.T.-O., M.P., L.F.; validation, E.P.-A., S.P., S.R.-M.; resources, T.P., and J.C.; writing—original draft preparation, E.P.-A., S.P., S.R.-M.; writing—review and editing, E.P.-A., S.P., S.R.-M., T.P. and J.C.; supervision, T.P. and S.R.-M.; funding acquisition, T.P. and J.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors acknowledge the financial support from AstraZeneca, the E. P.-A pre-doctoral grant (2019FI\_B01011), and the support of the Catalan Government (2017SGR00385).

**Acknowledgments:** Authors thank R. Rosell and M. A. Molina from laboratory of Oncology Pangaea (Barcelona, Spain) for kindly provided PC9 models and AstraZeneca (London, UK) for supplying gefitinib and osimertinib.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

### **Targeting Glycolysis with Epigallocatechin-3-Gallate Enhances the E**ffi**cacy of Chemotherapeutics in Pancreatic Cancer Cells and Xenografts**

#### **Ran Wei 1,2, Robert M. Hackman 2, Yuefei Wang <sup>1</sup> and Gerardo G. Mackenzie 2,3,\***


Received: 22 August 2019; Accepted: 3 October 2019; Published: 5 October 2019

**Abstract:** Pancreatic cancer is a complex disease, in need of new therapeutic approaches. In this study, we explored the effect and mechanism of action of epigallocatechin-3-gallate (EGCG), a major polyphenol in green tea, alone and in combination with current chemotherapeutics on pancreatic cancer cell growth, focusing on glycolysis metabolism. Moreover, we investigated whether EGCG's effect is dependent on its ability to induce reactive oxygen species (ROS). EGCG reduced pancreatic cancer cell growth in a concentration-dependent manner and the growth inhibition effect was further enhanced under glucose deprivation conditions. Mechanistically, EGCG induced ROS levels concentration-dependently. EGCG affected glycolysis by suppressing the extracellular acidification rate through the reduction of the activity and levels of the glycolytic enzymes phosphofructokinase and pyruvate kinase. Cotreatment with catalase abrogated EGCG's effect on phosphofructokinase and pyruvate kinase. Furthermore, EGCG sensitized gemcitabine to inhibit pancreatic cancer cell growth in vitro and in vivo. EGCG and gemcitabine, given alone, reduced pancreatic tumor xenograft growth by 40% and 52%, respectively, whereas the EGCG/gemcitabine combination reduced tumor growth by 67%. EGCG enhanced gemcitabine's effect on apoptosis, cell proliferation, cell cycle and further suppressed phosphofructokinase and pyruvate kinase levels. In conclusion, EGCG is a strong combination partner of gemcitabine reducing pancreatic cancer cell growth by suppressing glycolysis.

**Keywords:** pancreatic cancer; epigallocatechin-3-gallate (EGCG); gemcitabine; glycolysis; ROS; phosphofructokinase

#### **1. Introduction**

Pancreatic cancer is one of the top five causes of cancer-related death in the United States, with an overall five-year survival rate of approximately eight percent [1]. While surgery, which offers the only realistic hope, is a viable option in a limited number of patients, current radiation therapy and chemotherapy regimens offer no significant clinical benefit [2]. For example, gemcitabine is an antimetabolite recognized as the current standard of care for unresectable locally advanced or metastatic pancreatic cancer, but its therapeutic effect for patients is limited [3]. Thus, an urgent need exists for new ways to combat pancreatic cancer, with the exploration of novel therapeutic strategies being a critical component.

Deregulated energy metabolism is recognized as one of the hallmarks of cancer [4]. Unlike normal cells, tumor cells prefer to rely on aerobic glycolysis to produce energy and to obtain intermediate metabolites that can then be directed to other biosynthetic pathways that promote tumor growth [5,6]. Moreover, the higher glycolytic rate in cancer cells has been shown to correlate to chemotherapeutic

resistance [7]. Therefore, targeting tumor glycolysis remains attractive for therapeutic interventions, not only as a main target, but also for its ability to sensitize cancer cells to chemotherapeutics.

Epigallocatechin-3-gallate (EGCG), the most active polyphenol component found in green tea, presents several anticancer properties [8–10]. Accumulating evidence, including a previous study by our group, has shown that EGCG can inhibit cancer cell growth by modulating metabolic pathways [11–13]. An important consideration when exploring EGCG's mechanism of action is that in the presence of oxygen, EGCG undergoes auto-oxidation to generate reactive oxygen species (ROS) [14]. Since cancer cells are more susceptible to an increase in ROS levels [15], this has been proposed as a potential mechanism of action for the inhibitory cell growth effect by EGCG. However, to date, whether the glycolytic pathway is modulated by EGCG in pancreatic cancer, and whether it is ROS-dependent is not completely elucidated.

In the present study, we explored the effect and mechanism of action of EGCG alone and in combination with current chemotherapeutics on pancreatic cancer cell growth, focusing on glycolysis metabolism involved in pancreatic cancer cell growth, and investigated whether it is dependent on the ability of EGCG to induce ROS. Our results show that EGCG strongly reduced pancreatic cancer cell growth by suppressing glycolysis in a ROS-dependent manner. Moreover, EGCG enhanced the anticancer effect of gemcitabine in vitro and in vivo by inhibiting glycolysis and affecting cell kinetics.

#### **2. Results**

#### *2.1. EGCG A*ff*ects Glycolysis in Pancreatic Cancer Cells*

We initially evaluated the efficacy of EGCG on pancreatic cancer cells in culture. For this purpose, we treated multiple human pancreatic cancer cells with or without increasing concentrations of EGCG (20–100 μM) for 72 h. In parallel, we also compared the effect of EGCG in human pancreatic cancer cells to that of human pancreatic normal epithelial cells (HPNE). EGCG reduced human pancreatic cancer cell growth in a concentration-dependent manner. Of note, EGCG reduced cell growth more potently in human pancreatic cancer cells compared to HPNE cells. For instance, at 72 h, EGCG 80 μM reduced MIA PaCa-2 and Panc-1 cell growth by 84% and 64% (*p* < 0.05), respectively. In contrast, EGCG at 80 μM for 72 h had significantly less effect on the HPNE cells, reducing cell growth by only 24% (Figure 1A).

Given that alterations in cancer cell metabolism can lead to an inhibition of cell growth [16], and because EGCG has been shown to affect glycolysis in other tumor types [11–13], we evaluated the effect of EGCG on glucose metabolism in two human pancreatic cancer cells and one murine pancreatic cancer cell line (KPC). We assessed the impact of EGCG on glycolysis by measuring the extracellular acidification rate (ECAR) in the MIA PaCa-2, Panc-1, and KPC cell lines. After 24 h of treatment, EGCG strongly reduced ECAR in all three cell lines, revealing an inhibitory effect of EGCG on glycolysis (Figure 1B and Figure S1).

Next, we determined the effect of EGCG on pancreatic cancer cell growth under the conditions of either glucose deprivation or treatment with 2-Deoxy-D-glucose (2-DG), a glucose analog. As shown in Figure 1C, the reduction in cell growth induced by EGCG was enhanced under glucose deprivation or 2-DG treatment. For example, 2-DG alone reduced the cell growth rate by 5%, 27%, and 8% in Panc-1, MIA PaCa-2, and KPC cells, respectively, whereas when combined with EGCG, the cell growth was reduced by 71%, 58%, and 89%, respectively (*p* < 0.01; Figure 1C). Moreover, EGCG treatment reduced ATP levels concentration-dependently (Figure 1D). Treatment with EGCG at 40 μM for 24 h reduced ATP levels in Panc-1 and MIA PaCa-2 cells by 35% and 32%, respectively (*p* < 0.01 for both, Figure 1D).

**Figure 1.** Epigallocatechin-3-gallate (EGCG) inhibits pancreatic cancer cell growth through glycolysis suppression. (**A**) EGCG inhibits human pancreatic cancer cell growth in a concentration-dependent manner. Cell growth was determined in Panc-1, MIA PaCa-2, HPAF-II, BxPC-3, SU-86.86, CFPAC-1, and KPC pancreatic cancer cells, and in the human pancreatic normal epithelial (HPNE) cells after treatment with increasing EGCG concentrations for 72 h. Results are expressed as a percentage of control. (**B**) EGCG suppresses glycolysis capacity in Panc-1, MIA PaCa-2, and KPC cells after 24 h. Glucose (25 mM), Oligomycin (1 μM) and 2-Deoxy-D-glucose (2-DG) (75 mM) were injected and the extracellular acidification rate (ECAR) of live cells was monitored during the experimental period. Results are presented as the mean ± SD of ECAR. (**C**) Cell growth was measured in Panc-1, MIA PaCa-2, and KPC cells treated with or without EGCG (40 μM) under glucose deprivation or 2-DG (10 mM) treatment condition. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**D**) EGCG reduced ATP levels in Panc-1 and MIA PaCa-2 cells after 24 h. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

#### *2.2. EGCG Inhibits Glycolysis through Suppressing Rate-Limiting Enzymes*

Given the effect of EGCG on glycolysis, we evaluated whether EGCG could affect any particular step in the glycolytic pathway by measuring the activity and levels of glycolytic enzymes. EGCG treatment reduced both the activity and expression levels of phosphofructokinase (PFK) and pyruvate kinase (PK) in Panc-1 and MIA PaCa-2 cells, having a stronger effect on PFK (Figure 2A–C, Figure S5). For instance, EGCG at 40 μM reduced the levels of platelet-type phosphofructokinase (PFKP) and the pyruvate kinase M2 (PKM2), an isoform of PK, by 65% and 49%, respectively, in Panc-1 cells, and by 57% and 34%, respectively in MIA PaCa-2 cells (Figure 2B, Figure S4). However, EGCG failed to reduce hexokinases II (HK2) and lactate dehydrogenase A (LDHA) protein expression levels (Figure S2, Figure S12). In agreement with the in vitro results, EGCG reduced the levels of PFKP and PKM2 (*p* < 0.01 for both) in pancreatic tumor xenograft homogenates, obtained from mice treated with EGCG (Figure 2D, Figure S6).

**Figure 2.** *Cont*.

**Figure 2.** EGCG inhibits glycolysis through suppressing rate-limiting enzyme activity and expression. (**A**) Phosphofructokinase (PFK) and pyruvate kinase (PK) activities were determined in Panc-1 and MIA PaCa-2 cells after treatment with EGCG for 24 h. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**B**) Immunoblots for platelet-type phosphofructokinase (PFKP) and pyruvate kinase M2 (PKM2) in total cell protein extracts from Panc-1 and MIA PaCa-2 cells treated with escalating concentrations of EGCG, as indicated, for 24 h. Loading control: β-Actin. Bands were quantified and results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**C**) EGCG (40 μM) inhibited PFKP and PKM2 protein expression in a time-dependent manner in Panc-1 and MIA PaCa-2 cells. Results are expressed as percentage of control and presented as the mean ± SD. \* *p* < 0.05, \*\* *p* <0.01 vs. control. (**D**) Immunoblots of PFKP, PKM2 expression on tumor tissue from control- and EGCG-treated (10mg/kg/d) mice. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**E**) Effect of silencing PFKP on EGCG-induced cell growth reduction. Panc-1 and MIA PaCa-2 cells were transfected with either control or PFKP siRNA. After transfection, cells were treated with EGCG for 72 h and cell growth was evaluated. Results are expressed as percentage of control; \* *p* < 0.05, \*\* *p* < 0.01 vs. control. Immunoblots to verify PFKP silencing were performed on whole cell extracts obtained from these cells (top panel).

To confirm the role of PFKP on EGCG-induced cell growth inhibition, we silenced PFKP in Panc-1 and MIA PaCa-2 cells. Knocking-down PFKP reduced Panc-1 and MIA PaCa-2 cell growth by 15% and 19%, respectively. In Panc-1 and MIA PaCa-2 cells transfected with nonspecific siRNA, EGCG 40 μM decreased the number of viable cells by 49% and 42%, whereas in PFKP-silenced cells, EGCG reduced cell growth by 57% and 54% in Panc-1 and MIA PaCa-2 cells, respectively (Figure 2E, Figure S6). Taken together, these findings indicate that silencing PFKP function has a slight additive effect on the growth inhibitory response of EGCG in both cell lines, and suggest that regulating glycolysis represents an important mechanism of EGCG in inhibiting pancreatic cancer cell growth.

#### *2.3. EGCG A*ff*ects the Glycolytic Pathway in Pancreatic Cancer Cells through a ROS-Dependent Manner*

Because EGCG and other polyphenols have been shown to undergo rapid oxidization to generate free radicals in the presence of oxygen [17], we evaluated the ability of EGCG to induce reactive oxygen species (ROS) in pancreatic cancer cells. Compared to the control, treatment with EGCG significantly increased ROS levels concentration-dependently, as determined with the general probe 2- ,7- -Dichlorodihydrofluorescein diacetate (H2DCFDA). For example, EGCG at 40 μM increased ROS levels by 1.4- and 1.6-fold in Panc-1 and MIA PaCa-2 cells, respectively (*p* < 0.01 for both; Figure 3A).

Next, we determined the levels of hydrogen peroxide (H2O2) using the Amplex™ Red indicator probe. EGCG strongly increased H2O2 levels in pancreatic cancer cells. Compared to controls, EGCG at 40 μM increased H2O2 levels by 1.5- and 1.9-fold in Panc-1 and MIA PaCa-2 cells, respectively (Figure 3B). In both cell lines, cotreatment with catalase significantly abrogated the increase in H2O2 levels induced by EGCG (Figure 3B). Moreover, EGCG increased the levels of mitochondria superoxide anion in a concentration and time-dependent manner in both Panc-1 and MIA PaCa-2 cells (Figure 3C).

**Figure 3.** EGCG affects the glycolytic pathway in pancreatic cancer through a reactive oxygen species (ROS)-dependent manner. (**A**) 2- ,7- -Dichlorodihydrofluorescein diacetate (H2DCFDA) fluorescence was measured in Panc-1 and MIA PaCa-2 cells treated without (control) or with EGCG at various concentration for 24 h. Results are expressed as fold change of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**B**) AmplexTM Red was used to evaluate the levels of hydrogen peroxide level released by Panc-1 and MIA PaCa-2 cells. Results are expressed as fold change of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

(**C**) EGCG induces a concentration- and time-dependent increase in mitochondrial superoxide anion levels. The levels of superoxide anion in the mitochondria were determined by flow cytometry using the MitoSOX-Red fluorescent probe. Black, red and blue lines represent control, EGCG 40 μM, and EGCG 60 μM groups, respectively. (**D**) The reduction in the expression levels of PFKP and PKM2 triggered by EGCG (E) was reversed by catalase (CAT). Immunoblots for PFKP and PKM2 in total cell protein extracts from Panc-1 and MIA PaCa-2 cells treated with 40 μM EGCG (E), 1500 U/mL catalase (CAT) or both, for 12 h. Loading control: β-Actin. Bands were quantified and results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**E**) Catalase (CAT) partly ameliorated the cell growth inhibitory effect induced by EGCG (E). Cells were treated with 40 μM EGCG (E), 1500 U/mL CAT or both, for 24 h or 48 h. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**F**) N-Acetyl-L-Cysteine (NAC) partly reversed cell inhibition effect of EGCG after 48 h. Cells were treated with 40 μM EGCG (E), 5 mM NAC or both for 48 h. Results are expressed as percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**G**) L-Buthionine Sulfoximine (BSO) further aggravated cell inhibition effect of EGCG after 24 h. Cells were treated with 40 μM EGCG (E), 1 mM BSO or both for 24 h. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

To elucidate whether the effect of EGCG on glycolysis is mediated through the increase in ROS, Panc-1 and MIA PaCa-2 cells were incubated with or without EGCG at 40 μM in the presence or absence of catalase for 12 h. In both cell lines, catalase mostly abrogated the inhibitory effect of EGCG on PFKP and PKM2, suggesting that the effect of EGCG on glycolytic enzymes is dependent on the ability of EGCG to induce ROS (Figure 3D, Figure S7).

However, catalase could only, in part, prevent the reduction in pancreatic cancer cell growth induced by EGCG (Figure 3E). Somewhat similar results were obtained when cotreating with N-Acetyl-L-Cysteine (NAC) (Figure 3F). In contrast, treatment with L-Buthionine Sulfoximine (BSO), an inhibitor of glutamylcysteine synthetase, strongly enhanced the reduction in cell growth induced by EGCG in both cell lines (Figure 3G). For example, in Panc-1 cells at 24 h, BSO and EGCG alone reduced cell growth by 22% and 12%, respectively, and by 54% when combined. These results suggest that EGCG reduces pancreatic cancer growth, partly, through an ROS-dependent mechanism.

#### *2.4. EGCG Sensitizes Pancreatic Cancer Cells to Gemcitabine In Vitro*

Because chemotherapeutics often display limited effects, show resistance, and cause side-effects, we evaluated whether EGCG would represent a useful partner in combination with commonly used drugs. For this purpose, we treated Panc-1 and MIA PaCa-2 cells with EGCG alone or in combination with gemcitabine, Abraxane®, 5-Fluorouracil, irinotecan, or oxaliplatin, five commonly used chemotherapeutics in pancreatic cancer patients, for 72 h and analyzed their combination index (CI) by means of the Chou–Talalay method. In both Panc-1 and MIA PaCa-2 cells, EGCG strongly synergized with gemcitabine (Figure 4 and Table S1). For example, in Panc-1 cells, the CI of all tested groups, except one, showed synergistic effects. In MIA PaCa-2 cells, the CI effects between EGCG and gemcitabine were also indicative of a strong additive effect (Figure 4). The combination effect of EGCG with the other four chemotherapeutics tested presented effects that were more variable with some showing an additive effect (Figure 4 and Table S1).

**Figure 4.** EGCG enhances the cell growth inhibitory effect of gemcitabine in pancreatic cancer cells. Combination index (CI) plots of EGCG (20, 40, 60 μM) in combination with gemcitabine (1, 10, 20 and 40 nM), abraxane (1, 10, 20 and 40 nM), 5-Fluorouracil (1, 10, 20 and 40 μM), irrinotecan (1, 10, 20 and 40 μM), and oxialiplatin (1, 10, 20 and 40 μM) in Panc-1 (left) and MIA PaCa-2 (right) cells. Drug interactions were quantitatively determined using the Chou–Talalay method, and CI < 1, =1, and >1 indicates synergism, additive and antagonism effect, respectively. Of note, the CI value dots depicted are based on concentrations of EGCG and the chemotherapeutic drugs shown on Table S1.

#### *2.5. EGCG Enhances the Anticancer E*ff*ect of Gemcitabine in Pancreatic Cancer Xenografts through a Strong Cytokinetic E*ff*ect*

Next, we evaluated whether EGCG could enhance the chemotherapeutic effect of gemcitabine in vivo. For this purpose, murine pancreatic cancer KPC cells were injected subcutaneously into immunocompetent mice, which gave rise to exponentially growing tumors. Once the tumors reached ~300 mm3, the mice were treated either with EGCG (10 mg/kg) suspended in phosphate buffered saline (PBS) pH 7.4 given intraperitoneally (I.P.) once daily, gemcitabine (100 mg/kg) given I.P. twice per week, or both drugs. On day 16 of treatment, EGCG and gemcitabine, given as single agents, reduced tumor weight by 40% and 52%, respectively, compared to vehicle-treated controls (*p* < 0.05 and *p* < 0.01). In combination, EGCG plus gemcitabine reduced tumor weight by 67%, compared to controls (*p* < 0.01). Of note, the effect of the EGCG plus gemcitabine combination was significantly different from that of gemcitabine alone (*p* < 0.05; Figure 5A). Importantly, the drug combination was well tolerated by the mice, as indicated by the comparable mean body weight of the experimental groups throughout the treatment period (Figure 5B).

**Figure 5.** EGCG enhances the anticancer effect of gemcitabine in pancreatic cancer xenografts. (**A**) EGCG (E) promoted tumor weight lost with gemcitabine (G). Results are presented as the mean ± SD. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**B**) Mice body weight during treatment days for control, gemcitabine-treated groups (G), EGCG-treated groups (E), and the combination (G + E). Results are presented as the mean ± SD. (**C**) Immunostaining of the cell proliferation marker Ki-67 and of cleaved Caspase 3 were performed on KPC tumor sections and photographs were taken at 100× magnification. Representative images are shown. The consecutive section was stained with isotype Immunoglobulin G (IgG) as negative staining control and it is shown in the upper left corner. Results were expressed as percent of positive (+) cells ± SD per 100x field. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

Because EGCG has been shown to be hepatotoxic at higher doses [18], and to further evaluate the safety of EGCG plus gemcitabine, we performed an acute toxicity study and determined the levels of multiple liver enzymes and kidney markers. After 16 days of EGCG plus gemcitabine treatment, liver and kidney function markers (including activity of Alanine Transaminase, Alkaline Phosphatase, Aspartate Transaminase, and the levels of blood urea nitrogen, creatinine, etc.) were in the normal range (Table S2). Consistent with the efficacy study, the mean body weights were comparable between the groups (Figure S3).

To investigate the mechanism by which EGCG plus gemcitabine reduced tumor growth, we determined cell proliferation (Ki-67 expression), and apoptosis (cleaved Caspase 3 expression) levels by immunohistochemistry in tumor tissue sections from control and EGCG plus/minus gemcitabine-treated mice (Figure 5C). Compared to controls, EGCG alone and EGCG plus gemcitabine inhibited cell proliferation (Ki-67 positive cells) by 45% and 83% in the KPC xenografts, respectively (*p* < 0.01, Figure 5C). In addition, EGCG increased the percentage of cleaved Caspase 3 positive cells by 1.6-fold when given alone and by 2.7-fold when given in combination (*p* < 0.01, Figure 5C).

In vitro, we determined the effect of EGCG and gemcitabine on cell cycle progression and cell death by apoptosis. As expected, gemcitabine, an inhibitor of DNA synthesis [19], strongly induced Synthesis/Gap2 (S/G2) phase arrest after 48 h of treatment. While EGCG at 40 μM only slightly arrested cell cycle progression, it strongly enhanced the effect of gemcitabine, with the combination further blocking the cell cycle at S/G2 in Panc-1 and MIA PaCa-2 cells (Figure 6A). Concomitant with cell cycle arrest, expression levels of S/G2 checkpoint proteins, including phosphorylated checkpoint kinases 1 (p-Chk1), phosphorylated tumor protein p53 (p-p53), and cyclin-dependent kinase (cdk) inhibitor p21 Waf1/Cip1 (p21) further increased, while cdc2 and CyclinB1 were further reduced in the EGCG plus gemcitabine-treated cells compared to those treated with gemcitabine alone (Figure 6B, Figure S8).

**Figure 6.** *Cont*.

**Figure 6.** EGCG and gemcitabine together further affected cell kinetic in pancreatic cancer cells. (**A**) EGCG (E) enhanced the effect of gemcitabine (G) on the cell cycle. Following treatment with 40 μM EGCG (E), 20 nM gemcitabine (G), or both (G + E) for 48 h, cells were stained with propidium iodide (PI) and the number of cells in each phase of the cell cycle was measured by flow cytometry. Results are expressed as a percentage of control. (**B**) EGCG (E) and gemcitabine (G) modulated S/G2 phase protein expression. Immunoblots for phosphorylated checkpoint kinases 1 (p-Chk1), phosphorylated and total tumor protein p53 (p53), cyclin-dependent kinase (cdk) inhibitor p21 Waf1/Cip1 (p21), cell division cycle 2 (cdc2) and Cyclin B1 in total cell protein extracts from Panc-1 and MIA PaCa-2 cells treated with EGCG (E), gemcitabine (G), or both (G + E), for 48 h. Loading control: β-Actin. Bands were quantified and results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**C**) Panc-1 and MIA PaCa-2 cells were treated with EGCG (E), gemcitabine (G), or both (G + E) for 48 h, and the percentage of apoptotic cells were determined by flow cytometry using dual staining (Annexin V and propipium iodide). The percentages of Annexin V (+) cells was calculated, and results are expressed as the fold-increase over control. Co-treatment with EGCG (E) further increased the apoptosis rate induced by gemcitabine (G) alone after 48 h. Results are expressed as percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**D**) Immunoblots for full length and cleaved Caspases 3, 7 and 9 as well as full length and cleaved poly (ADP-ribose) polymerase (PARP) in total cell protein extracts from Panc-1 and MIA PaCa-2 cells treated with EGCG (E), gemcitabine (G), or both (G + E) for 48 h. Loading control: β-Actin. Bands were quantified and results are shown as the ratio between the cleaved/full length protein; \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**E**) EGCG sensitized gemcitabine on apoptosis induction by regulating B-cell lymphoma 2 (Bcl-2) family and Inhibitors of apoptosis proteins (IAP) family protein expression in Panc-1 and MIA PaCa-2 cells after 48 h. Results are expressed as percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

Besides blocking cell cycle progression and consistent with the in vivo results, gemcitabine also induced apoptosis in pancreatic cancer cells in culture. Compared to controls, treatment with gemcitabine for 48 h resulted in a 2.4- and 1.6-fold increase in apoptosis in Panc-1 and MIA PaCa-2 cells, respectively (Figure 6C). While EGCG at 40 μM induced apoptosis by 3.9- and 2.7-fold, the effect of EGCG plus gemcitabine was enhanced 4.7- and 3.4-fold over control in Panc-1 and MIA PaCa-2 cells, a response that was approximately two times that of gemcitabine alone (*p* < 0.01).

We then determined the expression of proteins involved in the mechanism of apoptosis by Western blot. Compared to controls, EGCG plus gemcitabine significantly affected the expression of cleaved Caspase 3, 7, 9, cleaved poly (ADP-ribose) polymerase (PARP), proapoptotic member of the Bcl-2 family protein Bad, antiapoptotic member of the Bcl-2 family protein Bcl-xl, and X-linked inhibitor of apoptosis protein (XIAP) levels, but not survivin (Figure 6D,E and Figure S9). Of note, no additive effect was observed on the expression of cleaved Caspase 3, 7, 9, Bad, Bcl-xl, or XIAP between EGCG plus gemcitabine and EGCG alone (Figure 6E, Figure S10), suggesting that the effect of the combination is most likely being driven by EGCG.

#### *2.6. EGCG Plus Gemcitabine Further Inhibits Glycolysis*

Given that EGCG strongly affected the glycolytic pathway, we evaluated whether combining EGCG with gemcitabine would lead to any additional glycolysis inhibitory effect. For this purpose, we determined the activity and levels of glycolytic enzymes in Panc-1 and MIA PaCa-2 cells, treated with EGCG and gemcitabine alone or in combination. The activity and levels of PFK and PK were significantly reduced compared to controls and gemcitabine alone groups. However, the effect of the combination group was not significant compared to the EGCG group (Figure 7A,B and Figure S11).

**Figure 7.** EGCG plus gemcitabine inhibit glycolysis. (**A**) The activity of PFK and PK was determined in Panc-1 and MIA PaCa-2 cells treated with 40 μM EGCG (E), 20 nM gemcitabine (G), or both (G + E) for 24 h. Results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control. (**B**) Immunoblots for PFKP and PKM2 in total cell protein extracts from Panc-1 and MIA PaCa-2 cells treated with EGCG (E), gemcitabine (G), or both (G + E) for 12 h. Loading control: β-Actin. Bands were quantified and results are expressed as a percentage of control. \* *p* < 0.05, \*\* *p* < 0.01 vs. control.

#### **3. Discussion**

Pancreatic cancer continues to be a significant health problem around the world. Given the lack of effective treatments that can meaningfully prolong a patient's life, an active search for agents that have additive or synergistic effect with chemotherapeutic drugs is critical in order to enhance their efficacy. In this study, we show that EGCG sensitizes pancreatic cancer cells to gemcitabine by suppressing glycolysis.

Our work identified the glycolytic pathway as one of the major signaling mechanisms involved in eliciting the growth inhibitory effect of EGCG. Glycolysis supports cell growth by rapidly generating ATP and metabolic intermediates for other biosynthetic pathways. A high rate of glycolysis is characteristic of cancer cells, even in the presence of sufficient oxygen [20]. EGCG strongly inhibited the glycolytic rate in pancreatic cancer cells by affecting the activity and expression of PFK and PK, two essential rate-limiting enzymes involved in glycolysis. The finding that glucose deprivation or 2-DG treatment enhances the growth inhibition effect of EGCG indicates that glycolysis is related to pancreatic cancer cell growth.

PFK catalyzes the first irreversible reaction of glycolysis and is usually highly expressed in tumor tissues. The platelet isoform of PFK, PFKP, functions as an important mediator in cancer cell proliferation and metastasis [21,22]. On the other hand, PKM2, an isoform of PK, is involved in regulating cell growth and metastasis [23]. Our results showed that EGCG decreased the enzyme activity and protein expression of PFK and PK in vitro and in vivo, but not HK2 or LDHA. Of note, silencing PFKP had a slight additive effect on the growth inhibitory effect of EGCG, suggesting that regulating glycolysis represents an important mechanism of EGCG in inhibiting pancreatic cancer cell growth. These results are consistent with other studies that also report inhibition of growth through the suppression of glycolysis by EGCG, such as in breast and hepatocellular cancer cell models [11,12].

The induction of oxidative stress plays a significant role in the effect of many anticancer agents [24]. Compared with normal cells, cancer cells exhibit higher levels of ROS and antioxidant levels to maintain redox homeostasis. Cancer cells are thus more susceptible to oxidative stress [15], which precedes the induction of apoptotic cell death [25]. EGCG induced ROS levels in pancreatic cancer cells in culture, consistent with its known pro-oxidant activity [26]. Interestingly, the inhibitory effect of EGCG on PFKP and PKM2 levels was mostly reversed by catalase, suggesting that the effect of EGCG on glycolytic processes is ROS-dependent. However, catalase only partly prevented the growth inhibition effect of EGCG, indicating that the ROS produced by EGCG could only explain part of the growth inhibitory effect by EGCG. Interestingly, we have recently shown that EGCG inhibits the protein kinase B (PKB, also known as Akt) pathway though a ROS-independent effect [27]. Thus, the effect of EGCG on the growth of pancreatic cancer cells appears to be the result of the sum of EGCG's ROS-dependent and independent effects. Moreover, the significance of oxidative stress in the reduction of cell growth can be further evidenced by two manipulations of the system affecting the levels of glutathione. Pretreatment with BSO, which depletes intracellular glutathione, strongly enhanced the growth inhibitory effect of EGCG. Alternatively, supplementing the cells with NAC attenuated the growth inhibitory effect of EGCG.

A common practice in the clinic is to administer multiple drugs concomitantly to cancer patients to help enhance a beneficial effect at lower doses while reducing side effects. The combination of gemcitabine with Abraxane® is the most widely used regimen for patients with newly diagnosed pancreatic cancer [28]. Another option is the combined therapy of leucovorin-modulated 5-Fluorouracil (5-FU), irinotecan, and oxaliplatin (FOLFIRINOX). Unfortunately, this regimen is given only to patients that can tolerate its toxicity, which includes a higher degree of neutropenia, diarrhea, and sensory neuropathy [29]. In this work, we explored whether EGCG could enhance the anticancer efficacy of the Food and Drug Administration (FDA)-approved drugs mentioned above. EGCG successfully enhanced the cell growth inhibition effect of gemcitabine in vitro and in vivo. Importantly, EGCG plus gemcitabine, at their effective doses, appeared to be safe to mice, being well tolerated and showing no signs of liver toxicity. As an analog of deoxycytidine, gemcitabine enters the cells, embeds into DNA, inhibits DNA synthesis, and induces cell cycle arrest. EGCG enhanced gemcitabine's suppression of cell growth through the inhibition of cell proliferation, the arrest of the cell cycle, and the induction of apoptosis through activation of execution Caspases [30], modulation of Bcl-2, and inhibition of apoptosis protein families. Mechanistically, the two agents together showed an additive effect of inhibiting glycolysis in pancreatic cancer cells. Of note, additional studies in complementary preclinical pancreatic cancer models are needed to validate the anti-tumor effect of EGCG with gemcitabine and advance the preclinical development of this drug combination. In summary, these results suggest the possibility to utilize EGCG as a useful adjuvant drug with gemcitabine to inhibit pancreatic cancer by further modulating cell kinetics and suppressing glycolysis.

#### **4. Materials and Methods**

#### *4.1. Chemicals and Reagents*

EGCG (≥98% purity) was purchased from Tocris (Minneapolis, MN, USA). 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (≥97.5%), D-(+)-Glucose (≥99.5%), Oligomycin (≥90%), 2-Deoxy-d-glucose (≥99%), Irinotecan hydrochloride, Oxaliplatin, Catalase, *<sup>N</sup>*-Acetyl-l-cysteine (≥99%), DL-Buthionine-sulfoximine (≥99%), RIPA lysis buffer, Halt Protease Inhibitor Cocktail, Phosphatase Inhibitor Cocktail, and the following kits: Phosphofructokinase Activity Colorimetric assay and Pyruvate Kinase Activity assay were purchased from MilliporeSigma (St. Louis, MO, USA). Seahorse XF24 Extracellular Flux assay kits were purchased from Agilent (Santa Clara, CA, USA). CellTiter-Glo® reagent and rATP were purchased from Promega (Madison, WI, USA). PFKP siRNA plasmid was purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Gemcitabine-HCL (>99%) was purchased from BIOTANG (Waltham, MA, USA). 5-Fluorouracil (≥99%) was purchased from Alfa Aesar (Haverhill, MA, USA). Annexin V-FITC conjugate, propidium iodide (PI), 2',7'-dichlorodihydrofluorescein diacetate (H2DCFDA), Amplex® Red Hydrogen Peroxide kit, MitoSOX™ Red Mitochondrial Superoxide Indicator, Lipofectamine™ 3000 Transfection reagent, and SuperSignal™ West Dura Extended Duration Substrate were purchased from ThermoFisher Scientific (Waltham, MA, USA). Antibodies for Western blot were purchased from Cell Signaling Technology (Danvers, MA, USA). Bradford protein assay reagent, 30% (*w*/*v*) Acrylamide/Bis Solution, 4xLaemmli sample buffer, and Immun-Blot® PVDF Membranes were purchased from Bio-Rad (Hercules, CA, USA).

#### *4.2. Cell Culture*

Human pancreatic cancer cell lines (BxPC-3, HPAF-II, CFPAC-1, Su.86.86, Panc-1 and MIA PaCa-2), and human pancreatic normal epithelial (HPNE) cells were sourced from the American Type Culture Collection (Manassas, VA, USA). FC1245 cells (KPC, murine pancreatic cancer cells) were a gift from Dr. David Tuveson (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA). All cell lines were grown as monolayers in a specific medium under conditions suggested by the vendor. Although these cells lines were not authenticated in our lab, they were characterized by cell morphology and growth rate, and cultured in our laboratory less than six months after being received. We also routinely test for mycoplasma contamination in all cell lines every three months.

#### *4.3. Cell Viability*

Following the treatment with EGCG or the various chemotherapeutic drugs for 72 h, the reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dye (MTT) was determined according to the manufacture's protocol (MilliporeSigma, St. Louis, MO, USA).

#### *4.4. Cellular Glycolytic Rate Measurements*

The cellular glycolytic rate, represented as the extracellular acidification rate (ECAR), was measured using a Seahorse XF24 Extracellular Flux Analyzer (Agilent, Santa Clara, CA, USA). Briefly, Panc-1 and MIA PaCa-2 cells, plated on XF24 cell culture plates, were incubated with the agents for 24 h, and then assayed with a glycolytic stress test kit, following the manufacturer's instruction (Agilent, Santa Clara, CA, USA).

#### *4.5. Cellular ATP Levels*

Cells, plated in white 96-well plates, were treated with test drugs for 24 h. After treatment, ATP levels were measured with the CellTiter-Glo® reagent, following the manufacturer's protocol (Promega, Madison, WI, USA). A standard curve was generated using escalating concentrations of ATP.

#### *4.6. Glycolysis-Related Enzymes Activity*

Cells were seeded into the 6-well plates overnight and treated with EGCG for 24 h. Following the incubation, cells were collected, washed with cold PBS, and homogenized. After centrifugation, the supernatant was collected and used to measure the phosphofructokinase (PFK) and pyruvate kinase (PK) enzyme activities following the manufacturer's protocol (MilliporeSigma, Saint Louis, MO, USA). Protein concentration was determined using the Bradford protein assay.

#### *4.7. Gene Silencing*

Cells were plated in 6-well plates overnight, and transiently transfected with PFKP siRNA or nonspecific control siRNA for several hours using LipofectamineTM 3000 reagent according to the manufacturer's instructions (ThermoFisher Scientific, Waltham, MA, USA). Following transfection, cells were replated and treated with EGCG for up to 72 h and cell viability was tested. The gene silencing efficiency was determined by immunoblotting.

#### *4.8. Determination of ROS Levels*

After treatment with EGCG for 24 h, cells were incubated with 10 μM H2DCFDA for 30 min at 37 ◦C and their fluorescence intensity was analyzed using a Synergy H1 microplate reader (Biotek, Winooski, VT, USA). Hydrogen peroxide (H2O2) levels were detected using an Amplex™ Red hydrogen peroxide kit according to the manufacturer's instruction (ThermoFisher Scientific, Waltham, MA, USA).

#### *4.9. Mitochondrial Superoxide Level Analysis*

After treatment with EGCG for 24 or 48 h, cells were collected and incubated with 5 μM MitoSOXTM Red mitochondrial superoxide probe at 37 ◦C for 30 min. The fluorescence intensity was determined by FACScan flow cytometry (Becton Dickinson, San Jose, CA, USA) and results were analyzed with FlowJo software (v7.6, Tree Star, Inc., Ashland, OR, USA).

#### *4.10. Cell Apoptosis*

After cells were treated with the test agents in 6-well plates, they were trypsinized and stained with Annexin V-FITC (100× dilution) and PI (0.5 μg/mL) for 15 min. Annexin V-FITC and PI fluorescence intensities were analyzed by FACScan (Becton Dickinson, San Jose, CA, USA). Annexin V (+)/PI (-) cells are apoptotic cells, Annexin V (+)/ PI (+) cells have undergone secondary necrosis, and Annexin V (-)/ PI (+) cells are necrotic cells. Results were analyzed by using FlowJo software.

#### *4.11. Cell Cycle Analysis*

Cells were seeded in 6-well plates and treated for 48 h. After each treatment, cells were trypsinized and fixed in 70% (*v*/*v*) ethanol overnight, stained with PI (50 μg/mL) and RNase A (10 mg/mL) for 15 min, and subjected to flow cytometric analysis (Becton Dickinson, San Jose, CA, USA).

#### *4.12. Western Blot*

Whole cell protein lysates were prepared, and electrophoresis and electroblotting were performed as previously described [31]. Membranes were probed overnight with the following primary antibodies (1:1000 dilution) from Cell Signaling Technology (Danvers, MA, USA): PFKP (Cat #8164), PKM2 (Cat #4053), HK2 (Cat #2867), LDHA (Cat #3582), phospho-Chk1 (Ser345) (Cat #2348), phospho-p53 (Ser15) (Cat #9286), p53 (Cat #2527), p21 Waf1/Cip1 (Cat #2947), cdc2 (Cat #28439), Cyclin B1 (Cat #12231), Caspase-3 (Cat #14220), Caspase-7 (Cat #12827), Caspase-9 (Cat #9508), and PARP (Cat #9542). β-Actin (Cat #8457) was used at the same time as a loading control. After incubation for 60 min at room temperature in the presence of the secondary antibody (HRP-conjugated; 1:5000 dilution), the conjugates were developed and visualized using a Molecular Imager FXTM System (BioRad; Hercules, CA, USA).

#### *4.13. Animal Studies*

All animal studies were approved by the Institutional Animal Care and Use Committee at the University of California, Davis (protocol # 20716; approved on September 6, 2018). For the efficacy study, C57BL/6J mice (4–6 weeks) were bilaterally, s.c injected with 0.3 <sup>×</sup> 106 KPC cells/tumor suspended in 0.1 mL sterile PBS. When KPC cells reached palpable tumor size (~300 mm3), mice (n = 5/group) were divided randomly into four groups. Mice were either given vehicle, EGCG 10 mg/kg, 7x/week by intraperitoneal injection (I.P.) injections, gemcitabine 100 mg/kg, 2x/week by i.p injections, or EGCG in combination with gemcitabine at the above doses. The dose of EGCG was based on our previous studies [11,27]. Mice were treated for 16 days. Tumor size and body weight were measured every two days, and tumor size was determined by the equation length × width × (length + width/2) × 0.56, in millimeters [32]. At the end of the study, animals were euthanized by carbon dioxide asphyxiation, and tumor weights measured. Tumor and liver tissues were collected for analysis. For the drug combination toxicity study, C57BL/6J mice (n = 4/group) were treated either with PBS (vehicle control), or EGCG 10 mg/kg, 7x/week by i.p injections, in combination with gemcitabine 100 mg/kg, 2x/week by i.p injections. On day 16, mice were euthanized, blood was drawn, serum was collected, and a liver-kidney function panel was performed.

#### *4.14. Immunohistochemistry*

Immunohistochemical staining for for Ki-67 (Cat #12202) and cleaved Caspase-3 (Cat #9661, both from Cell Signaling Technology, Danvers, MA, USA) was performed as previously described [33]. Briefly, paraffin-embedded sections (5 μm thick) were deparaffinized and rehydrated, followed by antigen retrieval performed by microwave-heating in 0.01 M citrate buffer (pH 6.0). 3% H2O2 was used to block endogenous peroxidase activity for 10 min at room temperature. Slides were blocked for 60 min with serum, and incubated with primary antibody overnight at 4 ◦C. The following morning, slides were washed thrice with PBS, and then incubated with the biotinylated secondary antibody and the streptavidin-biotin complex (Invitrogen, Carlsbad, CA, USA) for 1 h of each at room temperature. After washing with PBS three times, slides were stained with 3,3- -Diaminobenzidine tetrahydrochloride hydrate (DAB) solution, and then counterstained with hematoxylin. Images were taken at 100× magnification. At least five fields per sample were scored and analyzed using Image J software (v1.46, NIH, Bethesda, MD, USA).

#### *4.15. Statistical Analysis*

Data were obtained from at least three independent biological experiments and results expressed as mean ± SD. One-way analysis of variance (ANOVA) and the Duncan test were used to analyze differences among multiple groups. *T*-tests were performed to compare the difference between two groups. *p* < 0.05 was regarded as being statistically significant.

#### **5. Conclusions**

EGCG strongly suppresses glycolysis through the inhibition of PFK, an effect that is ROS-dependent. In addition, EGCG presents a strong additive effect when combined with gemcitabine in pancreatic xenografts by further inhibiting glycolysis and affecting cell kinetics. Although additional studies to validate the above findings in complemmentary preclinical models of pancreatic cancer are warranted, our results suggest that EGCG is a useful combination partner of gemcitabine in pancreatic cancer treatment.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6694/11/10/1496/s1, Figure S1: EGCG reduced cell glycolysis, glycolytic capacity and glycolytic reserve in Panc-1, MIA PaCa-2 and KPC cells. Figure S2: Effect of EGCG on HK2 and LDHA protein expression. Table S1: Cell growth combination effects of EGCG with various chemotherapeutics in MIA PaCa-2 and Panc-1 cells. Table S2: Serum levels of multiple biochemical enzymes and markers of liver and kidney function for control and EGCG plus gemcitabine the end of the treatment period. Figure S3: Mice body weight progression for control and gemcitabine+EGCG

treated groups. Figure S4. Western blot images with molecular weight for PFKP and PKM2 shown in Figure 2B. Figure S5. Western blot images with molecular weight for PFKP and PKM2 shown in Figure 2C. Figure S6. Western blot images with molecular weight for PFKP and PKM2 shown in Figures 2D and 2E. Figure S7. Western blot images with molecular weight for PFKP and PKM2 shown in Figure 3D. Figure S8. Western blot images with molecular weight for p-Chk1, p53, p21, cdc2 and CyclinB1 shown in Figure 6B. Figure S9. Western blot images with molecular weight for Caspase 3, Caspase 7, Caspase 9 and PARP shown in Figure 6D. Figure S10. Western blot images with molecular weight for Bcl-xl, Bad, XIAP, Survivin shown in Figure 6E. Figure S11. Western blot images with molecular weight for PFKP and PKM2 shown in Figure 7B. Figure S12. Western blot images with molecular weight for HK2 and LDHA shown in Figure S2.

**Author Contributions:** Conceptualization, R.W. and G.G.M.; methodology, R.W. and G.G.M.; validation, R.W.; formal analysis, R.W.; investigation, R.W.; resources, G.G.M.; data curation, R.W.; writing—original draft preparation, R.W.; writing—review and editing, R.W., R.M.H., Y.W. and G.G.M.; supervision, G.G.M.; project administration, G.G.M.; funding acquisition, G.G.M.

**Funding:** This study was supported by the funds from the University of California, Davis and NIFA-USDA (CA-D-XXX-2397-H) to GGM. Ran Wei is sponsored by a China Scholarship Council fellowship. Flow cytometry experiments were funded in part by the UC Davis Comprehensive Cancer Center Support Grant (CCSG) (NCI P30CA093373).

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

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Review*

### **EGCG Mediated Targeting of Deregulated Signaling Pathways and Non-Coding RNAs in Di**ff**erent Cancers: Focus on JAK**/**STAT, Wnt**/β**-Catenin, TGF**/**SMAD, NOTCH, SHH**/**GLI, and TRAIL Mediated Signaling Pathways**

#### **Ammad Ahmad Farooqi 1, Marina Pinheiro 2,\*, Andreia Granja 2, Fulvia Farabegoli 3, Salette Reis 2, Rukset Attar 4, Uteuliyev Yerzhan Sabitaliyevich 5, Baojun Xu <sup>6</sup> and Aamir Ahmad <sup>7</sup>**


Received: 16 March 2020; Accepted: 4 April 2020; Published: 12 April 2020

**Abstract:** Decades of research have enabled us to develop a better and sharper understanding of multifaceted nature of cancer. Next-generation sequencing technologies have leveraged our existing knowledge related to intra- and inter-tumor heterogeneity to the next level. Functional genomics have opened new horizons to explore deregulated signaling pathways in different cancers. Therapeutic targeting of deregulated oncogenic signaling cascades by products obtained from natural sources has shown promising results. Epigallocatechin-3-gallate (EGCG) has emerged as a distinguished chemopreventive product because of its ability to regulate a myriad of oncogenic signaling pathways. Based on its scientifically approved anticancer activity and encouraging results obtained from preclinical trials, it is also being tested in various phases of clinical trials. A series of clinical trials associated with green tea extracts and EGCG are providing clues about significant potential of EGCG to mechanistically modulate wide ranging signal transduction cascades. In this review, we comprehensively analyzed regulation of JAK/STAT, Wnt/β-catenin, TGF/SMAD, SHH/GLI, NOTCH pathways by EGCG. We also discussed most recent evidence related to the ability of EGCG to modulate non-coding RNAs in different cancers. Methylation of the genome is also a widely studied mechanism and EGCG has been shown to modulate DNA methyltransferases (DNMTs) and protein enhancer of zeste-2 (EZH2) in multiple cancers. Moreover, the use of nanoformulations to increase the bioavailability and thus efficacy of EGCG will be also addressed. Better understanding of the pleiotropic abilities of EGCG to modulate intracellular pathways along with the development of effective EGCG delivery vehicles will be helpful in getting a step closer to individualized medicines.

**Keywords:** EGCG; signaling pathways; non-coding RNAs; anti-cancer drug

#### **1. Introduction**

Genomic approaches such as whole genome sequencing and genetic mapping have helped considerably in the identification of many genetic variants in multiple components of cell signaling pathways. Moreover, advancements in functional genomics have marked a new frontier in molecular oncology. Epigallocatechin-3-gallate (EGCG) is a phenolic compound present in tea and has captivated tremendous attention in the past two decades because of its premium pharmacological properties. There is a wide variety of reviews published with reference to EGCG mediated anticancer effects [1–4]. However, in this review we focused on EGCG mediated modulation of deregulation cell signaling pathways in different cancers. We partitioned this multi-component review into different sections. We will open the review by critical analysis of layered regulation of the JAK-STAT pathway by EGCG.

#### **2. Targeting of JAK**/**STAT Signaling**

The JAK-STAT pathway constitutes a rapid membrane-to-nucleus signaling module that has been shown to play fundamental role in cancer development and progression (shown in Figure 1). In this section, we will discuss in detail how EGCG modulated JAK/STAT signaling. EGCG has been shown to interfere with the JAK/STAT pathway at different steps, which includes inhibition of STAT phosphorylation and restriction of nuclear transportation of STAT proteins.

EGCG remarkably reduced tyrosine and serine phosphorylation of signal transducer and activator of transcription 1 (STAT1) [5]. Moreover, phosphorylation of protein kinase C delta PKC-delta, Janus kinase 1 (JAK1), and Janus kinase 2 (JAK2), which are the upstream activators of STAT1 are also inhibited by EGCG in interferon gamma (IFNγ)-stimulated oral cancer cells (shown in Figure 1) [5]. EGCG-mono-palmitate (EGCG-MP), a highly active derivative of EGCG effectively activated Src homology 2 domain-containing tyrosine phosphatase-1 (SHP-1) which consequentially resulted in reduction of phosphorylated levels of BCR-ABL and signal transducer and activator of transcription 3 (STAT3) in human chronic myeloid leukemia (CML) cells (shown in Figure 1) [6]. EGCG-MP treatment more efficiently induced regression of tumor growth in BALB/c athymic nude mice [6]. EGCG potently inhibited BCR/ABL oncoprotein and the JAK2/STAT3/AKT pathway in BCR/ABL+ CML cell lines [7]. Curcumin worked synchronously with EGCG and considerably interfered with tumor conditioned media-induced transition of normal endothelial cells toward tumor endothelial cells by inhibition of the JAK/STAT3 signaling pathway [8].

EGCG significantly reduced phosphorylation of STAT3 on the 705th tyrosine residue and improved sensitivity of cisplatin-resistant oral cancer cells [9]. Fundamental role-play of STAT signaling had previously been studied in invasive breast cancers and matched lymph nodes using quantitative immunofluorescence [10]. STAT proteins were analyzed in lymph nodes and paired primary breast cancer tissues. There was higher expression of cytoplasmic STAT1, p-STAT3 (Ser727), STAT5, and nuclear p-STAT3 (Ser727) in the nodes [10]. c-Myb overexpression induced activation of NF-κB and STAT3 signaling to enhance proliferation, invasion, and resistance against cisplatin [11]. However, c-Myb silencing inhibited proliferation, invasive potential, and sensitized ovarian cancer cells to cisplatin. EGCG completely inhibited c-Myb-mediated proliferative and invasive abilities of ovarian cancer cells [11].

EGCG dose-dependently reduced phosphorylated levels of STAT1 and STAT3 [12]. Quercetin and EGCG worked synergistically and exerted inhibitory effects on cytokine-mediated upregulation of iNOS (inducible nitric oxide synthase) and ICAM-1 (intercellular adhesion molecule-1) via JAK/STAT cascade in cholangiocarcinoma cells (Figure 1) [12].

Indoleamine 2,3-dioxygenase (IDO) is a tryptophan catabolic enzyme. IDO mechanistically regulates immunological response and enables tumor cells to evade the immune system [13]. IFN-γ increased mRNA and protein levels of IDO in HT29 and SW837 colorectal cancer cells. EGCG dose-dependently decreased IFN-γ-induced expression of IDO in SW837 cells. Increase in p-STAT1 level induced by IFN-γ was also found to be markedly repressed by EGCG. Data obtained from reporter assays clearly revealed that EGCG inhibited the transcriptional activity of IDO promoter and blocked binding of p-STAT1 to gamma-activated sequence (GAS) sites on the promoters of target genes (Figure 1) [13].

Toxicological analysis of EGCG highlighted its efficacy and minimum off-target effects. Orally administered EGCG mitigated cisplatin-induced hearing loss along with a marked reduction in the loss of outer hair cells in the basal cochlear region. Importantly, chemotherapeutic drug-induced toxicity was also reduced mainly though suppression of apoptotic markers and oxidative stress [14].

**Figure 1.** Regulation of the JAK/STAT pathway by epigallocatechin-3-gallate (EGCG). (**A**,**B**) Janus kinase (JAK) mediated phosphorylation of STAT proteins promoted their accumulation in nucleus to regulate expression of a plethora of genes. (**C**–**E**) EGCG showcased remarkable ability to shut down the JAK/STAT pathway by inhibition of Janus kinase 1 (JAK1), Janus kinase 2 (JAK2), signal transducer and activator of transcription 1 (STAT1), signal transducer and activator of transcription 3 (STAT3). EGCG also activated negative regulators of STAT-driven signaling. Activation of Src homology 2 domain-containing tyrosine phosphatase-1 (SHP-2) was effective in inhibition of JAK/STAT signaling. Different oncogenes particularly, inducible nitric oxide synthase (iNOS), intercellular adhesion molecule-1 (ICAM-1), and indoleamine 2,3-dioxygenase have been shown to be under direct control of STAT signaling. (**F**,**G**) Vascular endothelial growth factor vascular endothelial growth factor receptor (VEGF/VEGFR) signaling is also regulated by EGCG. EGCG interacted with VEGF. Additionally, EGCG inhibited phosphorylation of VEGFR.

It has recently been reported that IFNγ-mediated PD-L1 levels were noted to be downregulated after treatment with green tea extracts and EGCG mainly through inhibition of JAK2/STAT1 signaling in A549 cells [15]. Likewise, EGF-stimulated PD-L1 upregulation was reduced in EGCG-treated Lu99 cells by inactivation of EGFR/AKT transduction cascade. Additionally, green tea extracts notably reduced average number of tumors and percentage of PD-L1<sup>+</sup> cells in lungs of A/J mice intraperitoneally injected with a cigarette smoke toxin. EGCG reduced mRNA levels of PD-L1 in F10-OVA cells and enhanced expression of interleukin-2 in tumor-specific CD3<sup>+</sup> T cells [15]. Collectively these findings suggested that green tea catechin acted as a useful immunological checkpoint inhibitor.

Confluence of information suggested central role of JAK/STAT signaling in different cancers. EGCG mediated inhibition of JAK/STAT signaling via activation of negative regulators (SHP-2) and inactivation of positive regulators (JAK1, JAK2) has gradually gained appreciation. Additionally, different fusion oncoproteins (BCR-ABL) are also exclusively targeted by EGCG.

#### **3. VEGF**/**VEGFR Signaling**

EGCG and silibinin worked synergistically and inhibited vascular endothelial growth factor/vascular endothelial growth factor receptor (VEGF/VEGFR) signaling. EGCG and Silibinin also reduced migratory potential of A549 cells [16]. EGCG interacted with VEGF mainly through hydrophobic interactions and induced a change in the secondary structure of the protein (Figure 1) [17].

Vandetanib (ZD6474), a VEGFR inhibitor was co-loaded with EGCG in mesoporous Silica-Gold nanoclusters for effective targeting of tamoxifen-resistant breast cancer cells [18]. Vandetanib and EGCG effectively reduced phosphorylated levels of EGFR2 and VEGFR2 in drug-resistant breast cancer cells [18]. EGCG also worked with superior efficacy when used in combination with tamoxifen. Tamoxifen worked powerfully with EGCG and reduced the levels of EGFR1, VEGF, and VEGFR1 in breast cancer cells [19]. SU5416 (Semaxanib) also worked remarkably with EGCG and induced apoptosis in malignant neuroblastoma SK-N-BE2 and SH-SY5Y cells [20]. SU5416 and EGCG also inhibited VEGFR2 expression [20].

EGCG dose-dependently decreased levels of VEGFR2 and p-VEGFR2 in HCC and colorectal cancer cells (Figure 1) [21,22]. EGCG induced regression of tumors in mice xenografted with either HuH7 or SW837 cells. EGCG decreased total and phosphorylated levels of VEGFR2 in these xnografts [21,22].

Detailed mechanistic insights revealed that p-STAT1 and p-STAT3 formed complexes with VEGFR1 and VEGFR2 in chronic lymphocytic leukemia (CLL) cells [23]. VEGF induced nuclear accumulation of p-STAT3 in primary CLL B cells. VEGF/VEGFR complex facilitated shuttling of STAT3 from the plasma membrane to perinuclear regions. VEGF induced co-localization of STAT3, VEGFR1 and VEGFR2 to the same perinuclear regions. Collectively these findings provided clear evidence that the VEGF/VEGFR pathway "switched on" STAT proteins which induced resistance against apoptosis. EGCG decreased levels of p-STAT3 [23]. EGCG also remarkably reduced phosphorylated levels of VEGFR1 and VEGFR2 in B-cell chronic lymphocytic leukemia cells [24].

#### **4. Regulation of Methylation-Associated Machinery**

PRC2 (Polycomb repressive complex-2) is a transcriptional repressive complex that consists of three essential proteins: EZH2 (enhancer of zeste-2), EED (embryonic ectoderm development), and SUZ12 (suppressor of zeste 12). A series of structural studies have shown that EZH2 context-dependent trimethylates lysine 27 on histone 3 (H3K27) to promote transcriptional inactivation of target genes (shown in Figure 2).

EZH2-mediated trimethylation of H3K27 induced transcriptional repression of TIMP3 (tissue inhibitor of metalloproteinases-3). However, EGCG demonstrated remarkable ability to inhibit EZH2-mediated trimethylation. There was a considerable reduction in the levels of enhancers of zeste homolog 2 (EZH2) and H3K27me3 repressive marks at the promoter region of TIMP-3. Additionally, there was an evident increase in histone H3K9/18 acetylation [25]. Essentially, green tea polyphenols and EGCG treatment significantly reduced class I histone deacetylases (HDAC) activity/expression in prostate cancer cells. Furthermore, levels of EZH2 and H3K27me3 were also found to be reduced in prostate cancer cells [25]. Data clearly suggested that EGCG efficiently demonstrated multi-layered regulation of HDACs and EZH2.

Due to the fundamental role of EZH2 in cancer progression, different inhibitors of EZH2 have been designed and tested for evaluation of efficacy. EGCG and GSK343 (EZH2 inhibitor) exerted inhibitory effects on the proliferation, invasive and migratory potential of the cells, and suppressed EZH2-mediated trimethylation of H3K27 [26].

Recent advancements in the biochemical characterization of polycomb-group (PcG) complexes have revealed a broad range of new proteins, which assemble to form multi-protein complexes. All PRC1 complexes contain Ring1B, which has the E3 ubiquitin ligase activity of the complex. Complexes also include PCGF4/BMI-1 in association with Ring1B to regulate epigenetic modifications [27]. EGCG reduced BMI-1 and EZH2 levels in SCC-13 cells [28].

PML–RARα homodimers worked synchronously with co-repressors and histone deacetylases (HDACs) and consequentially enhanced DNA methylation [29]. EGCG reduced the levels of HDAC1 and PML/RARα in leukemic cells (Figure 2) [30].

Groundbreaking discoveries in biology of epigenome have enabled us to develop a sharp comprehension of highly intricate and well-coordinated interplay of HDACs, histone methyltransferases, and DNA methyltransferases. EGCG has emerged as a master-regulator of epigenetic-associated machinery.

Chromatin immunoprecipitation (ChIP) analyses revealed that EGCG enhanced hyperacetylated H4 and acetylated H3K14 histones within promoter regions of p27, PCAF, C/EBP and reduced binding of PRC2 core component genes EZH2, SUZ12, and EED [31].

EGCG significantly reduced enzymatic activities of DNA methyltransferase (DNMT) and HDAC in HeLa cells [32]. Moreover, EGCG also reduced expression level of DNMT3B whereas expression levels of HDAC1 remained unchanged [32]. GTP/EGCG-promoted acetylation of p53 and enhanced its binding to the promoters of Bax and p21/waf1. Treatment of cells with GTPs and EGCG dose- and time-dependently inhibited class I HDACs [33].

Am80 is a structurally different synthetic retinoid from all-trans-retinoic acid. EGCG and Am80 increased acetylated-p53 and acetylated-α-tubulin through suppression of HDAC activity. Use of specific inhibitors against HDAC4 and HDAC5 strongly induced p21waf1 gene expression. Additionally, HDAC6 inhibition induced upregulation of GADD153 and p21waf1 [34].

UHRF1 (ubiquitin-like containing PHD and Ring finger 1) contributed to inactivation of tumor suppressor genes by directing the binding of DNA methyltransferase 1 (DNMT1) to hemi-methylated promoters [35]. EGCG downregulated DNMT1 and UHRF1 expression and consequently upregulated p73 and p16 (INK4A) in Jurkat cells. UHRF1 downregulation was dependent upon the generation of ROS by EGCG. Upregulation of p16 (INK4A) correlated strongly with reduction in the binding of UHRF1 to the promoter region. UHRF1 overexpression counteracted EGCG-induced apoptosis and upregulation of p73 and p16 (INK4A) [35].

EGCG effectively reduced 5-methylcytosine, DNMT activity, mRNA and protein levels of DNMT1, DNMT3a, and DNMT3b [36]. EGCG decreased HDAC activity and increased levels of acetylated H3K9 and H3K14, H4K5, H4K12, and H4K16 but decreased levels of methylated H3-Lys 9. Collectively, because of inhibition of DNMTs and HDACs, EGCG induced re-expression of p16INK4a and Cip1/p21 [36].

Gazing through a molecular lens clearly highlighted contextual push and pull between various versatile regulators associated with methylation. Substantial fraction of information gathered through high-quality research has unraveled that a broad range of tumor suppressors are epigenetically silenced during cancer progression. Selective targeting of DNMTs and HDACs specifically in cancer cells is very challenging and needs to be comprehensively investigated in EGCG-treated preclinical models. In the upcoming section we will analyze how EGCG modulated deregulated TGF/SMAD signaling.

**Figure 2.** Interconnected and orchestrated interplay among various regulators of epigenetic modifying machinery. (**A**) Protein enhancer of zeste-2 (EZH2), embryonic ectoderm development (EED), and suppressor of zeste 12 (SUZ12) worked synchronously to trimethylate H3K27 and transcriptionally repressed tissue inhibitor of metalloproteinases-3 (TIMP-3). (**B**) Class 1 histone deacetylases (HDACs) were inhibited by EGCG to increase acetylation at H3K9 and H3K18. (**C**) PML–RARα homodimers worked collaboratively with HDAC to regulate expression of target genes. However, EGCG effectively inhibited PML–RARα and HDAC. (**D**) Acetylation of proteins has also been investigated. Acetylated p53 stimulated expression of Bax and p21. (**E**) Ubiquitin-like containing PHD and Ring finger 1 (UHRF1) and DNA methyltransferase (DNMT) also notably downregulated p16 and p73.

#### **5. TGF**/**SMAD Signaling**

Binding of TGFβ superfamily ligands to a type II receptor facilitated closer positioning of type I receptor and phosphorylated it [37]. More importantly, type-I receptor mediated phosphorylation of receptor-regulated SMADs (R-SMADs), which promoted formation of a complex with common mediator SMAD (co-SMAD) (shown in Figure 3). Structural studies had shown that the R-SMAD/co-SMAD complex accumulated in the nucleus to transcriptionally modulate the expression of target genes [38]. Epithelial to mesenchymal transition (EMT) is a highly complex mechanism induced by TGF/SMAD signaling. SMAD2/3 proteins have been shown to stimulate the expression of Snail and Slug in different cancers [39].

In this section, we will provide an overview of multi-layered regulation of TGF/SMAD signaling by EGCG in different cancers. Inhibition of phosphorylation of R-SMADs will inhibit TGF/SMAD signaling. Consequentially, TGF/SMAD signaling inhibition will result in repression of EMT-associated markers.

EGCG effectively reduced p-SMAD3, Snail, and Slug levels in ovarian cancer cells [40]. EGCG considerably suppressed EMT, invasive and migratory capacity of anaplastic thyroid carcinoma (ATC) 8505C cells by regulation of the TGFβ/SMAD pathway [41]. EGCG exerted inhibitory effects on TGFβ1-induced expression of EMT markers (vimentin) in 8505C cells. EGCG was noted to completely block the phosphorylation of SMAD2/3 and nuclear accumulation of SMAD4 [41].

Apart from phosphorylation, acetylation of SMAD proteins is also an intricate mechanism. p300/CBP, a histone acetyltransferase, has been shown to post-translationally modify SMAD proteins. TGFβ1-driven activation of p300/CBP mediated EMT mainly through acetylation of SMAD2 and SMAD3 [42]. EGCG inhibited p300/CBP activity in lung cancer cells. EGCG strongly repressed

TGFβ1-induced EMT and reversed the upregulation of different target genes associated with EMT. EGCG inhibited TGFβ1-mediated activation of p300/CBP. EGCG inhibited TGFβ1-mediated EMT by interfering with the acetylated state of SMAD2 and SMAD3 in lung cancer cells [42].

TGFβ potently induced epithelial–mesenchymal transition (EMT) in NSCLC cells but EGCG reversed TGFβ-induced morphological alterations [43]. EGCG upregulated the expression of E-cadherin and downregulated the expression of vimentin. Data obtained through immunofluorescence also provided clear clues that E-cadherin was upregulated, and vimentin was downregulated by EGCG [43]. Moreover, EGCG effectively inhibited TGFβ-induced migratory and invasive potential of NSCLC cells. EGCG inhibited TGFβ-induced EMT at the transcriptional level. Expectedly, EGCG reduced phosphorylated levels of ERK1/2 (extracellular signal-regulated protein kinases 1/2) and SMAD2 and also inhibited the accumulation of SMAD2 in the nucleus. EGCG repressed the expression of transcriptional factors Slug, Snail, Twist, and ZEB1 and upregulated E-cadherin expression (Figure 3) [43].

Interestingly, different peptide aptamers have been designed to effectively inhibit interaction of SMAD2 and SMAD3 with SMAD4. Therefore, it might be advantageous to combine EGCG with different TGFβ signaling inhibitors to inhibit tumor growth in xenografted mice. More importantly, it will also be exciting to evaluate EGCG-mediated regulation of negative regulators (SMURFs and NEDDs) of the TGF/SMAD pathway.

**Figure 3.** (**A**,**B**) Binding of TGFβ superfamily ligands to a type II receptor induced juxtapositioning of type I receptor. Phosphorylation of SMAD2/3 promoted its accumulation in the nucleus. SMAD2/3 have been shown to stimulate expression of Snail and Slug. Apart from phosphorylation, additional post-translational modifications, particularly acetylation, have also been observed in TGF/SMAD signaling. EGCG inhibited acetylation of SMAD proteins.

#### **6. Regulation of Wnt**/β**-Catenin Pathway**

Detailed mechanistic insights revealed that in the absence of Wnt signal, β-catenin was phosphorylated by APC (adenomatous polyposis coli)/Axin/GSK3β complex and degraded by proteasome [44]. However, activation of the membrane receptors by Wnt signal resulted in the phosphorylation and degradation of GSK3β. EGCG inhibited phosphorylation of GSK3β, upregulated GSK3β expression, and decreased the levels of β-catenin in colorectal cancer cells [44].

O6-methylguanine (O6-meG) DNA-methyltransferase (MGMT) is a versatile mediator of temozolomide resistance in glioblastomas. TCF/LEF-binding sites within the promoter region of the MGMT gene have previously been identified [45]. Intriguingly, there is evidence of regulation of MGMT by WNT/β-catenin signaling. EGCG not only prevented translocation of β-catenin into the nucleus but also reduced the levels of transcriptional factors TCF1 and LEF1 [46]. Overall these findings clearly suggested that EGCG repressed MGMT expression via inhibition of the β-catenin-driven pathway.

EGCG not only reduced mRNA levels and transcriptional activities of β-catenin in p53 wild-type-expressing KB cells but also promoted ubiquitylation and degradation of β-catenin [47]. EGCG dose-dependently suppressed β-catenin expression in MDA-MB-231 cells [48]. EGCG worked synergistically with gemcitabine and exerted stronger inhibitory effects on β-catenin and N-cadherin in pancreatic cancer cells [49].

Clinical trials of CWP232291 (NCT01398462) and PRI-724 (NCT01302405; NCT01764477) are providing important clinically relevant information and it will be interesting to combine these agents with EGCG for evaluation of robust inhibition of β-catenin-driven signaling and tumor growth inhibitory effects in xenografted mice.

#### **7. Regulation of Notch Pathway**

The Notch signaling pathway consists of the Notch receptor, Notch ligand, DNA-binding protein, and Notch regulatory molecules. Notch is a transmembrane protein that mediates communication between neighboring cells. Binding of the ligands to a Notch receptor promoted proteolytic cleavage of NICD (Notch intracellular domain) and its consequential nuclear translocation where it complexed with CSL and formed NICD/CSL transcriptional activation assembly for stimulation of HES and HEY.

EGCG decreased mRNA levels of Notch1, Hey1, and Hes1 [50]. Western blot assay clearly indicated that EGCG dose-dependently reduced protein levels of Notch1 in cancer stem cells (CSCs) of head and neck squamous carcinoma (HNSC). Additionally, EGCG dose-dependently decreased the Notch promoter activity [50].

Tumor growth was significantly reduced in HuCC-T1 tumor-bearing mice subcutaneously injected with EGCG. Notch1 was found to be markedly reduced by EGCG treatment [51].

Expression levels of Hes1 and Notch2 were observed to be considerably reduced in EGCG treated colorectal cancer cells [52]. More importantly, EGCG inhibited Notch1 and cleaved-Notch1 in 5-fluorouracil-resistant colorectal cancer cells [53].

#### **8. Regulation of TRAIL Mediated Apoptosis**

Increasingly it is being realized that cancer cells harbor highly complex signaling networks that resist apoptotic programming. High-quality research works related to the TRAIL-mediated pathway in the past two decades have ignited the field of molecular oncology and yielded a stream of preclinical and clinical insights that have reshaped current knowledge of apoptotic cell death.

GCG and TRAIL synergistically reduced Bcl-XL, Bcl-2, and FLIP. Whereas, combinatorial treatment activated capase-8, -9, and -3 in nasopharyngeal carcinoma cells [54].

EGCG and TRAIL also worked effectively against renal cell carcinoma and pancreatic cancer cells [55,56].

EGCG restored sensitivity of HCC cells to TRAIL-induced apoptosis [57]. EGCG upregulated caspase-3 activity and simultaneously downregulated Bcl-2 levels. EGCG also induced upregulation of DR4 and DR5 [57]. EGCG and TRAIL robustly enhanced DR4 levels. Furthermore, FLIP levels were reduced in prostate cancer cells treated in combination with EGCG and TRAIL [58]. Collectively these findings suggested that EGCG might be helpful in increasing the protein levels as well as cell surface expression of death receptors. There is sufficient experimental evidence related to reduction in the cell surface levels of death receptors. Death receptors are internalized and degraded in various cancers. Therefore, EGCG might play its role in stabilizing the levels of death receptors.

PEA15 (phosphoprotein-enriched in astrocytes) is an oncoprotein [59]. It has previously been reported that AKT-induced PEA15 phosphorylation and increased its stability. EGCG downregulated PEA levels mainly through inactivation of AKT. However, overexpression of PEA15 severely impaired apoptotic cell death induced by EGCG and TRAIL [59].

Certain hints had emerged which highlighted that EGCG inhibited TRAIL-induced apoptosis and activated autophagic flux in colon cancer cells. Inhibition of autophagic flux induced death receptor-driven apoptosis in colon cancer cells [60].

These scientific findings are intriguing and future research must converge on identification of additional protein targets in the TRAIL-driven pathway. Essentially, the TRAIL mediated pathway is regulated by long non-coding RNAs as well. Therefore, it will be paramount to unravel underlying mechanisms of TRAIL resistance and identification of proteins, which can be targeted to restore apoptosis in TRAIL-resistant cancers. Keeping in view the fact that TRAIL-based therapeutics and death receptor-targeting agonistic antibodies have entered into various phases of clinical trials, any progress in improving the efficacy of TRAIL-based therapeutics will be advantageous.

#### **9. Regulation of Non-Coding RNAs by EGCG in Di**ff**erent Cancers**

Discovery of non-coding RNAs has revolutionized our current understanding of the mechanisms that regulate post-transcriptional processes. The field of non-coding RNA has been extensively explored and researchers have witnessed groundbreaking advancements in disentangling the complicated web ranging from biogenesis of non-coding RNAs to post-transcriptional regulation of a myriad of target mRNAs.

A wealth of information has unveiled a fundamental role of non-coding RNAs in different cancers and researchers have experimentally verified the effects of natural and synthetic products on wide-ranging microRNAs and long non-coding RNAs in different cancers.

#### *9.1. Tumor Suppressor miRNAs*

miR-485, a tumor suppressor microRNA, has been found to be frequently downregulated in various cancers. CD44 was directly targeted by miR-485 in A549-cisplatin resistant lung cancer cells. CD44 was overexpressed in A549-cisplatin resistant lung cancer cells but EGCG treatment exerted repressive effects on CD44 levels by enhancing miR-485-mediated targeting of CD44 [61]. EGCG also induced regression of tumors in mice xenografted with A549-cisplatin resistant lung cancer cells.

miR-485-5p directly targeted RXRα in drug-resistant lung cancer cells. EGCG repressed CSC-like properties via modulation of the miR-485-5p/RXRα axis [62]. miR-155 is a tumor suppressor miRNA reportedly involved in enhancing drug sensitivity of cancer cells [63]. EGCG promoted NF-κB mediated upregulation of miR-155. Resultantly, miR-155 enhanced drug sensitivity of colorectal cancer cells by directly targeting MDR1 [63]. IGF2BP1 and IGF2BP3 are direct targets of miR-1275 in different cancers [64]. EGCG stimulated the expression of miR-1275 and potentiated targeting of IGF2BP1 and IGF2BP3 by miR-1275 in HCC cells [65].

#### *9.2. Oncogenic miRNAs*

miR-221/222 played a central role in drug resistance. Knockdown of miR-221/222 inhibited proliferation of drug-resistant breast cancer cells [66]

Suberoylanilide hydroxamic acid (SAHA), an HDAC inhibitor worked effectively with EGCG and markedly reduced expression of miR-221/222 in triple-negative breast cancer cells [67].

#### *9.3. Targeting of Oncogenic LncRNAs*

SOX2OT variant 7 effectively promoted Notch3/DLL3 signaling in osteosarcoma stem cells (OSCs) [68]. NOTCH target genes HEY1 and HES1 were found to be notably enhanced in variant 7-expressing OSCs. EGCG efficiently inhibited the levels of HEY1 and HES1 in OSCs. However, EGCG mediated inhibitory effects were noted to be impaired in variant 7-expressing cells [68]. EGCG

mediated tumor regression was not observed in mice xenografted with variant 7-expressing OSCs. However, EGCG treatment and NOTCH3 knockdown induced reduction in tumor growth in mice inoculated with variant 7-expressing OSCs [68].

EGCG also downregulated lncRNA-AF085935 in HCC cells. It was suggested that lncRNA-AF085935 promoted proliferation of HCC cells [69]. However, the study did not clearly provide a link between lncRNA-AF085935 and its targets and how lncRNA-AF085935 regulated apoptosis and proliferation in HCC cells.

#### *9.4. Tumor Suppressor LncRNAs*

EGCG had been shown to induce the expression of cisplatin transporter CTR1 (copper transporter 1) in cancer cells [70]. EGCG upregulated CTR1 and enhanced accumulation of intracellular platinum in NSCLC cells. hsa-miR-98-5p suppressed CTR1, whereas NEAT1 (nuclear enriched abundant transcript 1) enhanced its expression. hsa-miR-98-5p had specific complementary binding sites for NEAT1. Essentially, NEAT1 acted as a competitive endogenous RNA and upregulated EGCG-triggered CTR1 by sponging away hsa-miR-98-5p in NSCLC cells [70].

It seems surprising to note that available scientific proof related to regulation of non-coding RNAs by EGCG is limited. Keeping in view the wealth of information about remarkable pharmacological properties of EGCG, it is paramount to uncover how EGCG modulated different miRNAs, lncRNA, and circular RNAs in different cancers. Identification of the list of tumor suppressor and oncogenic non-coding RNAs regulated by EGCG in different cancers will be highly valuable in combinatorial treatments.

#### **10. Nanotechnological Approaches for E**ff**ective Delivery of EGCG to the Target Sites**

Despite the ability of EGCG to modulate several cancer-related mechanisms there are still major hurdles for the establishment of EGCG in clinical settings. The therapeutic concentrations of EGCG (between 1 and 10 μM) in the majority of the studies are much higher than the concentrations monitored in the plasma or tissues of animals or in human plasma (usually lower than 1 μM) after tea ingestion. In fact, even after the consumption of 7–9 cups of tea the EGCG concentration in plasma was still lower than 1 μM [71] and for that reason the use of nanotechnology, particularly the development of nanoparticles (NPs) as drug delivery systems, represent a promising approach to increase the bioavailability of EGCG. Nanotechnology corresponds to the science that studies and creates materials with dimensions between 1 and 1000 nm. NPs have at least one of the dimensions in the nanoscale range [72]. There are several types of NPs and for more comprehensive and detailed information the reader can consult the following revisions [73–76]. The different properties of the NPs can be used for medical purposes. Due to their small scale, NPs are excellent drug carriers, and since they can be modified in various factors such as size, chemical composition, outer layer, and others they are very versatile [77]. Furthermore, NPs can modify the pharmacokinetics and the stability of the carrier compound, being, for that reason, a promising strategy to improve EGCG bioavailability profile [78]. Another interesting characteristic of NPs is the possibility to enhance the cellular uptake or even the cellular targeting by modifying the outer layer with different ligands expressed in the target cells to assign particular characteristics in a strategy known by active targeting [79]. This is a useful strategy to improve the bioavailability and stability of EGCG even further, enhancing the utilization options and ultimately enhancing the anti-cancer properties of EGCG. The main types of NPs used for the delivery of EGCG reported in the literature are gold, polymeric, lipid-based, and inorganic NPs (shown in Figure 4). The majority of the NPs are designed to be at the range of approximately 200 nm since this size allows the administration of the NPs by the oral and intravenous routes. Other types of NPs were also used for the encapsulation of EGCG for the purpose of cancer therapy, including carbohydrates, transition metals, and inorganic materials [80–82]. The use of targeting ligands further increased cancer cell specificity and improved the anti-tumor effects of EGCG and, for that reason, folic acid has been used frequently to functionalize the NPs, since the folic acid receptor is overexpressed in tumor cells. However, other ligands can also be used, including antibodies, carbohydrates, or polysaccharides and other molecules [83]. A summary of the studies using different EGCG nanocarriers for cancer management and carried out in cell lines and in animals is depicted in Table 1.


**Table 1.** Different types of EGCG nanocarriers for cancer management.

**Figure 4.** Main types of nanoparticles (NPs) used for the delivery of EGCG.

Gold NPs as EGCG delivery systems have been exploited in several types of cancer since gold has anti-cancer properties per se [85,100]. Several reports have described the in vitro and in vivo efficacy of gold NPs in conjugation with EGCG for cancer treatment, including for the bladder, melanoma, neuroblastoma, and hepatocarcinoma [84–87]. These nanocarriers also demonstrated a high biocompatibility, inducing low damage to human red blood cells and therefore no toxicity for the dose tested was observed. The NPs made of polymers approved and recognized as safe by the US Food and Drug Administration (FDA) are also suitable for cancer applications [89,101]. Several groups have already encapsulated EGCG into different polymeric NPs for cancer therapy, including for the treatment of prostate cancer, colorectal, breast cancer, melanoma, and gastrointestinal cancer [88–95]. Despite the high toxicity towards cancer cells these NPs demonstrated absence of toxicity for normal cells. Liposomes and lipid NPs are lipid-based NPs in composition and for that reason are biodegradable and present minimal levels of toxicity [97]. There are some studies reporting the use of lipid-nanocarriers for the delivery of EGCG to cancer cells [96–99]. All of the studies were used for the treatment of breast cancer with results that demonstrated efficacy and security in vitro and in vivo, including in the MDA-MB-231 cell line, which is a model of the triple-negative cancer and considered more aggressive and associated with poorer outcome than other types of breast cancer.

#### **11. Potential Clinical Applications**

EGCG drug delivery systems based in NPs might represent an extraordinary resource to improve the application of EGCG in chemoprevention or to introduce the use of EGCG in the therapy of cancer. The idea of using drug delivery systems, such as NPs for loading EGCG, preserving its structure, and allowing to circumvent the limitations of the low bioavailability associated with the oral administration of free EGCG has a tremendous potentiality since increasing the amount of EGCG inside the cells will potentialize the effect of EGCG in the molecular targets and the effect of deregulated oncogenic signaling cascades and, therefore, determine better cancer outcomes in comparison with free EGCG. For instance, EGCG loaded in polylactic acid–polyethylene glycol NPs preserved the biological activity and efficacy on molecular targets in vitro and in xenograft tumors with over 10-fold dose advantage in comparison with EGCG alone [91]. Indeed, in vitro and in vivo studies are mandatory to verify whether EGCG loaded in NPs maintain EGCG mechanism of action and to understand if the efficacy on molecular targets is at least retained or increased. In view of a safe application, the toxicity of engineered NPs associated with EGCG needs to be fully investigated. For instance, transition metal oxide NPs have been found to increase oxidative stress, disturb calcium homeostasis, and deregulate cell cycle [102]. The activation of the immune system, specifically macrophage activation and cytokine release has been also reported [103]. Thus, lipid-based NPs show higher level of biocompatibility and bioavailability, emerging as the best candidates for pharmaceutical and clinical applications. In this context, EGCG loaded solid lipid NPs as an oral delivery system did not show any toxicity in rats [104]. Different nanoformulations, including EGCG, also showed great biocompatibility with no or very modest toxicity in animal models [105,106]. All these findings encourage the efforts to invest in biocompatible EGCG NPs to be used on humans, as interventional studies in pre-cancerous lesions, including prostate, breast, colon, and Barret's Esophagus [107–110] demonstrated EGCG efficacy despite the poor bioavailability and low plasma concentrations. Therefore, EGCG NPs are expected to improve the chemopreventive effects and to widen the applications in pre-neoplastic lesions, where the results were unclear or incomplete. In addition, EGCG mechanism of action can be improved by the association with anti-cancer drugs already used in cancer treatment since numerous drugs used in cancer therapy, including doxorubicin, 5-flurouracile, cisplatin, paclitaxel, act synergistically with EGCG [111], the best combinations being predictable on the basis of in vitro and in vivo studies. Lastly, active targeting also represents a strategy to preferentially address NPs to cancer cells. Nanomedicine-based therapy is at the beginning, but in the context of cancer chemoprevention and therapy, EGCG NPs might become a powerful strategy over the conventional chemotherapy approach.

#### *Clinical Trials Evaluating EGCG*

Given the promising reports from preclinical studies, EGCG has been tested in various clinical studies. Postmenopausal women are at high risk of developing breast cancer and, therefore, EGCG safety clinical trials have been conducted targeting this population. EGCG can afford benefit in terms of regulating LDL-cholesterol as well as glucose and insulin, as reported by a double-blind, randomized, placebo-controlled intervention study in healthy postmenopausal women [112]. A subsequent ancillary study of a double-blind, randomized, placebo-controlled, parallel-arm trial further confirmed the benefit of EGCG but reported the total cholesterol levels reductions only in women with elevated baseline levels [113]. In postmenopausal women, a daily dose of 843 mg EGCG has been reported to be generally well-tolerated with only a small fraction (6.7%) of women reporting adverse events [114]. This dose of 843 mg EGCG, when administered for a year, can reduce mammographic density in relatively younger women (50–55 years) but not in postmenopausal women, as suggested by phase II trial [115]. Not only in breast cancer patients or women at high breast cancer risk, EGCG is well-tolerated by chronic lymphocytic leukemia (CLL) patients as well [116]. Further, EGCG, at a daily dose as low as 44.9 mg for 4 weeks prior to surgery, has been reported to result in increased bioavailability, including accumulations in breast tumor tissue, in early stage breast cancer patients [108].

A randomized trial reported no reduction in likelihood of prostate cancer in men with high-grade prostatic intraepithelial neoplasia, compared to placebo, after a year on 400 mg EGCG dose per day [117]. It is possible that this might be related to the dose tested in this study as a previous study which tested the effects of 800 mg EGCG administered to 26 patients with positive prostate biopsies reported significant reductions in PSA, HGF, and VEGF, with no associated liver toxicity [118]. Similarly, a phase II pharmacodynamic prevention trial in bladder cancer patients indicated a possible reduction in PCNA and clusterin levels upon 2–4 weeks administration of EGCG prior to transurethral resection of bladder tumor or cystectomy [119].

EGCG has been tested in cancer clinical trials not just for the direct anticancer effects, but also for possible effects on co-morbidities. In lung cancer patients with an unresectable stage III disease, a phase I study was conducted to evaluate the efficacy of EGCG against chemotherapy related esophagitis [120]. Patients, divided in six cohorts receiving six different doses of EGCG, were administered EGCG once grade 2 esophagitis occurred. The study reported dramatic regression of esophagitis to grade 0/1 in 22 of 24 patients (91.7% cases), thus underlying the effectiveness of EGCG. On similar lines, a prospective phase II trial confirmed that EGCG can be effective against acute radiation-induced esophagitis as well [121]. Topical administration of EGCG to the radiation field, post-mastectomy and radiotherapy, can resolve radiation dermatitis, as revealed in a phase I study [122].

#### **12. Concluding Remarks**

Recent breakthroughs in novel single-cell profiling and spatial transcriptomics have leveraged our understanding to a new level and helped us to find new answers to a critical question of how cancers move through space and time. Importantly, with rapidly increasing sensitivity of detection methods, we also require novel approaches to conceptually analyze single-cell data with observations at the tissue and organ level.

We have developed a near to complete understanding of VEGF/VEGFR signaling pathways. Studies have shown that relative abundance of the cell surface expression of various VEGFRs and their affiliations for specific VEGF ligands play a fundamental role in the initial set of dimeric constellations. Deeper knowledge of this multifaceted signaling web is key to result-oriented therapeutic targeting. Likewise, EGCG mediated targeting of Wnt/β-catenin has been explored and it needs to be tested comprehensively in different types of cancers. Henceforth xenografted mice bearing β-catenin-overexpressing cancer cells will be helpful in uncovering the true potential of EGCG. Likewise, there is a need to unveil if EGCG inhibited β-catenin activation by functionalization of negative regulators of Wnt signaling. Accordingly, TGF/SMAD signaling regulation by EGCG needs to be addressed more conceptually. Inhibition of SMAD phosphorylation by EGCG is a single

dimension of this highly intricate mechanism. Available evidence enlightens involvement of SMURFs and NEDDs in inhibition of TGF/SMAD signaling. Therefore, additional key players of TGF/SMAD signaling also need in-depth research. Regulation of Notch signaling by EGCG seems to be sparsely studied. Therefore, we still have incomplete information about targeting of proteolytically cleaved segment of Notch-intracellular domain (NICD) in regulation of the target gene network. Does EGCG inhibit NICD nuclear accumulation or whether it also interferes with repressor/co-repressor and activator/co-activator machinery needs more answers. On a similar note, SHH/Gli pathway regulation by EGCG requires initial cellular studies. Furthermore, Gli-overexpressing cancers have to be treated with EGCG and combinatorial treatments.

Despite the absence of clinical trials, the NPs loaded with EGCG might be an efficient and safe strategy for the treatment of several cancers, especially breast and prostate cancer. Thus, clinical trials should be conducted to establish the clinical potential of the NPs loaded with EGCG alone or in addition with the conventional anti-cancer drugs.

**Author Contributions:** Conceptualization, A.A.F. and M.P.; Writing Original Draft Preparation, A.A.F., M.P., A.G., F.F., S.R., R.A., U.Y.S., B.X., A.A.; Writing Review & Editing, A.A.F., M.P., A.G., F.F., S.R., R.A., U.Y.S., B.X., A.A.; Visualization, A.A.F., M.P., A.G., F.F., S.R., R.A., U.Y.S., B.X., A.A.; Supervision, A.A.F. and M.P.; Funding and Acquisition. M.P. and S.R. All authors have read and agreed to the published version of the manuscript.

**Acknowledgments:** This work received financial support from the European Union (FEDER funds through COMPETE POCI-01-0145-FEDER-30624) and National Funds (FCT, Fundação para a Ciência e Tecnologia) through project PTDC/BTM-MAT/30624/2017. Marina Pinheiro thanks FCT for funding through program DL 57/2016—Norma transitória. Andreia Granja thanks FCT and POPH (Programa Operacional Potencial Humano) for the PhD grant (SFRH/BD/130147/2017). This work was also supported by FCT through the FCT PhD Programs, specifically by the BiotechHealth Program (Doctoral Program on Cellular and Molecular Biotechnology Applied to Health Sciences).

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

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
