M1 Macrophages Promote TRAIL Expression in Adipose Tissue-Derived Stem Cells, Which Suppresses Colitis-Associated Colon Cancer by Increasing Apoptosis of CD133+ Cancer Stem Cells and Decreasing M2 Macrophage Population

We have previously reported that adipose tissue-derived stem cells (ASCs) cultured at high cell density can induce cancer cell death through the expression of type I interferons and tumor necrosis factor (TNF)-related apoptosis-inducing ligands (TRAIL). Here, we investigated whether TRAIL-expressing ASCs induced by M1 macrophages can alleviate colitis-associated cancer in an azoxymethane (AOM)/dextran sodium sulfate (DSS) animal model. M1 macrophages significantly increased the TRAIL expression in ASCs, which induced the apoptosis of LoVo cells in a TRAIL-dependent manner. However, CD133knockout LoVo cells, generated using the CRISPR-Cas9 gene-editing system, were resistant to TRAIL. In the AOM/DSS-induced colitis-associated cancer model, the intraperitoneal transplantation of TRAIL-expressing ASCs significantly suppressed colon cancer development. Moreover, immunohistochemical staining revealed a low CD133 expression in tumors from the AOM/DSS + ASCs group when compared with tumors from the untreated group. Additionally, the ASC treatment selectively reduced the number of M2 macrophages in tumoral (45.7 ± 4.2) and non-tumoral mucosa (30.3 ± 1.5) in AOM/DSS + ASCs-treated animals relative to those in the untreated group (tumor 71.7 ± 11.2, non-tumor 94.3 ± 12.5; p < 0.001). Thus, TRAIL-expressing ASCs are promising agents for anti-tumor therapy, particularly to alleviate colon cancer by inducing the apoptosis of CD133+ cancer stem cells and decreasing the M2 macrophage population.


Introduction
Death receptors (DR) targeted by the tumor necrosis factor α-related apoptosis-inducing ligand (TRAIL) are expressed only in tumor cells and not in normal cells [1][2][3]. TRAIL can regulate CD133 + cancer stem cells (CSC) and induce tumor cell-specific apoptosis [4,5]. CD133 is known as a potential than the ones detected for the ASC control group ( Figure 1A). Taken together, in macrophages and ASCs co-cultures, TRAIL was expressed by both cells. Still, the TRAIL expression in ASC was about 3.3 times higher than in macrophages, suggesting that ASCs are the major TRAIL source. In addition, the expression of TRAIL protein in cell lysate and conditioned medium (CM) was increased by 5.36 and 2.71 times in ASCs co-cultured with M1 macrophages and high-density cultured ASCs, respectively ( Figure 1B). Moreover, the concentrations of the secreted TRAIL in ASCs cultured at a high density and co-cultured with M1 macrophages were 135.37 ± 12.76 and 475.22 ± 18.55 pg/mL, respectively ( Figure 1C). These results suggest that M1 macrophages significantly increased the expression of TRAIL in ASCs.
Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 3 of 15 greater than the ones detected for the ASC control group ( Figure 1A). Taken together, in macrophages and ASCs co-cultures, TRAIL was expressed by both cells. Still, the TRAIL expression in ASC was about 3.3 times higher than in macrophages, suggesting that ASCs are the major TRAIL source. In addition, the expression of TRAIL protein in cell lysate and conditioned medium (CM) was increased by 5.36 and 2.71 times in ASCs co-cultured with M1 macrophages and high-density cultured ASCs, respectively ( Figure 1B). Moreover, the concentrations of the secreted TRAIL in ASCs cultured at a high density and co-cultured with M1 macrophages were 135.37 ± 12.76 and 475.22 ± 18.55 pg/mL, respectively ( Figure 1C). These results suggest that M1 macrophages significantly increased the expression of TRAIL in ASCs.

Figure 1.
Enhanced expression of tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) in adipose tissue-derived stem cells (ASCs) co-cultured with M1 macrophages. ASCs were cultured at high-density or co-cultured with M1 macrophages (THP-1) for 2 days and harvested to analyze the TRAIL mRNA and protein, and conditioned media were stored after centrifugation to detect the secreted TRAIL (sTRAIL). (A) Relative expression of TRAIL in ASCs cultured at a high density or with M1 macrophages. The relative expression of TRAIL was analyzed by a nextgeneration sequencing assay, and the relative values are presented in Fragments Per Kilobase Million (FPKM). ** p < 0.01. (B) Expression of TRAIL in cell lysates and conditioned medium (CM). TRAIL was detected in cell lysates and CM using immunoblotting. β-actin was used for normalization. (C) Expression of sTRAIL. * p < 0.05.

Toxicity of TRAIL and CM in LoVo Cells
To evaluate the TRAIL-dependent toxicity in colon cancer cells (LoVo), we used CM obtained from high-density cultures of ASCs. TRAIL (100 ng/mL) reduced the viability of LoVo cells in a timedependent manner (Figure 2A). Similarly, CM increased the LoVo toxicity by about 13.17% ( Figure  2B). Moreover, when the TRAIL in CM was neutralized with an anti-TRAIL antibody, the toxicity of CM in the LoVo cells was reduced ( Figure 2C). These results suggest that the toxicity in LoVo cells by CM was TRAIL-dependent. In addition, in LoVo cells co-cultured indirectly with ASCs, the Annexin +/7-Aminodactinomycin (7-AAD) + population (apoptotic cells) was 44.2 ± 7.8%, indicating Figure 1. Enhanced expression of tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) in adipose tissue-derived stem cells (ASCs) co-cultured with M1 macrophages. ASCs were cultured at high-density or co-cultured with M1 macrophages (THP-1) for 2 days and harvested to analyze the TRAIL mRNA and protein, and conditioned media were stored after centrifugation to detect the secreted TRAIL (sTRAIL). (A) Relative expression of TRAIL in ASCs cultured at a high density or with M1 macrophages. The relative expression of TRAIL was analyzed by a next-generation sequencing assay, and the relative values are presented in Fragments Per Kilobase Million (FPKM). ** p < 0.01. (B) Expression of TRAIL in cell lysates and conditioned medium (CM). TRAIL was detected in cell lysates and CM using immunoblotting. β-actin was used for normalization. (C) Expression of sTRAIL. * p < 0.05.

Toxicity of TRAIL and CM in LoVo Cells
To evaluate the TRAIL-dependent toxicity in colon cancer cells (LoVo), we used CM obtained from high-density cultures of ASCs. TRAIL (100 ng/mL) reduced the viability of LoVo cells in a time-dependent manner (Figure 2A). Similarly, CM increased the LoVo toxicity by about 13.17% ( Figure 2B). Moreover, when the TRAIL in CM was neutralized with an anti-TRAIL antibody, the toxicity of CM in the LoVo cells was reduced ( Figure 2C). These results suggest that the toxicity in LoVo cells by CM was TRAIL-dependent. In addition, in LoVo cells co-cultured indirectly with ASCs, the Annexin +/7-Aminodactinomycin (7-AAD) + population (apoptotic cells) was 44.2 ± 7.8%, indicating that apoptosis was increased by 1.7 times as compared to that in the control group (26.0 ± 4.2%) ( Figure 2D). that apoptosis was increased by 1.7 times as compared to that in the control group (26.0 ± 4.2%) ( Figure 2D). LoVo cells were treated with CM for 24 h and toxicity was detected using a WST-1 reagent. * p < 0.05. (C) TRAIL-dependent toxicity of CM in LoVo cells. To detect whether the toxicity of CM in LoVo cells was TRAIL-dependent, the TRAIL protein in CM was neutralized with anti-TRAIL antibodies and the cytotoxicity was evaluated using a WST-1 reagent. * p < 0.05. (D) ASC-induced apoptosis of LoVo cells. To investigate the direct effects of ASCs on the apoptosis of LoVo cells, the LoVo cells were co-cultured with ASCs and then analyzed by Annexin-V/7-AAD staining and flow cytometry.

TRAIL Resistance of LoVo-CD133 KO Cells
We investigated the TRAIL sensitivity of CD133 + CSCs by comparing the TRAIL effects on CD133 knockout LoVo cells (LoVo-CD133 KO). We generated the LoVo-CD133 KO cells ( Figure 3A) using the CRISPR-Cas9 gene-editing system, as in our previous report [33]. Interestingly, the TRAILinduced apoptosis was dramatically increased in the CD133 + LoVo cells, while there was no significant difference in the apoptosis of LoVo-CD133 KO cells ( Figure 3B,C). In addition, the CM obtained from ASCs cultured at high density increased the apoptosis of LoVo cells, while no apparent effect was detected for LoVo-CD133 KO cells ( Figure 3C).

TRAIL Resistance of LoVo-CD133 KO Cells
We investigated the TRAIL sensitivity of CD133 + CSCs by comparing the TRAIL effects on CD133 knockout LoVo cells (LoVo-CD133 KO). We generated the LoVo-CD133 KO cells ( Figure 3A) using the CRISPR-Cas9 gene-editing system, as in our previous report [33]. Interestingly, the TRAIL-induced apoptosis was dramatically increased in the CD133 + LoVo cells, while there was no significant difference in the apoptosis of LoVo-CD133 KO cells ( Figure 3B,C). In addition, the CM obtained from ASCs cultured at high density increased the apoptosis of LoVo cells, while no apparent effect was detected for LoVo-CD133 KO cells ( Figure 3C).

Suppression of Colon Cancer Development and CD133 Expression by TRAIL-Expressing ASCs
Next, we investigated whether TRAIL-expressing ASCs could alleviate cancer development in AOM/DSS-induced colon cancer model. AOM was injected once into the intraperitoneal cavity on d 0, DSS was administered three times a week with drinking water, and TRAIL-expressing ASCs were transplanted i.p. 1 d before the DSS administration ( Figure 4A). After 56 d, the colon length was found to be reduced in the mice from the AOM/DSS group ( Figure 4B,D), and an average of 12.8 colon tumors per mouse was observed ( Figure 4C). Similarly, mice from the AOM/DSS + ASCs group had a reduced colon length ( Figure 4B,D); however, the average number of tumors was 1.4 per mouse, which was significantly less than that of mice from the AOM/DSS group ( Figure 4C). Histologically, all the tumors induced by AOM/DSS were well to moderately differentiated adenocarcinomas ( Figure 5) and were variable in size, but AOM/DSS + ASC-treated mice exhibited tumors less than 2 mm in size. The non-tumor colonic mucosa showed multifocal scar alterations in the AOM/DSS and AOM/DSS + ASCs groups, but there was no significant active inflammation in any group.

Suppression of Colon Cancer Development and CD133 Expression by TRAIL-Expressing ASCs
Next, we investigated whether TRAIL-expressing ASCs could alleviate cancer development in AOM/DSS-induced colon cancer model. AOM was injected once into the intraperitoneal cavity on d 0, DSS was administered three times a week with drinking water, and TRAIL-expressing ASCs were transplanted i.p. 1 d before the DSS administration ( Figure 4A). After 56 d, the colon length was found to be reduced in the mice from the AOM/DSS group ( Figure 4B,D), and an average of 12.8 colon tumors per mouse was observed ( Figure 4C). Similarly, mice from the AOM/DSS + ASCs group had a reduced colon length ( Figure 4B,D); however, the average number of tumors was 1.4 per mouse, which was significantly less than that of mice from the AOM/DSS group ( Figure 4C). Histologically, all the tumors induced by AOM/DSS were well to moderately differentiated adenocarcinomas ( Figure 5) and were variable in size, but AOM/DSS + ASC-treated mice exhibited tumors less than 2 mm in size. The non-tumor colonic mucosa showed multifocal scar alterations in the AOM/DSS and AOM/DSS + ASCs groups, but there was no significant active inflammation in any group. Immunohistochemical (IHC) staining revealed CD133 expression in the luminal side of most tumor glands in AOM/DSS-induced colon cancer ( Figure 5H) but not in normal mucosa ( Figure 5G). In contrast, the AOM/DSS + ASCs group revealed focal CD133 expression in few tumor glands ( Figure 5I). Immunohistochemical (IHC) staining revealed CD133 expression in the luminal side of most tumor glands in AOM/DSS-induced colon cancer ( Figure 5H) but not in normal mucosa ( Figure 5G). In contrast, the AOM/DSS + ASCs group revealed focal CD133 expression in few tumor glands ( Figure 5I).

Decreasing M2 Macrophage Population Using TRAIL-Expressing ASCs
IHC staining for CD163, a specific marker for M2 macrophages, was diffusely scattered in the lamina propria of mucosa in all groups, including the control ( Figure 6A-C). In the AOM/DSS group, the number of CD163-positive M2 macrophages in the tumoral area was significantly increased (71.7 ± 11.2) in contrast to that of the control group (12.7 ± 2.5; p < 0.001). The AOM/DSS + ASCs group had a significantly lower number of CD163-positive M2 macrophages (45.7 ± 4.2) than that of the AOM/DSS group ( Figure 6G,H; p <0.001). Moreover, non-tumoral mucosa showed similar findings to the tumoral area (AOM/DSS (94.3 ± 12.5) vs. AOM/DSS + ASCs (30.3 ± 1.5), p < 0.001). For F4/80 immunostaining, there was a higher number of F4/80-positive cells than CD163-positive macrophages in the control samples ( Figure 6D,G; CD163, 12.7 ± 2.1 vs. F4/80, 35 ± 5). Further, mucosa from the AOM/DSS and AOM/DSS + ASCs groups showed intratumoral, peritumoral, and non-tumoral mucosa-infiltrating macrophages ( Figure 6E,F). The ASC treatment also lowered the number of F4/80-positive macrophages to 92.3 ± 5.9 in tumor and 103.3 ± 6.1 in non-tumoral mucosa compared with 105.0 ± 8.2 in tumor and 103.3 ± 6.1 in non-tumoral mucosa following the AOM/DSS treatment, but the difference was not statistically significant (p > 0.05). These results suggest that TRAIL-expressing ASCs can suppress tumor development by selectively reducing the number of M2 macrophages during colitis-associated colon cancer development.

Discussion
In this study, we observed that TRAIL expression in ASCs was significantly increased by M1 macrophages, and TRAIL-expressing ASCs could effectively alleviate colon cancer by suppressing CD133 + CSCs and M2 macrophages in the tumor microenvironment. TRAIL can selectively kill only cancer cells; therefore, studies using TRAIL to treat cancer have been actively conducted [4,5]. Based on the homing characteristic of MSCs into the site of inflammation or tumor, MSCs are used to treat cancers as drug and therapeutic gene carriers [5,21,34,35]. In particular, MSCs engineered with the TRAIL gene have demonstrated therapeutic effects in several tumor models [5,21,35]. We reported for the first time that ASCs co-cultured with M1 macrophages secreted high levels of TRAIL without gene manipulation. However, in this study the CM from ASCs cultured at a high density were used to evaluate the effect of TRAIL, excluding the anti-tumor activity of M1 macrophages. Although the cytotoxicity of CM in LoVo cells was about 13%, the ASC transplantation in the AOM/DSS tumor model reduced tumor development by about 90%. The difference in anti-cancer efficiency in vitro and in vivo could be explained by the presence or absence of M1 macrophages and/or expression level of TRAIL. That is, in vivo M1 macrophages might increase the TRAIL expression in ASCs and subsequently suppress the development and progression of colon cancer.
Furthermore, AOM/DSS can initiate and promote relatively strong and reproducible colitisassociated cancer through repeated cycles of colitis caused by DSS following DNA damage by AOM

Discussion
In this study, we observed that TRAIL expression in ASCs was significantly increased by M1 macrophages, and TRAIL-expressing ASCs could effectively alleviate colon cancer by suppressing CD133 + CSCs and M2 macrophages in the tumor microenvironment. TRAIL can selectively kill only cancer cells; therefore, studies using TRAIL to treat cancer have been actively conducted [4,5]. Based on the homing characteristic of MSCs into the site of inflammation or tumor, MSCs are used to treat cancers as drug and therapeutic gene carriers [5,21,34,35]. In particular, MSCs engineered with the TRAIL gene have demonstrated therapeutic effects in several tumor models [5,21,35]. We reported for the first time that ASCs co-cultured with M1 macrophages secreted high levels of TRAIL without gene manipulation. However, in this study the CM from ASCs cultured at a high density were used to evaluate the effect of TRAIL, excluding the anti-tumor activity of M1 macrophages. Although the cytotoxicity of CM in LoVo cells was about 13%, the ASC transplantation in the AOM/DSS tumor model reduced tumor development by about 90%. The difference in anti-cancer efficiency in vitro and in vivo could be explained by the presence or absence of M1 macrophages and/or expression level of TRAIL. That is, in vivo M1 macrophages might increase the TRAIL expression in ASCs and subsequently suppress the development and progression of colon cancer. Furthermore, AOM/DSS can initiate and promote relatively strong and reproducible colitis-associated cancer through repeated cycles of colitis caused by DSS following DNA damage by AOM [36][37][38][39][40]. DSS-induced colitis induces the massive infiltration of T and B lymphocytes, neutrophils, granulocytes, and macrophages, and they express various pro-inflammatory cytokines, including tumor necrosis factor (TNF)-α; interleukin (IL)-6, IL-8, IL-12, and IL-17; and IFN-γ [41,42]. IFN-γ plays an important role in the initiation of DSS-induced colitis by increasing the expression of three chemokines: monokines induced by IFN-γ (MIG), IFN-inducible protein 10 (IP-10), and monocyte chemoattractant protein-1 (MCP-1) [43]. In addition, IFN-γ activates M1 macrophages, which then produce TNF-α and reactive oxygen species (ROS) and respond to the pathogenesis of colitis [44,45]. M1 macrophages are potent scavengers of invading pathogens which activate the innate immune system and then induce the activity of the acquired immune system [46]. However, in the AOM/DSS tumor model, M2 macrophages have been shown to initiate, promote, and induce migration of the colon cancer [29]. In other words, M1-to-M2 polarization occurs during colon cancer development and metastasis, and M2 macrophages exhibit a pro-tumor role [47]. In our previous report, in the DSS colitis model the M1 macrophage population increased significantly, while ASC treatment decreased the M1 population but did not increase the M2 population [45]. In this study, M2 macrophages were increased in the tumor and non-tumoral mucosa of mice from the AOM/DSS group and remarkably reduced in the AOM/DSS + ASCs group. M2 macrophages are known to have a higher sensitivity to TRAIL than M1 macrophages do [32]. Therefore, in the AOM/DSS model, TRAIL-expressing ASCs exposed to M1 macrophages were expected to lower the colon cancer development by controlling the pro-tumor M2 macrophage and CD133 + CSC populations.
Another important characteristic of MSCs is to modulate the activity of various immune cells. MSCs express various kinds of immunomodulators-including indoleamine 2,3-dioxygenase, prostaglandin E2, TNF-α-stimulated protein/gene 6, NO, IL-6, IL-10, and HLA-G-to inhibit the activity of T-and B-cells, macrophages, and natural killer cells [48,49]. Chronic inflammation increases the risk of cancer development, promotes tumor progression, and increases metastasis [50][51][52]. In the early phase of tumor development, several cytokines and ROS secreted by tumor-infiltrating immune cells induce epigenetic changes in premalignant lesions and inhibit tumor suppressor genes [53]. Moreover, during tumor promotion, immune cells produce several cytokines and chemokines essential for tumor cell survival and proliferation, resulting in tumor progression and metastasis [54]. Therefore, modulating the activities of immune cells through MSCs may be a mechanism for reducing tumor development and progression. However, according to our previous results, the tumor-suppressive effect of ASCs cultured at a high density in the xenograft tumor model using athymic nude mice was not significant [25]. Moreover, although it has been reported that the function of macrophages and DCs is maintained in xenograft tumor models using athymic nude mice [55], it remains controversial whether the distribution and function of macrophages and DCs in the tumor microenvironment is normal. However, the current study demonstrated that in the colitis-associated cancer model, ASCs modulated the macrophage population during colon cancer development.
In summary, TRAIL-expressing ASCs were able to alleviate the development and progression of colon cancer by inducing the apoptosis of CD133 + CSCs and reducing the number of M2 macrophages. To understand the therapeutic mechanisms of MSCs in an in vivo system, future studies analyzing differences in immune cell populations in the tumor microenvironment after MSC transplantation are warranted. Moreover, to optimize the effectiveness of the treatment, it is necessary to determine at which stage of the tumor development MSCs must be transplanted. These findings would help advance the treatment of colon cancer using MSCs.

Cell Culture
This study was approved by the Institutional Review Board of the Yonsei University Wonju College of Medicine (IRB n • 2011-58, approval date: 22 December 2013) and the scientific use of human adipose tissues was permitted by written informed consent from three healthy donors (24-38 years of age). ASCs were isolated using a modified protocol [24,56] and sub-cultured with low-glucose Dulbecco's modified Eagle's medium (DMEM; Gibco, Rockville, MD, USA) supplemented with 10% fetal bovine serum (FBS, Gibco) and penicillin/streptomycin (Gibco). For the experiments, ASCs were seeded at 4 × 10 4 cells/cm 2 . After the 3-day culture period, the CM was collected, filtered with syringe-driven filters (0.45 µM), and stored at −80 • C until further use.
The LoVo cells were maintained in Roswell Park Memorial Institute-1640 medium (RPMI-1640) (Hyclone, Logan, UT, USA) supplemented with 10% FBS (Gibco) and penicillin/streptomycin (Gibco). The CD133 knockout LoVo cells (LoVo-CD133 KO) were generated using the CRISPR-Cas9 gene editing system, as described in our previous report [34]. For the indirect co-culture of ASCs and LoVo cells, a transwell plate (Costar, Kennebunk, ME, USA) was used. The ASCs were cultured for three days in the upper chamber without LoVo cells in the lower chamber, the LoVo cells were cultured in the lower chamber for one day without ASCs in the upper chamber, and then the upper and lower chambers were assembled to initiate the co-culture.

Next-Generation Sequencing (NGS)
The total RNA was extracted from 1 × 10 5 cells using TRIzol Reagent (Gibco BRL) according to the manufacturer's instructions. The libraries were prepared for 150 bp paired-end sequencing using TruSeq Stranded mRNA Sample Prep Kit (Illumina, San Diego, CA, USA). The mRNA samples were purified and fragmented from 1 µg of total RNA using oligo (dT) magnetic beads, and the fragmented mRNAs were synthesized as single-stranded cDNAs through random hexamer priming. These were used as templates for second-strand synthesis and the preparation of double-stranded cDNA. After sequential end repairing, A-tailing, and adapter ligation, cDNA libraries were amplified by polymerase chain reaction (PCR). The quality of these cDNA libraries was evaluated with the Agilent 2100 BioAnalyzer (Agilent, Santa Clara, CA, USA), and the cDNA libraries were quantified using the KAPA library quantification kit (Kapa Biosystems, Wilmington, MA, USA) according to the manufacturer's protocol. Following the cluster amplification of denatured templates, sequencing was performed of the paired-end reads (2 × 150 bp) using Illumina NovaSeq6000 (Illumina, San Diego, CA, USA).

Enzyme-Linked Immunosorbent Assay (ELISA)
The CM was recovered from ASCs cultured at a high density for 3 days, and the concentration of secreted TRAIL (sTRAIL) was measured by the human TRAIL Quantikine ELISA kit (R&D Systems) according to the manufacturer's instructions.

Cytotoxicity Assay
The LoVo cells were seeded at a density of 1 × 10 4 cells/cm 2 in 96-well plates and cultured for 24 h. The cells were treated with human recombinant TRAIL (R&D systems), incubated for 24 h, treated with 10 µL of water soluble tetrazolium salt (WST)-1 reagent (Roche, Indianapolis, IN, USA), and incubated for 1 h at standard culture conditions. Then, the absorbance was measured using a microplate reader (Molecular Devices, San Jose CA, USA) at 450 nm.

Apoptosis Assay
The PE-Annexin-V apoptosis detection kit I (BD Biosciences, San Diego, CA, USA) was used according to the manufacturer's instructions. The cells were harvested, washed twice with cold PBS, and re-suspended in a binding buffer. The cells were stained with PE-Annexin-V and 7-aminoactinomycin D (7-AAD) for 15 min at room temperature in the dark and then analyzed without washing using a flow cytometer (FACSAria III, BD Biosciences) within 1 h after the staining.

Animal Study
Six-week-old Balb/c mice (male, 18-22 g) were purchased from Orient Bio, Inc. (Seongnam, Korea). The mice were maintained in a 12-h light/12-h-dark cycle at 23 • C. All the animal care and experiments were conducted in accordance with the Guide for Animal Experiments published by the Korea Academy of Medical Sciences and approved by the Institutional Animal Care and Use Committee of Yonsei University, Wonju College of Medicine (YWC-180117-1, approval date: 23 September 2018). The mice were divided into the following 3 groups (n = 5 per group): AOM/DSS/ASCs-untreated control, AOM/DSS, and AOM/DSS + ASCs. To prepare the AOM/DSS model of colon cancer [35], the AOM (10 mg/kg, Sigma) was injected intraperitoneally (i.p.) and 7 d later mice were administered drinking water with 1.5% DSS (MP Biomedicals, Santa Ana, CA, USA) for 7 d, followed by normal water without DSS for 14 d. This DSS administration cycle was repeated twice and ASCs (1 × 10 6 cells/mouse) were injected i.p. 3 times, on day 6, 27, and 48. The mice were sacrificed on day 56 to analyze the colon tumor development.

Immunohistochemical (IHC) Staining
IHC staining of the paraffin-embedded tissue sections was performed as described in our previous report [57]. The slides were incubated with the monoclonal antibodies against CD133 (Miltenyi Biotec) for CSCs, CD163 (Abcam) for M2 macrophages, and F4/80 (cell signaling technology) for pan macrophages for 2 h at 37 • C in an autostainer using an Ultra View Universal DAB Detection Kit (Benchmark XT, Ventana Medical Systems, Tucson, AZ, USA). Then, the slides were analyzed using an Olympus BX51 microscope (Olympus, Tokyo, Japan), and the positive staining of CD163 and F4/80 was scored based on the number of positive cells/high power field (HPF, ×400) for comparison.

Statistical Analysis
Data are presented as the mean ± standard error of the mean. To compare the group means, Student's t-test and one-way analysis of variance were used, followed by Scheffe's test. A p-value of <0.05 was considered statistically significant.

Conclusions
M1 macrophages significantly increased the TRAIL expression in ASCs and subsequently induced the apoptosis of colon cancer LoVo cells. This TRAIL-dependent response alleviated the development and progression of colon cancer, particularly by inducing the apoptosis of CD133 + CSCs and reducing the M2 macrophage population. However, CD133 knockout LoVo cells were resistant to TRAIL. These results suggest that CD133 + CSCs are potential targets of the TRAIL-expressing ASCs for treating colitis-associated colon cancer. In conclusion, for chronic inflammatory diseases, including colitis ASCs can be used as therapeutic agents to slow the onset and progression of tumors and relieve inflammation through TRAIL expression.

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