3,3′-Diindolylmethane Promotes Gastric Cancer Progression via β-TrCP-Mediated NF-κB Activation in Gastric Cancer-Derived MSCs

Gastric cancer is a malignant tumor characterized by high morbidity and invasion. Surgery combined with chemo-radiotherapy is the most common treatment for gastric cancer, while multiple drug resistance always results in treatment failure. Once the anti-tumor drugs enter the tumor foci, tumor cells as well as those found in the microenvironment are affected. However, the effects of drugs on tumor microenvironment (TME) are easily overlooked. In this study, we investigated the effects of the anti-cancer drug 3,3’-diindolylmethane (DIM) on gastric cancer-derived mesenchymal stem cells (GC-MSCs) and their subsequent impact on cancer progression. Surprisingly, we found that the therapeutic concentration of DIM upregulated the expression level of tumor-related factors such as CCL-2, IL-6, and IL-8 in GC-MSCs. The conditioned medium of DIM-treated GC-MSCs promoted the proliferation, invasion, and migration of gastric cancer cells in vitro and tumor growth in vivo. Mechanistically, DIM enhanced the expression of β-TrCP, an E3 ubiquitin ligase leading to IκBα degradation and NF-κB activation in GC-MSCs. The β-TrCP knockdown partially eliminated positive results caused by DIM. Our results showed that the therapeutic dosage of DIM induced cell death in cancer cells, while enhancing MSC paracrine functions in the stroma to offset the original DIM effect on cancer cells. These findings provide a new mechanism of anti-cancer drug resistance and remind us to adjust the chemotherapeutic scheme by combining the anti-cancer drug with an appropriate signaling pathway inhibitor to block the side effects of drug on targeted TME cells.


INTRODUCTION
Gastric cancer is one of the most common malignancies worldwide and constitutes the second highest morbidity and the third leading cause of cancer-related deaths in China (1,2). Surgery and chemo-radiotherapy are the primary treatments for gastric cancer. However, these therapeutic regimens are not satisfactory in prolonging patient survival time, mostly due to insufficient pharmaceutical effects and multidrug resistance (3). Scientists have put considerable effort into determining the mechanisms of drug resistance in gastric cancer. Previous research suggests that drug degradation, anti-apoptosis, immune escape, epithelial-mesenchymal transition (EMT), cell stemness, autophagy, epigenetic modifications, and upregulation of multidrug resistance (MDR)-related genes are all involved in the potential risk of treatment failure of gastric cancer (4). Current research has mainly focused on the intrinsic or acquired drug resistance in gastric cancer cells; however, once drugs enter the tumor foci, both tumor cells and those in the tumor microenvironment (TME) are affected. Nevertheless, little attention has been paid to the influence of chemotherapeutic drugs on the TME and the subsequent feedback effect on the cancer cells.
TME is vital to the growth of cancer cells because it provides the necessary support and nutrition as in the relationship between soil and seeds (5). TME contains matrix cells, vascular endothelial cells, and immune cells in addition to tumor cells, with major characteristics including low oxygen, low pH, and high levels of proteolytic enzymes and cytokines. These particular features and peculiarity provide opportunities for tumor proliferation, invasion, and migration (6). Mesenchymal stem cells (MSCs) exist in TME and facilitate tumor progression by secreting various cytokines, suppressing immune responses, and remodeling the extracellular matrix of tumors. Houghton et al. (7) found that the fusion of MSCs and gastric epithelial cells under Helicobacter pylori infection can induce gastric cancer formation. Thus, alterations in the external environment, such as the treatment of chemotherapeutic drugs on MSCs, may influence its primary impact on tumor progression. We previously isolated and identified MSCs from gastric cancer (GC-MSCs) for the first time (8), and proved that GC-MSCs could prompt gastric cancer metastasis via EMT induction (9) and serve as a potential target for gastric tumor treatment (10). Thus, we hypothesized that alterations in the external environment as the treatment of chemotherapeutic drugs may influence the function of MSCs on gastric cancer progression. In this study, we aimed to investigate the relationship between the anti-cancer drug 3,3-diindolylmethane (DIM), GC-MSCs, and gastric cancer progression.
DIM is a small-molecule compound and a major active metabolite of indole-3-carbinol, which can be extracted from cruciferous vegetables. Many studies have shown that DIM can inhibit proliferation and induce apoptosis in various cancer types (11). Previously, we found that low levels of DIM activated Wnt4 autocrine signaling to enhance the progression of gastric cancer cells (12). Moreover, our research also indicated that DIM could promote the stemness of human umbilical cord-derived mesenchymal stem cells (hucMSCs) by increasing exosome mediated Wnt11 autocrine signaling (13), so that the stemnessenhanced hucMSCs could be used in tissue regeneration. However, the effects of DIM on TME-derived MSCs and their subsequent influence on tumors remains unknown.
In this study, we treated GC-MSCs with the regular dosage of DIM (depending on IC 50 ) and found that GC-MSCs expressed a high level of oncogenic factors such as CCL-2, IL-6, IL-8, and TGF-b. Furthermore, this expression was triggered by the activation of b-TrCP/NF-kB signaling pathway. The conditioned medium of GC-MSCs pretreated with DIM could promote proliferation, invasion, and migration of gastric cancer cells. b-TrCP knockdown eliminated positive results caused by DIM. Collectively, the therapeutic dosage of DIM could induce cancer cell death, while enhancing MSC paracrine functions in the stroma to offset the cell death induction, which provides a new vision on the application of anti-tumor drugs. A chemotherapeutic scheme that combines the use of a signaling pathway inhibitor to block the side effect from drug-targeted TME cells could be considered.

MATERIALS AND METHODS
The study was approved by the Medical Ethics Committee and Ethics Committee for Experimental Animals of Jiangsu University (IRB protocol number: 2020161).

Cell Culture, GC-MSC Isolation and Identification
Human gastric cancer cell lines HGC-27, SGC7901, and MGC-803 were purchased from the Institute of Biochemistry and Cell Biology at the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in high-glucose Dulbecco's modified Eagle medium (DMEM) (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS; Gibco, USA). Cells were cultured at 37°C in humidified air with 5% CO 2 . HucMSCs were isolated as previously described (14) and maintained in low-glucose DMEM (Gibco, Grand Island, NY, USA) containing 10% FBS.
The gastric cancer tissues were obtained from patients with gastric adenocarcinoma in The Affiliated People's Hospital of Jiangsu University, Zhenjiang, China. Fresh, sterile gastric carcinoma tissue specimens were collected during surgery. The specimens were immersed in 95% ethanol for 2-3 sec to avoid contamination, and then washed with PBS and antibiotics several times to remove the blood. The surface of the cancer tissues was removed and the inner parts were cut into 1-to 3-mm 3 -sized pieces and floated in Dulbecco's modified Eagle's medium with low glucose (LG-DMEM) (Gibco, USA), containing 10% fetal bovine serum (FBS, Gibco, USA), penicillin (100 U/ml) and streptomycin (100 lg/ml). The pieces of cancer tissues were subsequently incubated at 37°C in humidified air with 5% CO2. After culturing for 15 days, colonies of fibroblast-like cells appeared. When their confluens reached 80%, the cells were harvested by 0.25% trypsin-1 mM EDTA and re-plated into larger culture flasks at a 1:3 split ratio. These gastric cancer-derived MSC-like cells at passage four were used for subsequent experiments. As for the identification of GC-MSCs, the expression of specific surface antigens CD44 (BD Pharmingen), CD105 (Miltenyi), CD34 (BD Pharmingen), CD45 (BD Pharmingen) of GC-MSCs was detected by flow cytometry, and multi-directional differentiation potential was assessed through osteogenic and adipogenic differentiation assays according to the manufacturer's instructions (Cyagen). Cells were stained with alizarin red and Oil-Red-O (for lipid droplets) on Day 14.

Colony Formation Assay
Gastric cancer cells SGC-7901, MGC80-3 and HGC-27 were treated with CM from GC-MSC mentioned above for 48 h, harvested, and seeded into 35-mm plates (1000 cells/well) overnight under standard conditions for 10 days. The medium containing was changed at 3-day intervals. At the end of the incubation period, the cultures were fixed with 4% paraformaldehyde and stained with crystal violet.

MTT Assay
MTT assay was used to determine the viability of different cells (HGC-27, SGC-7901, hucMSC, and GC-MSC) treated with different concentrations of DIM. Cells were inoculated into 96well plates at a density of 1.5×10 4 cells/well. After 24 h incubation, the cells were treated with DIM at concentrations of 1, 10, 25, 50, 100, 200, 300, and 400 mM DIM or the same volume of DMSO (0 mM). After 24 h incubation, 20 mL of MTT (0.5 mg/mL) was added to each well and cultured for 4 h. The formazan crystals formed were solubilized in DMSO, and the absorbance of each well was read in a microplate reader at a wavelength of 490 nm.

RNA Inference
Specific siRNA against b-TrCP was produced by GenePharma (Suzhou, Jiangsu, China). The sequence of b-TrCP siRNA was 5'-GAGAGAGAAG ACUGUAAUAdTdT-3'. GC-MSCs (1×10 6 cells/well) were grown in six-well plates and transfected with siRNAs using LipoFiter transfection reagent (Hanbio, Shanghai, China) for 6 h. Following 24 h incubation, the transfected cells were harvested in 10 cm plates. The conditioned medium and cells were collected for subsequent experiments.

Transwell Migration Assay
Gastric cancer cells SGC-7901, MGC80-3 and HGC-27 were treated with CM from GC-MSC mentioned above for 48 h, harvested, and seeded into the top chamber, and 10% FBScontaining medium was placed into the bottom chamber. After incubation at 37°C in 5% CO 2 for 12 h, the cells remaining on the upper surface of the membrane were removed with a cotton swab. The cells that migrated through the 8-mm sized pores and adhered to the lower surface of the membrane were fixed with 4% paraformaldehyde, stained with crystal violet, and imaged.

Wound Healing Assay
Gastric cancer cells SGC-7901, MGC80-3 and HGC-27 were pretreated with CM from GC-MSC mentioned above for 48 h, harvested, and seeded at a density of 2×10 5 cells/well in six-well plates and incubated at 37°C in 5% CO 2 for 24 h to create confluent monolayers. The monolayers were scratched with a sterile pipette tip. To measure cell mobility, images from five random fields at 24 h after scratching were obtained. The width of the original scratch was measured using the NIH ImageJ image processing software (http://rsb.info.nih.gov/nih-image/). The migration ratio was calculated as follows: (the width of the original scratch-the width of the actual scratch)/the width of the original scratch×100.

Xenograft Mouse Model
Twenty male BALB/c nu/nu mice (Laboratory Animal Center of Shanghai, Academy of Science, Shanghai, China) aged 4-6 weeks were randomly divided into two groups (five mice/group). MGC80-3 cells were pretreated with different types of conditioned medium of GC-MSCs for 48 h, then 2.5×10 6 cells in 200 ml PBS were implanted subcutaneously into the right flanks of the mice. The mice were fed normally and the tumors were harvested 30 days after the implantation. Tumor size and weight were measured.

Apoptosis Assays
GC cells MGC-7901 were pretreated with different types of conditioned medium of GC-MSCs for 48 h, with PBS used as a control, and then cells were collected and seeded at a density of 2×10 5 cells/well in six-well plates and incubated at 37°C in 5% CO 2 for 24 h, then treated with 50 mM DIM for 48 h. Annexin V/PI staining of these cells was performed to detect apoptosis. In brief, cells were collected and stained with Annexin V/PI for 15 and 30 min, then the multicolor flow cytometry was performed according to standard protocols (FACS Canto II, BD Biosciences). The results were analyzed using CellQuest software.

Statistical Analysis
Data are expressed as mean ± SD. The statistical significance of differences between two groups was determined using two-tailed Student's t-test. The significance of differences among multiple groups was determined using one-way ANOVA. All experiments were performed at least in triplicates (n=3). P < 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism (GraphPad Software, La Jolla, CA).

DIM Increases the Expression of Tumor-Related Factors in GC-MSCs
DIM is a natural compound harvested from cruciferous vegetables, which belongs to the class of indole glucosinolates with the molecular formula C17H14N2 ( Figure 1A). To explore its influence on GC-MSCs, we firstly isolated and identified GC-MSCs by verifying their adipogenesis and osteogenesis capacity in inducing reagent as well as the representative markers ( Figures S1A, B). We used the MTT assay to determine the optimum concentration of DIM required to induce cell death.
We  Figure 1B). To further confirm the effects of DIM on GC-MSCs, we analyzed the colony formation ability in GC-MSCs pretreated with different concentrations of DIM (0, 1, 10, 25, and 50mM DIM) for 48 h. Ten days later, we found that the anticancer dosage of DIM not only had no inhibitory effect on GC-MSCs but promoted cell proliferation, especially at 50 mM ( Figure 1C). Based on these results, we evaluated the expression as well as the secretion level of inflammatory cytokines (CCL-2, IL-6, IL-8, TGF-b) in GC-MSCs pretreated with different concentrations of DIM. The data showed that 50 mM DIM significantly increased the expression of CCL-2, IL-6, and IL-8 ( Figure 1D; Figure S4). However, a higher concentration of 100 mM reversed these results ( Figure S2A). Thus, we hypothesized that GC-MSCs pretreated with DIM may influence the progression of gastric cancer by secreting proinflammatory cytokines.

Conditioned Medium of GC-MSCs Treated With DIM Promotes Gastric Cancer Cell Development
To verify our hypothesis, we collected the conditioned medium

DIM Increases the Expression of NF-kB in GC-MSCs by b-TrCP-Mediated IkBa Inhibition
The activation of NF-kB signaling pathway is regulated by IkBa, which is the substrate of E3 ubiquitin ligase b-TrCP ( Figure 4A). transwell assay. Treatment with DIM (50m M) enhanced the migration of gastric cancer cells, which was weakened in b-TrCP -/-GC-MSCs ( Figure 5D). Furthermore, animal models were used to confirm the proliferation of cancer cells in vivo.  Figure  5E), and their weight was evaluated ( Figure 5F). CM from DIM-treated GC-MSCs significantly increased the tumor weight, and there were no differences between ctrl+b-TRCP -/and DIM+b-TrCP -/mice. These results further confirmed that the dosage of DIM that was lethal for gastric cancer cells could not induce cell death in GC-MSCs from TME. Moreover, it promoted the progression of gastric cancer, which was mediated by increased expression of b-TRCP and NF-kB signaling activation to produce more tumor-related cytokines such as CCL-2, IL-6, and IL-8, contributing to the development of gastric cancer ( Figure 6).

DISCUSSION
The management of gastric cancer is a tough issue in the medical community because of the difficulty in early diagnosis and its multi-drug resistance (15). Mechanisms such as drug degradation, target site modification, expression of efflux pumps were shown to play important roles in chemotherapeutic failure (16). Even though new anti-cancer drugs and chemotherapies have been developed, there has been no significant improvement in the outcome due to drug resistance, which suggests that the mechanisms underlying this process require urgent investigation. Most studies have focused on the intrinsic or acquired resistance in cancer cells. For example, long non-coding RNA such as lncRNA UCA1 inhibits the apoptosis pathway and induces the resistance in gastric cancer cells to adriamycin and 5-fluorouracil functioning as a sponge of miR-27b to downregulate caspase-3 expression (17). Our group has been trying to discover the mechanism of drug resistance in gastric cancer in a different way by examining the relationship between the chemotherapeutic drugs DIM and mesenchymal stem cells derived from TME, while simultaneously studying the subsequent effects on cancer progression. DIM is a natural small-molecule anti-cancer drug that has been shown to inhibit cell proliferation and induce cell apoptosis in prostate and breast cancer cells through the regulation of the Akt/FOXO3a/GSK-3/b-catenin/AR signaling axis (18) and reduction of NF-kB activity (19). However, it has also been shown that anti-cancer dosage of DIM can activate autocrine Wnt4 signaling to promote gastric cancer (12). Our previous  study indicated that DIM can enhance the stemness of hucMSCs through Wnt/b-catenin signaling (13). MSCs exist in many tissues, including TME, supply necessary nutrition for cancer cells, and have multiple effects on tumors. Taro Ikeda et al. (20) found that bone marrow-derived MSCs can promote gastric cancer by secreting CXCL16 to activate STAT3 and mediate the expression of Ror-1. Our group was the first to isolate GC-MSCs from gastric cancer tissues (8). GC-MSCs share the characteristic of bone marrow-derived MSCs and display unique biomarkers of gastric cancer. Furthermore, GC-MSCs can promote gastric cancer progression by inducing the EMT and triggering M2 macrophage polarization (9,10). When studying the effects of DIM on MSCs in clinical applications, we should not only consider the safety of DIM with regard to its the lethal effect on cancer cells, but also its action on TME, especially the TMEderived MSCs. Our study found that 50 mM of DIM (corresponding to IC 50 in gastric cancer cell lines) had almost no lethal effect on GC-MSCs, but could activate the NF-kB signaling pathway and increase the expression of CCL2, IL-6, IL-8, TGF-b. These molecules have been known to promote tumor progression (21). Paracrine signaling is the most important mechanism by which the MSCs exert biological effects on the target cells. The treatment of gastric cancer lines with conditioned medium of GC-MSCs (22) significantly enhanced the proliferation, invasion, and migration of these cells. These effects could be inhibited by the knockdown of b-TrCP in GC-MSCs. Our study proved that although a regular dosage of DIM could induce cell death in gastric cancer cells, it had a simultaneous effect on MSCs in TME and accelerated the development of tumor cells through paracrine signaling by oncogenic factors. These findings provide a novel direction for a rational usage of anti-cancer drugs and help to establish more effective anti-tumor programs such as adjusting the dosage or combining anti-cancer drugs with inhibitors of the associated signaling pathways, such as NF-kB inhibitors.
In summary, we found that the anti-cancer drugs DIM affected both gastric cancer cells and TME-derived GC-MSCs. The effective dose of DIM reversed the upregulation of the expression of CCL-2, IL-6, and IL-8 in GC-MSCs by b-TrCPmediated ikBa degradation and NFkB signaling pathway activation, thereby enhancing tumor progression ( Figure 6). This study results provide a different angle of view on the application of anti-tumor drugs, and imply that a combination of signaling pathway inhibitors with antineoplastic agents may achieve a better curative effect.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.