Pentagalloyl glucose inhibits TNF‐α‐activated CXCL1/GRO-α expression and induces apoptosis‐related genes in triple-negative breast cancer cells

In triple-negative breast cancer (TNBC), the tumor microenvironment is associated with increased proliferation, suppressing apoptotic mechanisms, an altered immune response, and drug resistance. The current investigation was designed to examine the natural compound pentagalloyl glucose (PGG) effects on TNF-α activated TNBC cell lines, MDA-MB-231 and MDA-MB-468. The results obtained showed that PGG reduced the expression of the cytokine GRO-α/CXCL1. PGG also inhibited IƙBKE and MAPK1 genes and the protein expression of IƙBKE and MAPK, indicating that GRO-α downregulation is possibly through NFƙB and MAPK signaling pathway. PGG also inhibited cell proliferation in both cell lines. Moreover, PGG induced apoptosis, modulating caspases, and TNF superfamily receptor genes. It also augmented mRNA of receptors DR4 and DR5 expression, which binds to TNF-related apoptosis-induced ligand, a potent and specific stimulator of apoptosis in tumors. Remarkably, PGG induced a 154-fold increase in TNF expression in MDA-MB-468 compared to a 14.6-fold increase in MDA-MB-231 cells. These findings indicate PGG anti-cancer ability in inhibiting tumor cell proliferation and GRO-α release and inducing apoptosis by increasing TNF and TNF family receptors' expression. Thus, PGG use may be recommended as an adjunct therapy for TNBC to increase chemotherapy effectiveness and prevent cancer progression.

In the US, breast cancer (BC) is the second leading cause of mortality among women. In breast cancer treatment, one of the most significant challenges in treating this disease is chemotherapeutic resistance. However, it is well-known that chemotherapeutic resistance in BC is associated with abnormal growth factor signaling and aberrant hormonal response. The exact mechanisms of this phenomenon in cancer treatment are still unknown 1,2 . Current insights in molecular targets lead to the recognition of BC therapy in the subset HER2/neu. Regardless of new advances, no treatments for triple-negative breast cancers (TNBC) have been identified. TNBC patients, which present a lack of estrogen receptor (ER), progesterone receptor (PR), and basal level of the HER2/neu oncogene, have an aggressive phenotype, poor prognosis, with an increased chance of metastasis and deficiency of specific and therapeutic targets 1,3-6 . The tumor microenvironment (TME), observed in TNBC, is linked to increased cell proliferation, migration, angiogenesis, suppression of apoptotic mechanisms, an altered immune response, and drug resistance. The oncogenic transformations in the TME involve the infiltration of immune cells and the stimulated fibroblasts that produce cytokines, chemokines, and growth factors, stimulating tumor initiation and progression 7,8 . Chemokines and cytokines may affect tumor immunity and cancer progression directly or indirectly, impacting cancer therapy outcomes 9 . Meanwhile, recent investigations linked TME to tumor heterogeneity, apoptosis repression, oncogenic genes overexpression, activation of tumor suppressors, and deregulation of signal transduction pathways [10][11][12][13] . On the other hand, more studies have focused on finding molecules and pathways that can successfully induce apoptosis and tumor regression, impeding cancer cell progression without affecting normal cells 14 .
The current investigation aims to examine the modulating effect of the polyphenolic natural compound pentagalloyl glucose (PGG), found in many medicinal herbs, on cytokine release. In vivo and in vitro investigations indicated the potential use of PGG in the treatment and prevention of many cancers and inflammatory diseases [15][16][17] . Our previous studies verified PGG potential in down-regulating pro-inflammatory cytokines in BV-2 microglial cells activated by LPS/IFNγ. In this work, PGG decreased the expression of MCP5 and MMP-9 cytokines, whose overexpression is involved in chronic inflammation and neurodegeneration. Also, MCP-5 and MMP-9 are known to modulate genes and proteins that participate in NFƙB and MAPK signaling, suggesting a possible molecular mechanism of action for the PGG inhibitory effect 18,19 . NFƙB and MAPK signaling regulate several genes associated with inflammatory processes and cancer, controlling cell proliferation and survival, chronic inflammation, apoptosis, and the conversion to cancer cells 20 .
Considering the cytokines' critical function in cancer cell proliferation, metastasis, and angiogenesis, this study hypothesized that PGG would inhibit cytokines' expression, leading to reduced cell proliferation and induction of apoptosis through the modulation of the expression of apoptotic-associated genes. Since the genetic variability of TNBCs contributes significantly to the tumor's microenvironment, the current study evaluated PGG effects on two distinct genetically different TNBC cell lines, MDA-MB-231 (MM-231) and MDA-MB-468 (MM-468) cells.

PGG inhibits cell viability and cell growth in MM-231 and MM-468 TNBC cells. Cells were treated
with PGG (3.125-200 μM) for 24, 48, and 72 h to examine the PGG effect on cell viability. Figure 1A,B show that PGG acted on both cell lines in a dose and time-dependent manner. After 24 h-treatment, PGG concentrations of 12.5 and 25 μM were more effective in MM-231 cells than MM-468. However, in the concentration of 50 μM and over, PGG cytotoxic effect was higher in MM-468 cells (Fig. 1A), showing that PGG affects both cell lines differently. In both cell lines, concentrations lower than 12.5 μM caused a non-significant effect. The IC 50 of 50.23 ± 2.16 µM for MM-231 and 35.72 ± 0.75 µM for MM-468 indicate that PGG affects the cell lines differently, inducing more significant cytotoxicity in MM-468 cells (Fig. 1A). After 48 and 72 h, the PGG effect was more pronounced compared to the period of 24 h, leading to higher toxicity (Fig. 1B). However, there was higher toxicity in the MM-468 cells in concentrations over 50 μM, following the same pattern observed after the 24 h-period incubation.
To investigate the PGG effect on cell proliferation, TNBC cells were treated with PGG (1.56-200 μM) for 96 h. The results demonstrated that the proliferation was inhibited by PGG in both breast cancer cell lines in a dose-dependent manner (Fig. 1C,D). Interestingly, at lower concentrations of PGG showed to be more potent in MM-231 cells; however, in the concentration of 50 μM, PGG was more effective in MM-468 cells, causing 83% suppression in cell proliferation compared to 67% in MM-231. The same pattern was consistent in concentrations above 50 µM. (Fig. 1C,D).

Discussion
Our study's primary focus is to investigate the downstream effects of PGG on the transcription regulation of genes for the treatment of TNBC by comparing genetically different MM-231 and MM-468 cell lines. These cells were previously described as expressing GRO-α/CXCL1 mRNA and protein in significantly increased levels 21 . In the present investigation, the polyphenol PGG downregulated the expression of the chemokine GRO-α/CXCL1 in MM-231 and MM-468 cells stimulated by TNF-α with a higher impact in MM-468 comparing to MM-231 cells, decreasing levels of expression in GRO-α/CXCL1 mRNA and proteins in these TNBC cells. Reports have described that stromal and immune cells may produce the chemokine CXCL1, which works in a paracrine manner in the tumor environment through cancer development 22,23 . G protein-coupled receptor chemokine (C-X-C motif) receptor 2 (CXCR2) functions as the receptor for CXCL1 24 . The overexpression of CXCL1 has been associated with many malignancy types linked to oncogenesis, metastasis, angiogenesis, and chemoresistance 23,25,26 . In ER-negative cells, higher levels of CXCL1 mRNA and protein expression were detected compared to ER-positive cells. CXCL1 stimulated invasion and migration of BC cells via ERK/MMP2/9 signaling, indicating that tumorderived CXCL1 may be connected with the high capacity of invasion observed in ER-negative BC cells, thus suggesting that CXCL1 may be a possible target in ER-negative BC therapeutics 27 . In BC, CXCL1-knocked down MM-231 cells presented a decrease of 40% in cell proliferation and even a more significant inhibition in migration (43%) and invasion (60%) studies 21 . Corroborating with the literature, in our studies, the inhibitory effect of PGG over GRO-α/CXCL1 expression may also be associated with the decline in the proliferation rate observed in the MM-231 and MM-468 cell lines since CXCL1 is associated with increased cell growth in MM-231 cells 21 . www.nature.com/scientificreports/ To investigate a possible molecular mechanism for PGG inhibitory effect over GRO-α/CXCL1, the levels of expression of genes belonging to NFƙB and MAPK signaling pathways were evaluated after PGG treatment. The results indicate that PGG decreases TNF-α-stimulated GRO-α/CXCL1 release via inhibition of IқBKE and MAPK expression. Among the IқB proteins that participate in NFƙB activation, IқBKE showed to be highly expressed in human breast cancer samples, suggesting its participation in the regulation of breast cancer growth, and identifying IқBKE as a new oncogene in breast cancer 28,29 . By silencing IқBKE expression in breast cancer cells, studies showed that tumor cell death rate was increased, emphasizing the regulatory role of IқBKE in tumor cell proliferation 28 . Furthermore, another signaling pathway that has a significant role in TNBC progression is the MAPK, which is associated with invasion, metastasis, and prognosis of TNBC 30 . The decision to test MAPK1/ ERK2 only in the RT-PCR assays was based on reports that show MAPK3/ERK1 association with a better prognosis and its role in modulating the YAP1 signaling pathway, indicating that MAPK3 is a negative regulator of breast cancer development. However, higher expression of MAPK1 predicted a poor prognosis 31 . The studies in the literature agree with our findings showing that TNF-α stimulates the expression of CXCL1 and supports the fact that PGG inhibitory effect over GRO-α/CXCL1 expression through inhibition of NFƙB and MAPK signaling genes may decrease cell proliferation and slow inflammation and cancer progression.
To further investigate PGG anti-cancer potential, this study also focused on another hallmark of cancer, apoptosis suppression. The investigation of apoptosis-related genes revealed that PGG could stimulate intrinsic and extrinsic apoptotic pathways in MM-231 and MM-468 TNBC cell lines. PGG increased Caspases 3, 4, and 14 expressions in MM-231 cells, but in MM-468, only Caspase 3 was significantly upregulated. Caspase 3 is considered an executioner caspase, and its activation leads to the breakdown of proteins and cytoskeleton cleavage, killing the cell 32   www.nature.com/scientificreports/ both cell lines, higher expression levels detected in MM-231. BNIP3 role in the cell has been linked to cell death and cell viability 33,34 . BIRC3 is a gene that codes for inhibitors of apoptosis proteins (IAP) family of proteins cIAP2, which controls caspases and cell death, regulates signaling during inflammation, and controls mitogenic kinase signaling, cell growth, cell invasion, and metastasis [35][36][37] .
The current results showed variability in PGG ability to induce the two cell lines, anti-and pro-apoptotic genes. In MM-231 cells, PGG induced the expression of BCL2A1, BIRC6, BNIP3L, BLC2L11, BNIP2, RIPK2, and TP53BP2 genes, while BCL2 and BAK1 were increased only in the MM-468 cells. These genes are critical initiators of the intrinsic or extrinsic apoptotic pathways and establish the balance between cell death cell survival. Remarkably, the highest upregulation was seen in tumor necrosis factor (TNF) expression with a 154.6-fold increase in MM-468 cells, compared to 14.6-fold in MM-231. Additionally, PGG induced several TNF receptor superfamily members' gene expressions, with MM-231 cells being more responsive to PGG treatment than MM-468. The MM-231 cells presented a higher death receptor 4 (DR4), death receptor (DR5), and CD137 expression. In MM-468 cells, a 27-fold upregulation in the death receptor 6 (DR6) gene expression was observed, with no significant increase in MM-231 cells. The difference in TNF and TNF receptor expression levels after PGG treatment shows how genetically different TNBC cells may respond differently to a treatment.
Clinical studies with TNF demonstrated that it could be toxic due to its systemic toxicity. However, locoregional drug delivery systems have shown that TNF alone or combined with supplementary pharmacological agents could be a possible approach for tumor therapy by inducing tumor sensitivity to the treatment 38 . The targeted delivery of TNF to the tumor may considerably increase the local concentration in tumor cells, minimizing TNF doses and consequently decreasing the systemic toxicity 38 . Wu et al. 39 demonstrated that TNF could preserve radiotherapy's sensitizing ability and strength cytotoxicity of chemotherapeutics against breast cancer cells in vitro and in vivo. These results indicate that TNF may be a promising candidate for further clinical application 39 . Therefore, our results show that the PGG effect as an apoptotic inducer through TNF increased expression may have a clinical significance in cancer treatment. Our data also suggests that unique approaches may be considered for different TNBC types since they may have different response levels.
Moreover, members of the TNF superfamily, such as TRAIL, can selectively stimulate apoptosis in numerous tumorigenic cells but not in healthy cells 40 . TRAIL binds to DR4 41 and DR5 42,43 , which contain functional cytoplasmic death domains that can activate the extrinsic pathway for apoptosis 44 . TRAIL stimulates cell death in a range of cancerous cells despite p53 status. Thus, it may be a helpful strategy in BC therapy, mainly in cells where the response signaling of p53 is not activated, hence avoiding chemotherapy and radiotherapy resistance 45,46 . However, TRAIL shows limited therapeutic benefits since most primary cancer cells present resistance to TRAIL. This limitation happens because TRAIL induces apoptosis, but it can also activate cell survival mechanisms, including NFқB, MAPKs, and phosphatidylinositol-3-kinases (PI3K/AKT) 47 . In this regard, the present study showed that PGG upregulated DR4 and DR5 mRNA expression in TNBC cells, suggesting that PGG could be combined with TRAIL enhancing the anti-cancer effect of TRAIL. Furthermore, PGG modulatory effect over NFқB signaling, observed in our studies, could also improve apoptosis induced by TRAIL, corroborating with many preclinical models of the tumor where other NFқB inhibitors were used.
The current study indicates PGG potential as an anti-cancer compound, inhibiting the expression of TNFαinduced CXCL1 via inhibition of the expression of genes and proteins involved in the NFқB and MAPK signaling in MM-231 and MM-468 TNBC cell lines. NFқB was shown to be a crucial modulator of numerous cytokines, including CXCL1. By stimulating the breast cancer cells with TNF-α, we induced a well-known cascade that leads to the phosphorylation of NFқB, which translocates to the nucleus and, together with co-factors, promotes the increased transcription of cytokines 48 . Regarding MAPK signaling, it also has an essential role in activating the transcription of cytokines. By stimulating the cells with TNF-α, MAPK pathway and its associated genes are activated. MAPK/ERK signaling, in particular, is a critical regulator of AP-1 transcriptional activity in TNBC cells 49 . Therefore, Fig. 7 describes the mechanisms used by TNF-α to activate NFқB and MAPK signaling (including activation of NFқB and AP-1 transcription factors) and subsequent transcription of cytokines, such as CXCL1; a mechanism already described in the literature. The figure also shows the proposed mechanism of the PGG effect on IқBKE and MAPK expression, which would lead to a decrease in the expression of CXCL1. PGG inhibitory effect over CXCL1 may also be associated with a cell growth decrease observed in both cell lines.

Conclusion
The obtained data show that PGG inhibits TNF-α-stimulated CXCL1 mRNA and protein expression and propose that this inhibitory effect may be caused by repression in the translation of genes that modulate NFқB and MAPK signaling as a possible molecular mechanism. The results also suggest PGG potential as an inducer of apoptosis in MM-231 and MM-468 TNBC cells, stimulating expression of numerous apoptosis-associated genes, including caspases. Remarkably, PGG upregulated TRAIL DR4 and DR5 receptors mRNA expression and highly induced TNF expression in MM-468 cells. Thus, PGG use may be recommended as an adjunct therapy for TNBC to increase chemotherapy effectiveness and prevent cancer progression. Further investigations are still needed to support the use of PGG in the induction of TNF and its receptors, which might be a helpful strategy to increase tumor sensitivity to chemotherapy.

Materials and methods
Cells, chemicals, and reagents. TNBC    Real time-polymerase chain reaction (RT-PCR). The pellet of cells for the RT-PCR reaction was obtained after 24 h of the following treatments: control (cells + DMSO), PGG-treated (6.25 µM for MM-231 and 25 µM for MM-468 cells), TNF-α-stimulated (50 ng/ml) and co-treated with PGG + TNF-α. RNA was extracted using the Trizol reagent. iScript advanced reverse transcriptase was used to synthesize cDNA strands from the mRNA. The advanced reaction mix, reverse transcriptase, sample (1.5 µg/reaction), and water were combined in a volume of 20 µl. The reverse transcription protocol for thermal cycling consisted of two steps: 46 °C for 20 min and 95 °C for 1 min. For the RT-PCR assay, the sample (200 ng cDNA/reaction), master mix, primer, and water were placed together in each well. According to the protocol from Bio-Rad, the program for the thermal cycling consisted of an initial step at 95 °C for 2 min and denaturation at 95 °C for 10 s, followed by 39 cycles of 60 °C for 30 s (annealing/extension), and 65-95 °C for 5 s/step (melting curve) (Bio-Rad CFX96 Real-Time System-Hercules, CA, USA). The UniqueAssay ID for the primers is described as follows: GRO-α: qHsaCED0046130; IқBKE: qHsaCID0014831; MAPK1: qHsaCED0042738.
Capillary electrophoresis Western analysis. Cell pellets corresponded to control (cells + DMSO), PGG-treated (6.25 µM-MM-231 and 25 µM-MM-468 cells), TNF-α-stimulated (50 ng/ml) and co-treated (PGG + TNF-α) cells after 24-h treatment. A lysis buffer containing a protease inhibitor cocktail was added to each pellet, and total protein expression was determined using Western analysis. Samples containing 0.2 mg/ ml of protein were used. Following the manufacture's protocol (ProteinSimple), the microplate was loaded and placed in the instrument. The reaction took place inside the capillary system; using specific antibodies, the proteins were identified, the chemiluminescence reaction was determined, and the digital blot images were taken. The antibodies used in this assay are: Apoptosis detection using annexin V-FITC and apoptosis PCR arrays. Apoptosis Kit (Annexin V-FITC) was used to establish PGG apoptotic effect on MM-231 and MM-468 TNBC cells. The cells were seeded (5 × 10 5 cell/well) and incubated overnight using 6-well plates. The next day cells were treated with PGG in a concentration ranging from 25 to 200 µM (final volume 3 ml/well). Control cells consisted of cells + DMSO (concentration < 0.1%). Following 24 h of incubation, cell pellets were collected and resuspended in 500 µl of the buffer. Following the manufacture's protocol (RayBiotech), 5 µl of Annexin V-FITC and 5 µl propidium iodide was added to the samples. PGG apoptotic effect was determined by flow cytometry (FACSCalibur-Becton Dickinson, San Jose, CA, USA), and data analysis was performed with CellQuest software. For the apoptosis PCR arrays, cell pellets from control (cells + DMSO) and PGG-treated cells (100 µM for MM-231 and 50 µM for MM-468 cells) were lysed with Trizol reagent. iScript advanced reverse transcriptase was used to synthesize cDNA strands from the mRNA, as described previously. For the RT-PCR reaction, the sample (10 ng cDNA/ reaction) and master mix were combined in the plate. The thermal cycling program was the same one used previously.
Data analysis. GraphPad Prism (version 6.07) (San Diego, CA, USA) was used to determine the results' statistical significance. All the experiments were performed at least in triplicates. The IC 50 calculation in cell viability studies was calculated using nonlinear regression. The statistical significance assessment between the treatments was established using a one-way ANOVA analysis and Dunnett's multiple comparison tests. Human Cytokine Array software C1000 (Code: S02-AAH-CYT-1000-RayBiotech) was used to establish cytokine expression on the arrays. The expression of specific genes was determined with CFX 3.1 Manager software (Bio-Rad, Hercules, CA). Total protein expression was determined using ProteinSimple Compass software, as well as the digital blot images.