Differential effects of peptidoglycan on colorectal tumors and intestinal tissue post-pelvic radiotherapy

Immediate medical intervention is required after pelvic tumor radiotherapy to protect the radiosensitive intestine and also to mitigate tumor growth. Toll-like receptors (TLRs) have been shown to promote tissue repair processes. Here, we analyzed the effect observed upon combining the TLR2 agonist, peptidoglycan (PGN), with radiation therapy on tumors as well as intestinal tissue, both in vitro and in vivo. In contrast to radiotherapy alone, PGN when combined with ionizing radiation (IR) elicited enhanced antitumor effects and also reduced the IR-induced intestinal damage. Mechanistic studies showed that PGN first induced an IL13 response in the irradiated intestine, but was decreased in tumor cell models screened by Th1/Th2 FlowCytomix assay and validated by the application of IL13 and anti-IL13 neutralizing antibodies. Next, PGN stimulated Akt3, but not Akt1/2, as was verified by AKT1/2/3 plasmid transfection assay and in AKT1/2/3 knockout mice in vivo. Akt3 expression was inhibited in 20 μg/mL PGN-treated tumor cells and in 1.5 mg/kg PGN-treated mouse tumor models. However, Akt3 was raised via IL13 in the irradiated intestine and human intestinal cell line after the same treatment. Finally, PGN activated mTOR via IL13/AKT3 in the intestine and restored intestinal structure and function. As an adjuvant to radiotherapy, PGN inhibited tumorigenesis by suppression of mTOR activity. To summarize, the IL13/AKT3/mTOR pathway was lessened in PGN-treated irradiated tumors but was raised in the normal intestine tissue. This distinct effect of PGN on normal and tumor tissues during pelvic radiotherapy suggests that PGN may be a promising adjuvant therapy to radiation.


INTRODUCTION
About 300,000 patients with gynecologic, bladder, rectal, and prostate cancers undergo pelvic radiotherapy worldwide every year. However, nine out of ten patients develop a permanent change in their bowel habits following radiation [1] irrespective of the quantum of radiation dose [2]. Bowel frequency, loose or liquid stools, fecal incontinence, and the need for undergarment protection were significantly more frequent in radiotherapy patients [3]. The side effects of pelvic radiotherapy such as bleeding, fistula formation, bowel obstruction, and secondary malignancy are very grave [4]. There is a need to develop anti-tumor drugs having high efficacy, but minimal bowel toxicity.
The effect of TLRs on tumor cells is still unclear [10]. In some tumor types, TLRs promote tumor proliferation and survival, as seen with TLR9 agonists [11], Pam2 lipopeptides (TLR2/6 ligands) [12], and flagellin [13]. However, TLRs have also been shown to be directly involved in tumor apoptosis. For example, TLR5 activation by flagellin elicited an innate immune response, causing decreased tumorigenesis in breast cancer cells [14]. TLR2 inhibited the production of the inflammatory cytokine interleukin-18 (IL-18) and protected mice from diethylnitrosamine (DEN)-induced hepatocellular carcinoma [15]. Loxoribin, a TLR7 ligand, inhibited tumor growth in xenograft models of colon and lung cancers, and these anti-tumor effects were mediated by increased CD4 + T cell proliferation and reversal of Treg-mediated immunosuppression via dendritic cells (DCs) [16]. TLR9 agonists, such as CpG oligodeoxynucleotides, are under clinical trials for the treatment of several hematopoietic and solid tumors [17]. In light of these dual functions, it is imperative to further explore and delineate the functional role of TLRs in tumorigenesis.
The possibility of combining radiation and immunebased therapies to achieve better microenvironmental protection and tumor immunogenicity has recently emerged [18]. A TLR7 agonist was shown to possess adjuvant activity when combined with local radiotherapy [19][20][21]. However, the authors did not describe the effect of the therapy on normal tissues. In contrast, a TLR9 agonist limited the efficacy of cancer radiotherapy [22,23]. We have previously reported the protective effect of peptidoglycan (PGN) against toxicity induced by ionizing radiation (IR) [24]. However, the effect of PGN administration on the intestine and pelvic tumor during pelvic cancer radiotherapy has not been established.

Radiotherapy combined with PGN inhibited tumor growth more effectively
As shown in Figure 1A, tumor volumes increased steadily to approximately 5 times their initial size in both untreated and mice treated with PGN alone. As a monotherapy, PGN at a dose of 1.5 mg/kg per mouse had no effect on tumor burden or survival relative to the PBS control. However, 15 Gy local radiotherapy and 15 Gy + PGN treatment significantly reduced tumor size, suggesting that radiation has an inhibitory effect on tumor growth. The effects of 15 Gy radiation alone were transient, and tumor volumes began to increase 12 days after radiotherapy; however, the tumor volumes in the 15 Gy radiation + PGN-treated group continued to decrease, with significant differences between the radiation alone and radiation + PGN treatment groups at 16,18, and 20 days after initial local irradiation (p<0.05, Figure 1A). An additional 15 Gy dose of radiation was administered on day 18, which also had an inhibitory effect on tumor growth. At day 70, all mice in the combination treatment group were alive, while 90% of mice treated with radiation alone were alive ( Figure 1B). All untreated and mice treated with PGN alone died by day 70 ( Figure 1B). Tumor sections at 1.25 and 3.5 days after IR showed morphological characteristics of apoptosis, specifically cell shrinkage and chromatin condensation ( Figure 1C). Sections from mice that were irradiated showed more apoptotic and karyolitic cells than those that were not. On post-IR day 9, tumor weights from irradiated mice were less than 0.5 g, while tumors from mice that were not irradiated were more than 2.5 g (p<0.05, Figure 1D, 1E).
The downstream target of the phosphatidylinositol-3-kinase (PI3K) /AKT pathway is the mammalian target of rapamycin (mTOR). mTOR is a serine/threonine-specific protein kinase that boosts cell growth and proliferation. Irradiation of HCT116 cells transiently reduced expression of both mTOR and phosphorylated mTOR between 24 and 32 h; however, IR in conjunction with PGN treatment continuously decreased mTOR and phosphorylated mTOR expression from 24-48 h. Similar effects were observed in CT26 cells ( Figure 1F).

PGN can promote the recovery of intestinal structure and function after irradiation
Stool formation assessments were carried out in order to detect malabsorption and hypoperistalsis. The number of fecal particles in the colon did not significantly differ between the radiation-alone and radiation + PGN groups at 1.25 days after IR, but was significantly less at 3.5 and 9 days after IR in the radiation alone group (p<0.05). At 3.5 days after IR, the feces of the radiationalone mice were soft, thick, and gray, but appeared hard and normal-colored in the radiation + PGN group ( Figure  2A, 2B).
The body weight of mice in the PGN-alone and untreated groups did not change up to 20 days, but significantly decreased in the irradiated mice within 10 days after IR. In addition, mice exposed to radiation alone lost more weight in 20 days and recovered more slowly than irradiated mice who also received PGN ( Figure 2C).
The untreated and PGN-alone treated mice had integrated villus epithelium and crypts. The intestinal epithelium of irradiated mice exhibited severe radiation damage 3.5 days after IR, as evidenced by eroded and truncated villi tips as well as significant necrosis of epithelial cells, vacuolization, and loose structure. Administration of PGN resulted in reduced intestinal damage at 3.5 days after IR, and the histology showed return to normalcy by day 9. The average lengths of villi in mice of the radiation-alone group were 158 μm and 250 μm, but were 312 μm and 376 μm at 3.5 and 9 days after IR, respectively, in the radiation + PGN group (p<0.05, Figure 2D, 2E).
H&E staining of the crypts of irradiated mice 3.5 days after IR showed sustained and significant crypt damage: elongated shape, deformation, and crypt numbers were reduced to approximately 96/circumference. Co-treatment with PGN resulted in intact crypts arranged in neat rows, with a significantly higher number of crypts/ circumference (114, Figure 2F). Immunochemistry analysis demonstrated that PGN induced mTOR expression after IR treatment ( Figure 2G).

PGN attenuated the effects of radiation on intestinal crypts
The proliferative capacities of the intestines and tumors were measured by Ki67 staining at 3.5 days after irradiation. There were significantly more number of Ki67-positive cells in the intestines of the radiation + PGN group than in the radiation-alone group; however, fewer Ki67-positive cells were observed in the tumors of the radiation + PGN group than in the radiation alone group (p<0.05, Figure 3).

PGN differentially regulated IL13 and TNF-α expression in the intestine and tumor 9 days after IR
The immune microenvironment influences the response to therapy. Intestinal epithelial cell (IEC) homeostasis and repair which is mediated through microbe-sensing, TLRinduced inflammatory pathways and inflammation-associated cancer development, is also influenced by inflammatory cytokines. As shown in Figure 4A, IL13, IL1α, IL22, IL2,  IL5, IL21, IL6, IL10, IL27, IFNγ, TNFα, IL4, and IL17 cytokines were expressed in subsets of mice intestines and tumors 9 days after IR. Among these cytokines, TNF-α expression did not show any marked difference among intestines of mice of any of the treatment groups, but was   assays using bead technology were utilized to detect cytokine production (μg cytokine/mg tissue weight) in intestinal and tumor tissues that were untreated, treated with PGN alone, radiation alone, or radiation + PGN at 9 days after IR. B. Western blot analysis of mTOR and phospho-mTOR expression in HCT116 cells following treatment with radiation alone or radiation + PGN in the presence or absence of IL13 (0.8 and 1.2 ng/mL), IL13 neutralizing antibody (0.12 and 0.2 μg/mL), or TNF-α neutralizing antibody (0.04 and 0.08 μg/mL). increased in tumors of mice treated with radiation + PGN compared with untreated, PGN-alone, and radiation-alone treated mice. IL13 expression was significantly increased in irradiated small intestines; it was highest in the intestines treated with IR combined with PGN. However, IL13 tended to be lower in irradiated tumors as compared to untreated tissues; it was the lowest in tumors treated with IR and PGN.
Different concentrations of IL13, IL13 neutralizing antibody, and TNFα neutralizing antibody were applied to PGN + radiation-treated HCT116 cells. mTOR and phospho-mTOR expression increased with increasing concentrations of IL13. Increasing concentrations of IL13 neutralizing antibody decreased phospho-mTOR expression; however, addition of TNFα neutralizing antibody had the opposite effect. This suggested that PGN promoted the secretion of TNFα by tumors following irradiation and that TNFα plays a role in radiation-induced inhibition of tumor growth. TNFα expression did not change significantly in irradiated intestines, suggesting that PGN does not have a similar effect in the normal tissue ( Figure 4B).

Akt3 but not Akt1/2 plays an important role in PGN's differential effects on tumor and intestinal tissues after IR
Expression of a series of proliferation-related genes was investigated in tumor and intestinal tissues by real-time PCR. These included EGFR (two transcripts: EGFR-1,2), AKT2 (two transcripts: AKT2-1,2), PIK3R1, PIK3R2, PIK3R3, β-catenin, AKT3, AKT1, Casp9, PTEN, PIK3CB, and EGF. Within the AKT family, the expression of AKT1 was highest and AKT3 was lowest in both the intestines and the tumors. AKT1 and AKT2 transcript variant 1 (AKT2-1) expressions did not change markedly, but AKT2 transcript variant 2 (AKT2-2) and AKT3 levels were significantly increased in the intestines but were decreased in the tumors of mice treated with PGN + radiation compared to radiation alone ( Figure 5A). However, AKT2-2 expression was lowest in the untreated tumors among all four groups. These results suggest that AKT3 may be a key player in PGN's differential biological effects on intestinal and tumor tissues following IR.
FHs 74 Int small intestine epithelial cells and HCT116 colorectal carcinoma cells were used to represent intestinal and tumor tissues, respectively. Akt3 expression was detected in irradiated cells treated with 10, 20, or 40 μg/mL PGN. As shown in Figure 5C, 10 μg/mL PGN significantly decreased Akt3 expression in irradiated FHs 74 Int and irradiated HCT116 cells. At increasing PGN concentrations, expression of Akt3 increased in irradiated FHs 74 Int and irradiated HCT116 cells. Akt3 expression in irradiated FHs 74 Int treated with 20 μg/mL PGN was higher than in cells treated with radiation alone; however, this effect was not observed in irradiated HCT116 cells. Irradiated HCT116 cells treated with 40 μg/mL PGN closely resembled irradiated FHs 74 Int cells ( Figure 5C). PGN at 20 μg/mL was selected for the treatment of cells.
AKT1 -/mice die shortly after birth. Therefore, we utilized AKT1 +/mice in these studies. The intestinal crypts of irradiated AKT1 +/mice and AKT2 -/mice treated with PGN had features consistent with those of irradiated C57Bl/6 mice, but the crypts of irradiated AKT3 -/mice did not proliferate significantly following PGN treatment, as shown by Ki67 staining (Figure 5D, 5E).

IL13 stimulates AKT3 expression
As shown in Figure 6, irradiated HCT116 cells were treated with PGN and increasing concentrations of IL13. Akt2 and Akt3 expression increased at the higher IL13 concentrations. Although the addition of IL13 neutralizing antibody increased Akt2 expression even further, Akt3 expression was reduced ( Figure 6A), suggesting that IL13 activated Akt3 but not Akt1 or Akt2 expression. Akt3 expression was abrogated when IL13 was knockdown by shIL13 ( Figure 6B).

DISCUSSION
After IR treatment with PGN, a TLR2 agonist, resulted in an enhanced antitumorigenic effect as compared to radiotherapy alone. Moreover, PGN mitigated the intestinal toxicity induced by IR. This effect was through differential production of IL13 and amplified via Akt3 and mTOR.
The dose of PGN was a key determinant of this effect. Akt3 was lowered in HCT116 cells but was markedly increased in FHs 74 Int cells upon treatment with 20 μg/ml PGN. A dose of 1.5 mg/kg of PGN in vivo induced intestinal proliferation and colorectal tumor suppression. According Fichera and Giese's [25], about 200 μg of PGN was required to saturate the binding sites in 1 μg of membranes. PGN may bind to the membrane receptor sites of the susceptible cells. Thereafter, the large amounts of PGN captured by the membranes may no longer be available for binding and inhibition of other cells, leading to their survival and proliferation.
Schaub et al. [26] have previously reported that pulmonary administration of PGN resulted in IL13 secretion in vivo. Ruíz-González et al. [27] also showed that keratinocytes treated with PGN increased IL13 production. Stimulation of IL13, a Th2 immune cytokine, led to an increase in crypt cell proliferation and goblet cell size and number via activation of PI3-kinase/AKT [28]. Additionally, Farmer et al. [29] reported that IL13 could protect mouse intestine from ischemia and reperfusion injury, and IL13 has been shown to promote colon carcinoma cell survival in a www.impactjournals.com/oncotarget PI3-kinase dependent manner [30]. In this study, both PGN and IR treatments stimulated intestinal IL13, with combined PGN and IR treatment causing the highest IL13 production. Both PGN and IR decreased IL13 production in the tumor, with the lowest level detected in the PGN + IR-treated tumor.
Apart from the differences in the expression of IL13 in the small intestine and tumor, IL13 also has two kinds of receptor [31]. IL4Rα and IL13Rα1 dimerize to form the receptor, which binds IL13 with high affinity [32] and the complex mediates signal transduction through the JAK/STAT6 pathway [33]. Apart from IL13Rα1, IL13 has another cognate receptor, IL13Rα2, which binds IL13 with a markedly high affinity, although it lacks any significant cytoplasmic domain and, therefore, does not mediate signal transduction [34,35]. It was reported that IL13Rα2 may suppress IL13 signal transduction through internalization of IL13 [36] and that the extracellular domain of IL13Rα2 may serve as a decoy receptor for IL13 [37,38]. IL13Rα2 is highly expressed in several human tumors such as colorectal cancer [39] but is absent in normal tissues such as intestinal epithelium. IL13mediated epithelial architectural and functional effects were dependent on the IL4R/IL13Rα1 signaling pathway but were IL13Rα2-independent 40]. Both IL13 dose and receptors demonstrated that IL13 signal transduction was higher in the intestine but lower in colorectal cancer.
Akt3 significantly activated the downstream mTOR kinase signaling. mTOR is known to stimulate cell growth and proliferation through regulation of ribosomal biogenesis and mRNA translation [54].
In conclusion, PGN has distinct biological effects on irradiated intestinal tissues and colorectal tumors via the IL13-Akt3-mTOR pathway. This pathway was inhibited in irradiated tumors but activated in irradiated intestines.
The role PGN in enhancing the radiation-induced antitumorigenic effect and in reducing the side effects of radiotherapy to the intestine is clinically relevant, and could potentially alter the current standard-of-care of colorectal cancer patients to include TLR2 agonists. With decreased radiotoxicity and increased antitumor activity, a larger therapeutic window could be established, which may result in better patient outcomes.

Cell culture and treatment
CT26.WT is a BALB/c colon carcinoma cell line that was maintained in RPMI1640 medium supplemented with 10% fetal bovine serum (FBS). FHs 74. Int, a normal human intestinal cell line, was purchased from ATCC and was cultured in Hybri-Care Medium ATCC 46-X supplemented with 30 ng/ml epidermal growth factor (EGF, Peprotech, NJ, USA) and 10% FBS. HCT116 is a human colorectal carcinoma cell and was cultured in DMEM medium supplemented with 10% FBS. All cells were incubated at 37°C and 5% CO 2 . Both FHs 74. Int and HCT116 cells were treated with 10, 20, and 40 μg/mL PGN (from Staphylococcus aureus, Sigma, St. Louis, MO, USA). PGN was used at 20 μg/mL in all other in vitro experiments. HCT116 cells were treated with 20 μg/mL of PGN alone, 15 Gy irradiation alone, 15 Gy irradiation followed by 20 μg/mL PGN at 24 h, 15 Gy irradiation followed by 0.8 or 1.2 ng/mL IL13 (Peprotech) 2 h prior to 20 μg/mL of PGN at 24 h, 15 Gy irradiation followed by 0.12 or 0.2 μg/mL anti-IL13 (Peprotech) 2 h prior to 20 μg/mL of PGN at 24 h, or 15 Gy irradiation followed by 0.04 or 0.08 μg/mL anti-TNFα (Peprotech) 2 h prior to 20 μg/mL of PGN at 24 h.

Established in vivo Matrigel-tumor growth assays and treatment
All animal studies were performed in accordance with the Animal Care Guidelines of Soochow University. Five-to seven-week-old male BABL/c mice (SLACCAS, Shanghai, China) were kept in animal maintenance facilities under conditions of controlled illumination (12:12 h light/dark cycle), humidity (30-50%), and temperature (18-22°C) and were fed a normal rodent laboratory diet and water. Mice (112 total) bearing BABL/c colon carcinoma at left abdominal derived from Matrigel (Becton Dickinson, San Jose, CA) suspensions 10 6 CT26.WT cells (ATCC, Manassas, VA) were used. Mouse weights and tumor volume were determined using caliper measurements and the formula volume (mm 3 ) = (length*width 2 )/2.
In the untreated group, 100 μl PBS was administered. In the pharmacotherapy group, an injection of 1.5 mg/kg PGN (1.5 mg/kg) was administered intraperitoneally (i.p). High-dose hypofractionated radiotherapy was adopted so as to reduce the frequency of animals were anesthetized and favor to observe intestinal damage. Irradiation (15 Gy) of the abdomen was performed every 18 days on anesthetized mice (i.p. administration of 0.36% chloral hydrate at 0.8 mL/100 g body weight) using a Philips SL18 X-ray system (9 MeV electron beam irradiation, Redhill, UK) at a dose rate of 200 cGy/min following the biosafety guidelines observed in China. For combination treatments, 15 Gy irradiation of the abdomen was followed by i.p. administration of 1.5 mg/ kg PGN at 24 h. Following irradiation, mice were returned to cages (4 mice/cage) and were given free access to food and water. Ten mice per group were used for recording body weight, tumor size and survival studies every two days.

Vector construction and transfection
Full length coding sequences of Akt1, 2, 3 genes were cloned and inserted into the pEGFP-C3 vector (Clontech, Mountain View, CA, USA) and transfected into HCT116 cells via DNA Transfection Reagent (Biotool, Houston, TX, USA) per the manufacturer's instructions. Cells were exposed to 15 Gy irradiation 24 h after transfection and half of these cells were treated with 20 μg/mL PGN 48 h after transfection. All cells were collected 3,5,8,10,22, and 24 hours after PGN treatment.

Assays for stool formation
BALB/c mice were sacrificed 1.25, 3.5, and 9 days after IR and the entire colon starting from the anus was harvested. Loose, yellow content in the lumen was defined as poor stool formation or diarrhea, while solid, dark, granulated content was defined as formed stool.

Haematoxylin and eosin (H&E) staining and immunohistochemistry
Tissues (tumor and normal intestine) were collected from mice and fixed in formalin solution. Tissue alterations were evaluated by histological analysis after H&E staining. For intestines, at least 20 villi were measured by length in each mouse. Complete crypts were also counted.
For immunohistochemical staining, paraffin blocks were cut into 4-μm sections that were mounted, deparaffinized in xylene, and rehydrated in decreasing concentrations of ethanol. Antigen retrieval was performed using citrate buffer, heating sections in a pressure cooker for 5 min and subsequently cooling to room temperature. Blocking of endogenous peroxidases was accomplished by incubating sections in 3% hydrogen peroxide for 5 min. Ki67 (BD Pharmingen, San Diego, CA, USA, 1:200)/mTOR (Cell Signaling, 1:100) antibody was incubated with sections overnight at 4°C. Immunostaining was performed using an Envision System and diaminobenzidine visualization (Dako) according to the manufacturer's instructions. Sections were counterstained with haematoxylin for 1 min, rinsed in water, dehydrated in increasing concentrations of ethanol followed by clearance with xylene, and cover-slipped permanently for light microscopy.

Statistical analyses
The data were analyzed using Windows SPSS version 10.0 (SPSS Inc., Chicago, IL, USA). Statistical analyses were performed using a Student's t-test. The data are presented as mean ± standard deviation. Survival data were assessed using Kaplan-Meier analysis. A p-value <0.05 was considered statistically significant.

CONFLICTS OF INTEREST
The authors declare that they have no conflict of interest.

Financial support
The National Natural Science Foundation of China (Grant No. 81001317/81172597/81372920); the Priority Academic Program Development of Jiangsu Higher Education Institutions; Defense basic research projects.