Eritoran Suppresses Colon Cancer by Altering a Functional Balance in Toll-Like Receptors That Bind Lipopolysaccharide

Published OnlineFirst June 21, 2016; DOI: 10.1158/0008-5472.CAN-16-0172

Cancer Microenvironment and Immunology Research

Eritoran Suppresses Colon Cancer by Altering a Functional Balance in Toll-like Receptors That Bind Lipopolysaccharide Wei-Ting Kuo1, Tsung-Chun Lee1,2, and Linda Chia-Hui Yu1

Abstract

Colorectal carcinogenesis is affected by overexpression of the cell lines displayed increased cell proliferation and cell-cycle lipopolysaccharide (LPS) receptors CD14 and TLR4, which antag- progression following LPS challenge. This effect was inhibited onize each other by affecting epithelial cell proliferation and by eritoran and by silencing CD14 or TLR4. In contrast, apoptosis apoptosis. Eritoran is an investigational drug for sepsis treatment induced by eritoran was eliminated by silencing CD14 or protein that resembles the lipid A moiety of LPS and therefore acts as a kinase Cz (PKCz) but not TLR4. Lastly, LPS and eritoran caused TLR4 inhibitor. In the present study, we explored the potential hyperphosphorylation of PKCz in a CD14-dependent and TLR4- therapeutic uses and mechanisms of action of eritoran in reducing independent manner. Blocking PKCz activation by a Src kinase colon cancer progression. Eritoran administration via intracolo- inhibitor and a PKCz-pseudosubstrate prevented eritoran- nic, intragastric, or intravenous routes signiﬁcantly reduced tumor induced apoptosis. In summary, our work offers a preclinical burden in a chemically induced mouse model of colorectal proof of concept for the exploration of eritoran as a clinical carcinoma. Decreased proliferation and increased apoptosis were treatment, with a mechanistic rationale to reposition this drug observed in mouse tumor cells after eritoran treatment. In vitro to improve the management of colorectal cancer. Cancer Res; 76(16); cultures of mouse primary tumor spheroids and human cancer 4684–95. 2016 AACR.

Introduction plasmic signaling, resulting in the production of proinflammatory cytokines, such as myeloid differentiation factor (MyD88), Colorectal carcinoma is characterized by unlimited cell pro- MAPK, and inhibitor of kB(IkB)/nuclear factor-kB(NFkB; liferation and resistance to cell death. This is in contrast to refs. 11, 12). Accumulating evidence suggests that LPS receptors normal intestinal tissue, which has a balanced epithelial turn- are involved in the dysregulation of epithelial apoptosis and over that is maintained by crypt renewal and surface/villus proliferation that predisposes the tissue of the colon to tumor- apoptosis (1, 2). The development of colon cancer has been igenesis. Previous animal studies have indicated that TLR4/ linked to the aberrant recognition of enteric bacterial bypro- MyD88 signaling causes tumor cell hyperproliferation via path- ducts through both myeloid and epithelial immune receptor ways dependent on cyclooxygenase (COX), EGFR, and b-cate- signaling (3, 4). The surface of the intestinal mucosa is con- nin (5, 13–15). In vitro studieshaveshownadirecteffectofLPS stantly exposed to a large amount of bacterial lipopolysaccha- via TLR4/MD2 signaling on the stimulation of intestinal epi- ride (LPS). Upregulated expression of the LPS receptor com- thelial cell proliferation and a resistance to TNF-related apo- plex, CD14/Toll-like receptor (TLR4)/MD2, has been identified ptosis-inducing ligand (TRAIL)–induced apoptosis (16–18). in the intestinal mucosa of animal models and patients with Our recent study demonstrated a functional antagonism inflammatory bowel disease and colorectal cancers (5–8). between CD14 and TLR4 in the regulation of epithelial apo- The binding of LPS to CD14 on the lipid raft domain ptosis and intestinal carcinogenesis (19). TLR4 counteracted activates a cascade of lipid messengers and protein kinase Cz CD14-dependent epithelial apoptosis and promoted survival (PKCz) to recruit TLR4 to form a complex with CD14 (9, 10). in developing cancer cells (19–21). The transfer of LPS to TLR4/MD2 subsequently initiates cyto- Eritoran (E5564) is an investigational drug for the treatment of severe sepsis that functions as a TLR4 inhibitor because of its structural similarity to the LPS lipid A moiety, which is composed 1Graduate Institute of Physiology, National Taiwan University College of 6 acyl chains (22, 23). Recent advances have been made in 2 of Medicine,Taipei,Taiwan. Department of Internal Medicine, National understanding the interaction between LPS and its receptor based Taiwan University Hospital, Taipei, Taiwan. on studying the crystal structures of TLR4/MD2 complexed with Note: Supplementary data for this article are available at Cancer Research LPS or eritoran (24). It is recognized nowadays that LPS binding to Online (http://cancerres.aacrjournals.org/). CD14 leads to ligand transfer onto the TLR4/MD2 complex, Corresponding Author: Linda Chia-Hui Yu, National Taiwan University College which dimerizes to initiate downstream signaling (11, 25, 26). of Medicine, Suite 1020, #1 Jen-Ai Rd. Sec. I, Taipei 100, Taiwan. Phone: 886-02- However, eritoran (with only 4 acyl chains) does not induce 23123456-88237; Fax: 886-02-23964350; E-mail: [email protected] TLR4/MD2 complex dimerization or cytoplasmic signaling doi: 10.1158/0008-5472.CAN-16-0172 (27). Animal studies using eritoran have demonstrated an inhi- 2016 American Association for Cancer Research. bition of proinflammatory TLR4 signaling, improvement of

4684 Cancer Res; 76(16) August 15, 2016

Tumor Progression by Dysregulated LPS Signaling

intestinal microcirculation, and higher survival rates in endotoxe- tumor formation was clearly observed on day 63 following the mic rodents (22, 28) and in influenza- and syncytial virus– start of the AOM/DSS protocol (19), and eritoran was given on infected mice (29, 30). However, whether eritoran binds to CD14 days 63, 70, and 77. Based on pilot studies, the intracolonic (i.c.) and triggers CD14-mediated apoptosis in colon cancer cells dose of eritoran was 10 mg dissolved in 150 mL of PBS per bolus per remains unclear. mouse. For intragastric (i.g.) administration, 500 mg of eritoran The aim of our study was to investigate whether the adminis- was dissolved in 250 mL of PBS per bolus per mouse. For tration of eritoran inhibits colorectal carcinoma development in intravenous (i.v.) administration, mice were injected with 5, mouse models. In the current study, we explored the potential 10, or 20 mg/kg of eritoran dissolved in 250 mL of PBS per bolus therapeutic use of eritoran in tumor reduction and examined its per mouse (22, 28, 29). underlying anticancer mechanisms using primary mouse tumor spheroids and human adenocarcinoma cell lines. Measurement of cell apoptosis Intestinal sections were stained using a terminal deoxynucleo- Materials and Methods tidyl transferase dUTP-biotin nick end labeling (TUNEL) detec- Animals tion kit (Merck). The percentage of TUNEL-positive cells was 2 BALB/c mice 7 to 10 weeks of age were used in this study. calculated, normalized, and expressed per 1 mm from 10 tumor Experimental procedures were approved by the Laboratory Ani- sections from each group of mice. In addition, cancer cells were mal Care Committee of the National Taiwan University College of homogenized, and the supernatant was measured for DNA frag- Medicine. mentation by using a cell death detection ELISA kit (Roche; refs. 19, 20, 31, 32). Antibodies All antibodies were purchased from Cell Signaling Technology, Primary intestinal spheroid cultures except otherwise stated. Primary antibodies included mouse anti- Colonic tumors were cultured in vitro as spheroids following human PCNA (1:2,000), phospho (p)- and total (t)-IkB previously methods (33, 34) with modification. Briefly, intestinal (1:2,000), p- and t-ERK1/2 (1:4,000), p- and t-JNK (1:2,000), fragments with tumors were incubated in a chelation buffer for 60 p- and t-Akt (1:2,000), p-PKCz (Thr410; 1:2,000), p-PKCz minutes on ice to remove most of the normal epithelial cells. (Thr560; 1:5,000; Abcam Epitomics), t-PKCz (1:2,000), p-tyro- Tumor fractions were then incubated in a digestion buffer with sine (1:2,000; EMD Millipore), CD14 (1:50 for immunofluores- collagenase and dispase for 2 hours at 37C. After counting cence and 1:2,000 for Western blot; R&D Systems), TLR4 (1:100 isolated single tumor cells, approximately 10,000 cells were for immunofluorescence and 1:2,000 for Western blot; Santa Cruz plated in 300 mL of ice-cold Matrigel (BD Biosciences; 3:1 ratio Biotechnology), TLR2 (1:2,000; Santa Cruz Biotechnology), with crypt culture medium) per well in 24-well plates. After MyD88 (1:2,000), bromodeoxyuridine (BrdUrd; 1:200; Abcam), Matrigel polymerization, 300 mL of crypt culture medium was mouse IgG1 and IgG2a isotype controls (R&D Systems), rat IgG2a overlaid. The overlaying medium was refreshed every 2 to 3 days, isotype controls (Abcam), and b-actin (1:10,000; Sigma; refs. 19, and images were captured under a light microscope. Spheroids 31). The secondary antibodies used were goat anti-mouse IgG were collected by dissolving the Matrigel with a recovery solution conjugated to horseradish peroxidase (1:2,000), and to Alexa 594 (BD Biosciences). or 488 (1:1,000; ThermoFisher). Spheroids grown for 5 days were challenged by adding LPS (from E. coli O26:B6; Sigma) or eritoran to the overlaying medi- Models of chemically induced colorectal carcinoma um, and the effects this had on cell proliferation and cell death Mouse models of colitis-associated colorectal carcinoma were measured. Spheroid area was determined using imaging were prepared according to previously established methods software (AxioVision Rel. 4.8). Alternatively, time lapse imaging (19). Briefly, mice were injected i.p. with azoxymethane (AOM; for spheroid growth was performed using a real-time cultured cell 10 mg/kg body weight; Sigma) at the beginning of the exper- monitoring system (ASTEC Co.). iment (day 0). After 7 days, 2% dextran sodium sulfate (DSS; MP Biomedicals) was administered via the drinking water for 4 Cell lines days, and this was followed by 3 days of regular water. This All cells were purchased from the ATCC/Bioresource Col- cycle of AOM/DSS was performed 3 times. Body weight was lection and Research Center (BCRC), and STR-PCR profile was measured every week, and the animals were sacrificed on day 84 performed by BCRC. Human colon adenocarcinoma HT29 for macroscopic inspection and histologic analysis. The num- and Caco-2 cells were exposed to LPS or eritoran for 48 or bers of tumors were determined macroscopically under a dis- 24 hours, and the effects of this exposure on cell proliferation secting microscope, and the area covered by tumors was mea- and cell death, respectively, were measured (20, 35). In some sured using an imaging software (AxioVision, Zeiss). Paraffin- experiments, cells were pretreated with inhibitors 30 minutes embedded tissues were used for histologic assessment and prior to challenge. These inhibitors included 1 mmol/L Src tumor grading by a pathologist blinded to mouse treatment inhibitor-1 (Sigma), 20 mmol/L inhibitory PKCz-pseudosub- groups. Four to 6 tumors per mouse and 10 mice per group strate (Merck), and 120 mmol/L z-DEVD-FMK (Calbiochem). were examined. In other settings, cells were challenged with lipoteichoic acid (LTA), which is a TLR2 ligand (36) for 48 hours to assess Administration of eritoran proliferation levels. Eritoran tetrasodium, a TLR4 inhibitor (a kind gift from Eisai Human embryonic kidney (HEK)-293T cells were transfected Inc.), was administered intracolonically, intragastrically, and with a huCD14 plasmid (see below) prior to eritoran challenge to intravenously after the induction of colorectal carcinoma. Colon analyze cell death response and signaling pathways.

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4685

Kuo et al.

4686 Cancer Res; 76(16) August 15, 2016 Cancer Research

Tumor Progression by Dysregulated LPS Signaling

Evaluation of cell-cycle progression Results Cell cycle was assessed by using a BrdUrd incorporation and Suppression of colon tumor burden by eritoran correlates with propyl iodide assay. Briefly, cells challenged with LPS or eritoran inhibition of cell proliferation and augmentation of cell for 48 hours were incubated with 10 mmol/L BrdUrd during the apoptosis last 2 hours at 37 C, and then fixed with ice-cold 70% ethanol for A colorectal carcinoma mouse model administering three 24 hours. Cells were denatured with 2N HCl and incubated with cycles of AOM/DSS was utilized to evaluate the anticancer anti-BrdUrd for 1 hour followed by secondary antibody conju- effects of eritoran through i.c., i.g., and i.v. administration. gated to Alexa 488. Finally, cells were incubated with propyl Eritoran was administered on days 63, 70, and 77 after the iodide for 30 minutes and analyzed by a flow cytometry with first AOM injection, and mice were sacrificed on day 84 FlowJo Software (Treestar). The percentage of cells in the G –G , 0 1 (Fig. 1A). A significant decrease in tumor cell multiplication S, and G –M phases was determined using Mod Fit LT cell cycle 2 and tumor area were observed in mice treated with eritoran analysis software (Verity Software). via i.c., i.g, or i.v. routes (Fig. 1B–D). Colon tumor stages in mice administered eritoran were less advanced and were siRNA-mediated knockdown mostly present as low-grade dysplasia (Fig. 1E and F). The Gene silencing by siRNA was conducted before the chal- doses used for the i.c. and i.g. routes were based on our pilot lenge with LPS or eritoran. Spheroids were transfected with 50 studies,whichshowednoanimalmortality,andthei.v. nmol/L siRNA oligonucleotides from a pool of four target- concentrations were based on results obtained from experi- specific 20- to 25-nt siRNAs for CD14, TLR4, MyD88, PKCz, mental models of endotoxemia. A dose-dependent inhibitory or scrambled controls (Santa Cruz Biotechnology) using Lipo- response in tumor growth with i.v. injection of eritoran was fectamine RNAiMAX Reagent (Life Technologies) in a serum- observed (Supplementary Fig. S1). fi free basal medium for 6 hours. The basal medium was Tumor cell proliferation was signi cantly lower after eritoran replaced with a serum-containing complete medium for 2 treatment, as measured by immunoblotting of the proliferating days prior to the challenge. Human cell lines were transfected cell nuclear antigen (PCNA; Fig. 1G). Using the TUNEL assay to with 50 nmol/L of siRNA as described above. After 24 hours, determine cell apoptosis, we observed higher numbers of TUNEL- cells were trypsinized and reseeded in culture plates for 2 days positive cells per area of cancerous tissue in mice treated with prior to challenge (19). eritoran compared with those administered the vehicle (Fig. 1H and I).

Plasmid constructs and cell transfection The expression vectors pMyc-CMV1-huTLR4 and pcDNA3- LPS/TLR4-induced epithelial proliferation is prevented by huCD14 were kind gifts from Dr. Douglas T. Golenbock in the eritoran in primary mouse tumor spheroids University of Massachusetts Medical School. Cells were trans- To clarify whether eritoran has a direct effect on colonic fected with pMyc-CMV1-huTLR4 or pcDNA3-huCD14 plasmids cancer cells or an indirect effect by altering the tumor micro- or respective plasmids for 24 hours. Cells were trypsinized and environment, primary cultures of colonic tumor spheroids in reseeded into cell culture plates for 2 days prior to challenge with the absence of immune/myeloid cells were developed (Fig. 2A). LPS or eritoran (19). Transcripts of Lgr5 (a marker of stem cells), CD14, TLR4, and MD2 were identified in colonic tumor spheroids (Fig. 2B). Absence of CD68 (a macrophage marker) and CD3 (a T-cell Statistical analysis marker) was also confirmed in primary cultures of tumor All data are expressed as the mean SEM. Data that met spheroids (Fig. 2B). the assumptions of normality were analyzed by one-way Tumor spheroids were challenged with various concentrations ANOVA followed by the Student–Newman–Keuls post-hoc of LPS and eritoran to determine a nonapoptotic dose and to test for multiple comparisons. Significance was established examine parameters of cell proliferation. No signs of cell apo- at P < 0.05. ptosis were observed at any test dose of LPS (Fig. 2C). In contrast, Detailed information on the animals, histopathologic exami- eritoran triggered cell death at a concentration of 10 to 50 mg/mL nation and tumor grading, primary intestinal spheroid cultures, but not at 1 mg/mL (Fig. 2C). Therefore, 1 mg/mL of LPS or eritoran human colon adenocarcinoma cell lines, siRNA-mediated knock- was used to assess proliferation, and 50 mg/mL was used in the cell down, plasmid constructs and cell transfection, RT-PCR, Western death experiments. blotting, and immunofluorescent staining is provided in the Mouse tumor spheroids treated with 1 mg/mL of LPS showed a Supplementary Text. 2-fold increase in spheroid area (Fig. 3A and B). Addition of

Figure 1. Decreased colon tumor growth after eritoran treatment in mice. Mice were subjected to three cycles of AOM/DSS and treated with vehicle or eritoran intracolonically (i.c., 10 mg per mouse), intragastrically (i.g., 500 mg per mouse), or intravenously (i.v., 20 mg/kg per mouse). A, scheme of the eritoran treatment regimen (TR). E, days eritoran was administered; X, day of sacriﬁce. B, representative images of mouse distal colons. Arrowheads, tumors. Scale bar, 1 mm. C and D, tumor multiplicity and mean tumor area per mouse were decreased after eritoran treatment. E and F, representative histopathologic images of colon tumors. Tumor stage in each mouse group is expressed as the percentage of tumors classiﬁed as low- and high-grade dysplasia, and carcinoma. G, Western blot analysis of PCNA in tumor tissues after eritoran treatment. H and I, representative images of TUNEL-positive cells (arrowheads) in tumor tissues. The number of apoptotic cells per mm2 in colonic tumors was increased after eritoran treatment. Scale bar, 50 mm. n ¼ 8–12/group. , P < 0.05 vs. vehicle.

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4687

Kuo et al.

Figure 2. Primary cultures of mouse colonic tumor spheroids for the assessment of responses to LPS and eritoran. A, representative images of mouse tumor spheroids cultured for various days. Magniﬁcation, 200. B, transcript levels of Lgr5 and the LPS receptors CD14, TLR4, and MD2 in tumor spheroids (a). Negative controls are without templates. Absence of CD68 (a macrophage marker) and CD3 (a T-cell marker) in primary spheroid cultures is also shown (b and c). RAW264.7 and mouse spleen tissues served as positive controls for CD68 and CD3, respectively. C, tumor spheroids were treated with various doses of LPS or eritoran for 24 hours to examine cell apoptosis and to determine a nonapoptotic dose for the following cell proliferation assays. , P < 0.05 vs. 0 mg/mL.

eritoran inhibited the increase of spheroid growth caused by LPS ing of CD14 or PKCz eliminated eritoran-induced cell apopto- challenge (Fig. 3A and B). The 48-hour time lapse imaging of sis (Fig. 4A). Apoptotic levels in double knockdowns of TLR4/ spheroid growth is shown in Supplementary Videos S1–S3. Gene CD14 or TLR4/PKCz were comparable with those in scrambled silencing of CD14 or TLR4 also diminished the LPS-induced controls (Fig. 4A). spheroid growth (Fig. 3B). The knockdown efficiency in spheroid PKCz is composed of a catalytic enzymatic site and an auto- cultures was confirmed by a decrease in protein levels (Supple- regulatory domain with a pseudosubstrate sequence (37, 38). mentary Fig. S2). Previous studies have demonstrated that lipid messengers simul- Cell-cycle analysis by PI nuclear staining revealed that a higher taneously induce the removal of pseudosubstrate-dependent proportion of cells were in the S and G2–M phases following the autoinhibition and the phosphorylation of the catalytic loop at challenge with LPS but not following an eritoran challenge (Fig. Thr410, which then causes Thr560 autophosphorylation (37). In 3C). The ratio of S–G2–MtoG1–G0 cells was significantly greater addition, Src kinase–activated Tyr428 phosphorylation may be in spheroids challenged with LPS compared with the ratio in lipid-independent (39). Phosphorylation of Thr410/Thr560 or untreated controls or in the eritoran-treated group (Fig. 3D). Tyr428 is crucial for the complete activation of the enzymatic Similarly, increased percentage of cells in the S phase by LPS domain of PKCz (37, 39). Our results demonstrated an increased challenge was shown by using a BrdUrd incorporation assay (Fig. phosphorylation of PKCz at Thr410, Thr 560, and Tyr sites 3E and F). The LPS-induced acceleration of cell-cycle progression following eritoran treatment in tumor spheroids, which was was prevented by the addition of eritoran (Fig. 3D and F). attenuated by CD14, but not by TLR4 or MyD88 knockdown Moreover, a challenge with LPS caused increased PCNA levels in (Fig. 4B). spheroid cultures. This increase was inhibited by the addition of Additional transfection experiments were conducted on eritoran (Fig. 3G). HEK293T cells (CD14- and TLR4-negative cells) to confirm that Furthermore, the activation of signaling molecules down- induced expression of CD14 is sufficient to cause a cell death stream of TLR4 was confirmed, which demonstrates an increased response following eritoran challenge, in the absence of TLR4 (Fig. phosphorylation of MAPKs, IkBa, and Akt following LPS chal- 4C). Moreover, increased phosphorylation of PKCz at Thr410, Thr lenge (Fig. 3H). Addition of eritoran or gene silencing of MyD88 560, and Tyr sites after eritoran treatment was observed in the ablated LPS-induced TLR4 signaling and cell proliferation in CD14-transfected HEK cells (Fig. 4D). mouse tumor spheroids (Fig. 3H and I). Eritoran exposure inhibits LPS/TLR4-induced cell proliferation Eritoran-induced apoptosis is dependent on CD14/PKCz, but in human colorectal adenocarcinoma cell lines not on TLR4, in mouse tumor spheroids Human colorectal adenocarcinoma cell lines were used to Primary mouse colonic tumor spheroids showing TLR4 and further investigate the molecular mechanism underlying the CD14 expression were subjected to multiple gene knockdowns eritoran-induced antiproliferation and proapoptotic effects. Dis- to evaluate the molecules involved in the eritoran-induced cell tinct expression patterns of LPS receptor subunits were detected in death response. Previous studies have indicated that the bind- HT29 and Caco-2 cells (Fig. 5A). The receptor expression in HT29 þ þ ing of LPS to CD14 activates a cascade of lipid messengers and cells (CD14 TLR4 ) resembles that of human cancerous tissues, þ protein kinase Cz (PKCz) prior to the recruitment of TLR4 into whereas Caco-2 cells (CD14 TLR4 ) resemble normal colono- lipid rafts (9, 10). Single and double gene silencing of TLR4, cytes (6, 19, 40). CD14,and/orPKCz in spheroids was specificanddidnotalter Human cell lines were challenged with a nonapoptotic dose of any proteins other than the target molecules (Supplementary LPS to examine cell proliferation and cell-cycle progression (Sup- Fig. S2). Compared with scrambled controls, the knockdown of plementary Fig. S3). HT29 showed a higher proportion of cells in TLR4 or MyD88 did not affect the levels of cell death in tumor the S–G2–M phase following LPS challenge, whereas no changes in spheroids following eritoran treatment. In contrast, the silenc- cell-cycle phases were observed in Caco-2 cells (Fig. 5B and C). The

4688 Cancer Res; 76(16) August 15, 2016 Cancer Research

Tumor Progression by Dysregulated LPS Signaling

Figure 3. LPS/TLR4-induced epithelial proliferation is inhibited by eritoran in primary mouse colonic tumor spheroids. Cell proliferation levels were determined in tumor spheroids after exposure to LPS and/or eritoran at a nonapoptotic dose (1 mg/mL) for 48 hours. A, representative images of tumor spheroids after exposure to LPS and/or eritoran. B, areas of tumor spheroids were quantiﬁed after exposure to LPS and/or eritoran. Spheroid cultures were silenced with CD14 or TLR4 siRNA or with scrambled controls. C and D, cell-cycle analysis by propidium iodide (PI) staining in tumor spheroids after exposure to LPS and/or eritoran.

Proportions of proliferative cells are expressed as the ratio of cells in S–G2–M to those in G1–G0 phases. E and F, cell-cycle analysis by using a BrdUrd incorporation assay in tumor spheroids after exposure to LPS and/or eritoran. The percentage of BrdUrd-positive cells represents those in the S phase. G, Western blot analysis of PCNA in tumor spheroids. H, changes in TLR4 downstream signaling pathways in tumor spheroids after exposure to LPS and/or eritoran. I, gene silencing of MyD88 reduced cell proliferation and signaling pathways induced by LPS. n ¼ 6–8/group. , P < 0.05 vs. vehicle.

LPS-induced acceleration of cell cycles in HT29 was eliminated by TLR4-expressing Caco-2 cells by transfection with a pMyc-CMV1- gene silencing of CD14 and TLR4, or by the addition of eritoran huTLR4 plasmid (Fig. 5C). Eritoran inhibited LPS-induced cell (Fig. 5B). In contrast, cell proliferative response was restored in proliferation in TLR4-transfected Caco-2 cells (Fig. 5C).

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4689

Kuo et al.

Figure 4. Eritoran-induced apoptosis in primary mouse tumor spheroids is dependent on CD14/PKCz activation in a TLR4-independent manner. A, exposure to eritoran (50 mg/mL) for 24 hours induced elevated apoptosis in tumor spheroids. This effect was eliminated by knockdown of CD14 and PKCz but not by silencing of TLR4 or MyD88. B, Western blotting analysis of eritoran-induced hyperphosphorylation of PKCz at Thr410, Thr560, and tyrosine sites in tumor spheroids, which were attenuated by the knockdown of CD14 but not of TLR4 or MyD88. C and D, HEK293T cells, which are negative of CD14 and TLR4, were transfected with CD14 by plasmids. Elevated cell apoptosis was observed following LPS challenge in CD14-transfected cells but not in those mock-transfected. , P < 0.05 vs. respective vehicle. n ¼ 4–6/group.

Other than TLR4, TLR2 also activates MyD88-dependent sig- Caco-2 cells (Fig. 5D and E). Gene silencing of CD14, but not naling pathways and increases epithelial cell proliferation (36, 41, TLR4, inhibited the eritoran-induced cell death in HT29 cells (Fig. 42). Bacterial cell wall products, such as LTA and peptidoglycan, 5D). Knockdown of CD14 or PKCz eliminated the eritoran- activate TLR2 signals upon initial binding to CD14 (43, 44). induced cell death in Caco-2 cells (Fig. 5E). In addition, eri- Therefore, we also examined whether eritoran counteracts TLR2 toran-induced phosphorylation of PKCz at Thr410, Thr 560, and þ signaling and cell proliferation in Caco-2 cells (CD14 TLR4 ). Tyr sites was diminished by knockdown of CD14 in both HT29 Our results showed that LTA stimulation increased the proportion and Caco-2 cells (Fig. 5F and G). þ of cells in the S–G2–M phase and the PCNA levels, which were Furthermore, Caco-2 cells (CD14 TLR4 ) pretreated with Src- reduced by gene silencing of TLR2, MyD88, or CD14 (Supple- I1 (a Src kinase inhibitor), PKCz-ps (a PKCz pseudosubstrate), or mentary Fig. S4A and S4B). LTA-induced phosphorylations of IkB z-DEVD-FMK (a caspase-3 inhibitor) displayed a reduction in and Akt were prevented by knockdown of TLR2, MyD88, or CD14 eritoran- and LPS-induced apoptosis (Fig. 6A). Inhibition of Src (Supplementary Fig. S4B). Addition of eritoran ablated the kinase attenuated the phosphorylation of PKCz at Thr410, increase of cell proliferation and downstream signaling induced Thr560, and Tyr sites. This suggests that Src kinase is upstream by LTA (Supplementary Fig. S4C and S4D). of PKCz activation (Fig. 6B).

Eritoran triggers CD14/Src/PKCz-mediated cell apoptosis in Changes in signaling pathways in colon tumors of mice treated human colorectal cancer cells with eritoran Responses were next assessed in the two human cell lines using Lastly, the CD14- and TLR4-dependent signaling pathways an apoptotic dose of eritoran. A 2-fold increase in cell apoptosis were validated in mouse colon cancers. Eritoran treatment was observed following eritoran treatment in both HT29 and stimulated the hyperphosphorylation of PKCz and diminished

4690 Cancer Res; 76(16) August 15, 2016 Cancer Research

Tumor Progression by Dysregulated LPS Signaling

Figure 5. Eritoran inhibits TLR4-dependent cell proliferation and triggers CD14/PKCz-mediated apoptosis in human colorectal adenocarcinoma cells. A, representative immunostaining of CD14 and TLR4 proteins in HT29 and Caco-2 cells. Scale bar, 5 mm. B, cell-cycle analysis of HT29 cells (CD14þTLR4þ) after exposure to a nonapoptotic dose (10 mg/mL) of LPS and/or eritoran for 48 hours. LPS-induced proliferation was eliminated by gene silencing of CD14 or TLR4 in HT29 cells. C, cell-cycle analysis of Caco-2 cells (CD14þTLR4) after exposure to a nonapoptotic dose (1 mg/mL) of LPS and/or eritoran for 48 hours. Transfection with huTLR4 plasmids restored LPS-induced proliferation in Caco-2 cells. D, exposure to eritoran (50 mg/mL) for 24 hours triggered apoptosis in HT29 cells, which was prevented by knockdown of CD14 but not by knockdown of TLR4. E, exposure to eritoran (50 mg/mL) for 24 hours triggered apoptosis in Caco-2 cells, which was inhibited by gene silencing of CD14 and PKCz. F and G, hyperphosphorylation of PKCz at Thr410, Thr560, and tyrosine sites after exposure to eritoran (50 mg/mL) for 20 minutes was reduced by CD14 knockdown in HT29 and Caco-2 cells. Representative blots of total (t)- and phospho (p)-PKCz in cells. , P < 0.05 vs. respective vehicle. Experiments were repeated at least twice. n ¼ 4–6/group. the activation of MAPKs, IkBa, and Akt pathways (Fig. 7), ulation of the functional antagonism by eritoran exerts anti- which correlated to the reduction in tumor burden. cancer effects. In the present study, eritoran (a putative TLR4 inhibitor) signiﬁcantly reduced the tumor burden in the mouse colon. Discussion Our results are in agreement with previous studies showing that Suppression of colonic tumor growth following eritoran TLR4 is involved in colon cancer growth (5, 13–15, 45). We treatment is a combination of effects caused by a reduction concluded that the anticancer effect of eritoran in in vivo animal of TLR4-dependent cell proliferation and an elevation of models is predominantly via direct actions on tumorous epi- CD14-dependent cell apoptosis. These dual modes of action thelial cells. Nevertheless, a minor role of eritoran inhibition by eritoran involve changes in the functional antagonism on myeloid/immune cells cannot be ruled out. An elegant study between LPS receptors (CD14/TLR4) on colon cancer cells. using bone marrow chimera of wild-type and TLR4-mutant Our ﬁndings indicate that dysregulation of the LPS receptor mice indicated that TLR4-dependent signaling originating from subunits is involved in tumorigenesis and that strategic manip- epithelial cells plays a more dominant role than that from

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4691

Kuo et al.

Figure 6. LPS- and eritoran-induced apoptosis in CD14þTLR4 Caco-2 cells is dependent on Src/PKCz activation. Caco-2 cells were exposed to LPS or eritoran (50 mg/mL) for 24 hours to examine cell death. A, pretreatment with Src Inhibitor-1 (a Src kinase inhibitor), PKCz-pseudosubstrate (ps; a speciﬁc inhibitor to PKCz), or z-DEVD-FMK (a caspase-3 inhibitor) prevented LPS- and eritoran-induced apoptosis. B, the hyperphosphorylation of PKCz at Thr410, Thr560, and tyrosine sites caused by LPS and eritoran exposure in Caco-2 cells was eliminated by these inhibitors. Representative blots of total (t)- and phospho (p)-PKCz are shown. Experiments were repeated at least twice. , P < 0.05 vs. respective vehicle; #, P < 0.05 vs. LPS or eritoran exposure in untreated cells. n ¼ 4–6/group.

myeloid cells in causing epithelial hyperproliferation and in increased COX-2/prostaglandin E2 (PGE2) production by epi- shaping a tumor-promoting microenvironment (46). In trans- thelial cells was observed at baseline and correlated with the genic mice with epithelial-speciﬁc TLR4 overexpression, higher numbers of colon tumors upon chemical induction

Figure 7. Eritoran-induced PKCz phosphorylation is accompanied by deactivation of MAPK, IkBa, and Akt signals in mouse colon tumors. Colonic tumor tissues from mice administered eritoran (i.v.) were harvested for examination of signaling pathways. A, Western blots showing the phosphorylation status of PKCz, MAPKs, IkBa, and Akt molecules in mouse colonic tumors after eritoran treatment. B, densitometric analysis of signaling molecules in mouse tumors after eritoran treatment. , P < 0.05 vs. vehicle. n ¼ 8–12/group.

4692 Cancer Res; 76(16) August 15, 2016 Cancer Research

Tumor Progression by Dysregulated LPS Signaling

(15). However, mucosal macrophages are involved in the activation induces tumor cell apoptosis via the Src/PKCz sig- synthesis of amphiregulin (an EGFR ligand) following COX- naling pathways. 2/PGE2 activation, and hence, indirectly promoting neoplasia It is worth noting that the i.v. dose of eritoran needed to (47). Taken together, epithelial TLR4 plays a key role in driving suppress colon cancer was higher than the amount adminis- carcinogenesis and is the dominant site for eritoran-induced tered to endotoxemic animals to inhibit proinflammatory suppression in vivo. signaling and cytokine production (22, 28). The massive loads In primary tumor spheroid studies in vitro, the proliferative of enteric bacteria and LPS produced in the gut lumen and death response caused by LPS or eritoran represented a may account for this discrepancy. The basal level of LPS direct effect on epithelial cells because myeloid cells were (1.8 mg/mL) documented in the luminal content of rodent absent in the culture. The findings in primary spheroids and intestines is much higher than the blood concentration that cell lines provide evidence that TLR4/MyD88-dependent colo- has been reported to stimulate monocyte cytokine production rectal tumor proliferation can be uncoupled from immune cell and septic responses (10 ng/mL). Therefore, the high dose of activation. i.v. eritoran needed for cancer suppression might reflect the Blocking TLR4 by eritoran not only inhibited cell prolif- concentration necessary for effective competitive binding to eration but also triggered apoptosis in mouse and human epithelial receptors in the presences of high levels of enteric colon tumor cells. Previous studies in our laboratory have bacterial LPS. shown that, in the absence of TLR4 signaling, LPS exposure In addition to i.v. injection, topical administration of eritoran, caused epithelial cell death in a CD14-dependent manner either intracolonically or intragastrically, caused tumor regression (19). Moreover, TLR4 overexpression counteracted epithelial in the present study. The i.c. dose needed to achieve tumor apoptosis induced by its coreceptor, CD14 (19). Other suppression was significantly lower than the amount required reports have shown that LPS/TLR4-dependent NFkB activa- when it was delivered intravenously. Intracolonic application tion confers resistance to TRAIL-induced apoptosis in human may be beneficial for limiting unwanted nonepithelial effects. colon cancer cells (18, 48), suggesting that TLR4 signaling Moreover, the bioactivity of eritoran appears to withstand the inhibits various types of stimuli for cell death. Signaling low pH and enzymatic degradation that could occur with i.g. molecules downstream of TLR4/MyD88, such as MAPK and administration. IKK/IkB/NFkB, and noncanonical Akt pathways, have been Eritoran, acting as a CD14 agonist and TLR4 antagonist, may associated with hyperproliferative, prosurvival, and antiapop- be clinically useful in blocking aggressive tumor growth and totic effects (13, 49–51). In the current study, eritoran induced extraintestinal metastasis. A previous study showed that the CD14/Src/PKCz-mediated cell apoptosis in primary spheroid intrasplenic injection of eritoran in nude mice inhibited liver cultures and in human adenocarcinoma cells, and it caused metastasis of colorectal cancer cells (54). Other studies in tumor regression in the mouse colon, which further supports human colorectal and esophageal cancer cell lines have indi- the hypothesis that changes in CD14/TLR4 antagonism exert cated that eritoran blocks LPS-induced TLR4 signaling for b1 anticancer effects. integrin– and selectin-dependent cell adhesion and that it Compared with the extensively studied TLR4 pathways, inhibits their migratory capability (54, 55). Furthermore, eri- signals mediated by CD14 are not well understood. Due to toran has previously been shown to inhibit monocytic inflam- its lack of a cytoplasmic tail, the membrane-bound CD14 has matory responses in endotoxemic rats and virus-infected mice traditionally been regarded as merely a binding component for (22, 29, 30). The possibility that the inhibitory effect of the transfer of LPS to TLR4/MD2. Other studies indicate eritoran on inflammatory cytokine production may help sup- that CD14 is also the binding component for bacterial cell press tumor malignancy cannot be ruled out. Overall, our wall products (e.g., LTA) for subsequent transfer to TLR2- findings demonstrate that eritoran may serve as a novel immu- mediated signals (43, 44). Similar to the findings in TLR4, notherapeutic agent for colorectal cancer, with a significant epithelial-specific deletion of TLR2 or MyD88 markedly advantage over other treatments because of its multiple modes reduced colon tumor formation (52). In the present study, we of action for inducing apoptosis and for its antiproliferative demonstrated that eritoran ablated MyD88 signaling and can- and anti-inflammatory effects. cer cell proliferation induced by LPS and LTA. The results In conclusion, the progression of bacterial LPS-induced suggested that eritoran suppressed tumorigenesis by inhibition colon cancer was inhibited by eritoran treatment through dual of TLR4- and TLR2-activated cell proliferation, partly through mechanisms involving the induction of CD14/Src/PKCz-medi- competitive binding to CD14. Furthermore, biochemical stud- ated apoptosis and the blockade of TLR4-dependent prolifer- ies have shown that LPS/CD14 binding initiates the production ation. Eritoran-induced changes in the functional antagonism of lipid messengers, such as sphingolipids and ceramide, and of epithelial LPS receptors represent a novel therapeutic the activation of PKCz prior to TLR4 recruitment to lipid rafts approach for the suppression of colorectal cancer. (9, 10). Our results from mouse colon tumors and from human cell lines demonstrate that eritoran is capable of binding to CD14 and triggering PKCz activation in a TLR4-independent Disclosure of Potential Conflicts of Interest manner. Moreover, Src kinase, a tyrosine kinase to phosphor- No potential conflicts of interest were disclosed. ylate Tyr 428 on PKCz, may also act upstream of its serine/ threonine phosphorylation sites. Our findings are in agreement Authors' Contributions with previous reports showing that Src kinase also phosphor- – Conception and design: T.-C. Lee, L.C.-H. Yu ylates 3-phosphoinositide dependent protein kinase-1 (PDK1; Development of methodology: W.-T. Kuo, T.-C. Lee, L.C.-H. Yu ref. 53), which is responsible for the phosphorylation of Acquisition of data (provided animals, acquired and managed patients, Thr410/560 in PKCz. Taken together, eritoran-induced CD14 provided facilities, etc.): W.-T. Kuo, T.-C. Lee, L.C.-H. Yu

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4693

Kuo et al.

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, Grant Support computational analysis): W.-T. Kuo, T.-C. Lee, L.C.-H. Yu This study was supported by grants from the Ministry of Science and Writing, review, and/or revision of the manuscript: W.-T. Kuo, T.-C. Lee, Technology (MoST 102-2628-B-002-009-MY3, 105-2811-B-002-014) and L.C.-H. Yu National Taiwan University (NTU-CDP-104R7798, 105R7798). Administrative, technical, or material support (i.e., reporting or organizing The costs of publication of this article were defrayed in part by the data, constructing databases): W.-T. Kuo, T.-C. Lee payment of page charges. This article must therefore be hereby marked Study supervision: L.C.-H. Yu advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Acknowledgments We thank the staff of the imaging and sequencing facility at the First Core Laboratory of National Taiwan University College of Medicine and the Disease Received January 21, 2016; revised May 13, 2016; accepted June 5, 2016; Animal Research Center for technical assistance. published OnlineFirst June 21, 2016.

References 1. Hanahan D, Weinberg RA. Hallmarks of cancer: The next generation. Cell 18. Tang X, Zhu Y. TLR4 signaling promotes immune escape of human colon 2011;144:646–74. cancer cells by inducing immunosuppressive cytokines and apoptosis 2. Huang CY, Yu LC. Pathophysiological mechanisms of death resistance in resistance. Oncol Res 2012;20:15–24. colorectal carcinoma. World J Gastroenterol 2015;21:11777–92. 19. Kuo WT, Lee TC, Yang HY, Chen CY, Au YC, Lu YZ, et al. LPS receptor 3. West AC, Jenkins BJ. Inflammatory and non-inflammatory roles for Toll- subunits have antagonistic roles in epithelial apoptosis and colonic car- like receptors in gastrointestinal cancer. Curr Pharm Des 2015;21: cinogenesis. Cell Death Differ 2015;22:1590–604. 2968–77. 20. Yu LCH, Flynn AN, Turner JR, Buret AG. SGLT-1-mediated glucose uptake 4. Sussman DA, Santaolalla R, Strobel S, Dheer R, Abreu MT. Cancer in protects intestinal epithelial cells against LPS-induced apoptosis and inflammatory bowel disease: Lessons from animal models. Curr Opin barrier defects: A novel cellular rescue mechanism? FASEB J 2005;19: Gastroenterol 2012;28:327–33. 1822–35. 5. Fukata M, Chen A, Vamadevan AS, Cohen J, Breglio K, Krishnareddy S, et al. 21.YuLCH,TurnerJR,BuretAG.LPS/CD14activationtriggersSGLT-1- Toll-like receptor-4 promotes the development of colitis-associated colo- mediated glucose uptake and cell rescue in intestinal epithelial cells via rectal tumors. Gastroenterology 2007;133:1869–81. early apoptotic signals upstream of caspase-3. Exp Cell Res 2006;312: 6. Wang EL, Qian ZR, Nakasono M, Tanahashi T, Yoshimoto K, Bando Y, et al. 3276–86. High expression of Toll-like receptor 4/myeloid differentiation factor 88 22. Mullarkey M, Rose JR, Bristol J, Kawata T, Kimura A, Kobayashi S, et al. signals correlates with poor prognosis in colorectal cancer. Br J Cancer Inhibition of endotoxin response by e5564, a novel Toll-like receptor 4- 2010;102:908–15. directed endotoxin antagonist. J Pharmacol Exp Ther 2003;304: 7. Doan HQ, Bowen KA, Jackson LA, Evers BM. Toll-like receptor 4 activation 1093–102. increases Akt phosphorylation in colon cancer cells. Anticancer Res 23. Opal SM, Laterre PF, Francois B, LaRosa SP, Angus DC, Mira JP, et al. Effect 2009;29:2473–8. of eritoran, an antagonist of MD2-TLR4, on mortality in patients with 8. Ortega-Cava CF, Ishihara S, Rumi MA, Kawashima K, Ishimura N, Kazu- severe sepsis: The ACCESS randomized trial. JAMA 2013;309:1154–62. mori H, et al. Strategic compartmentalization of Toll-like receptor 4 in the 24. Carpenter S, O'Neill LA. Recent insights into the structure of Toll-like mouse gut. J Immunol 2003;170:3977–85. receptors and post-translational modifications of their associated signal- 9. Cuschieri J, Billgren J, Maier RV. Phosphatidylcholine-specific phospholi- ling proteins. Biochem J 2009;422:1–10. pase C (PC-PLC) is required for LPS-mediated macrophage activation 25. Jiang Q, Akashi S, Miyake K, Petty HR. Lipopolysaccharide induces physical through CD14. J Leukoc Biol 2006;80:407–14. proximity between CD14 and toll-like receptor 4 (TLR4) prior to nuclear 10. Cuschieri J, Bulger E, Billgrin J, Garcia I, Maier RV. Acid sphingomyelinase is translocation of NF-kappa B. J Immunol 2000;165:3541–4. required for lipid Raft TLR4 complex formation. Surg Infect(Larchmt) 26. Park BS, Song DH, Kim HM, Choi BS, Lee H, Lee JO. The structural basis of 2007;8:91–106. lipopolysaccharide recognition by the TLR4-MD-2 complex. Nature 11. da Silva CJ, Soldau K, Christen U, Tobias PS, Ulevitch RJ. Lipopolysac- 2009;458:1191–5. charide is in close proximity to each of the proteins in its membrane 27. Kim HM, Park BS, Kim JI, Kim SE, Lee J, Oh SC, et al. Crystal structure of the receptor complex. transfer from CD14 to TLR4 and MD-2. J Biol Chem TLR4-MD-2 complex with bound endotoxin antagonist Eritoran. Cell 2001;276:21129–35. 2007;130:906–17. 12. Dunzendorfer S, Lee HK, Soldau K, Tobias PS. TLR4 is the signaling but 28. Yeh YC, Ko WJ, Chan KC, Fan SZ, Tsai JC, Cheng YJ, et al. Effects of eritoran not the lipopolysaccharide uptake receptor. J Immunol 2004;173: tetrasodium, a toll-like receptor 4 antagonist, on intestinal microcircula- 1166–70. tion in endotoxemic rats. Shock 2012;37:556–61. 13. Fukata M, Chen A, Klepper A, Krishnareddy S, Vamadevan AS, Thomas LS, 29. Shirey KA, Lai W, Scott AJ, Lipsky M, Mistry P, Pletneva LM, et al. The TLR4 et al. Cox-2 is regulated by Toll-like receptor-4 (TLR4) signaling: Role in antagonist Eritoran protects mice from lethal influenza infection. Nature proliferation and apoptosis in the intestine. Gastroenterology 2006;131: 2013;497:498–502. 862–77. 30. Rallabhandi P, Phillips RL, Boukhvalova MS, Pletneva LM, Shirey KA, 14. Santaolalla R, Sussman DA, Ruiz JR, Davies JM, Pastorini C, Espana CL, Gioannini TL, et al. Respiratory syncytial virus fusion protein-induced toll- et al. TLR4 activates the beta-catenin pathway to cause intestinal neoplasia. like receptor 4 (TLR4) signaling is inhibited by the TLR4 antagonists PLoS One 2013;8:e63298. Rhodobacter sphaeroides lipopolysaccharide and eritoran (E5564) and 15. Fukata M, Shang L, Santaolalla R, Sotolongo J, Pastorini C, Espana C, et al. requires direct interaction with MD-2. MBio 2012;3. Constitutive activation of epithelial TLR4 augments inflammatory 31. Huang CY, Hsiao JK, Lu YZ, Lee CT, Yu LCH. Anti-apoptotic PI3K/Akt responses to mucosal injury and drives colitis-associated tumorigenesis. signaling by sodium/glucose transporter 1 reduces epithelial barrier dam- Inflamm Bowel Dis 2011;17:1464–73. age and bacterial translocation in intestinal ischemia. Lab Invest 16. Ruemmele FM, Beaulieu JF, Dionne S, Levy E, Seidman EG, Cerf-Bensussan 2011;91294–309. N, et al. Lipopolysaccharide modulation of normal enterocyte turnover by 32. Huang CY, Kuo WT, Huang CY, Lee TC, Chen CT, Peng WH, et al. Distinct toll-like receptors is mediated by endogenously produced tumour necrosis cytoprotective roles of pyruvate and ATP by glucose metabolism on factor alpha. Gut 2002;51:842–8. epithelial necroptosis and crypt proliferation in ischemic gut. J Physiol 17. Grondin V, Seksik P, Dumont S, Thomas G, Trugnan G, Flejou JF, et al. 2016 Apr 28. [Epub ahead of print]. Regulation of colon cancer cell proliferation and migration by MD-2 33. Sato T, Stange DE, Ferrante M, Vries RG, Van Es JH, Van den Brink S, et al. activity. Innate Immun 2011;17:414–22. Long-term expansion of epithelial organoids from human colon,

4694 Cancer Res; 76(16) August 15, 2016 Cancer Research

Tumor Progression by Dysregulated LPS Signaling

adenoma, adenocarcinoma, and Barrett's epithelium. Gastroenterology 45. Eyking A, Ey B, Runzi M, Roig AI, Reis H, Schmid KW, et al. Toll-like 2011;141:1762–72. receptor 4 variant D299G induces features of neoplastic progression in 34. Chen Y, Lee SH, Tsai YH, Tseng SH. Ischemic preconditioning increased the Caco-2 intestinal cells and is associated with advanced human colon intestinal stem cell activities in the intestinal crypts in mice. J Surg Res cancer. Gastroenterology 2011;141:2154–65. 2014;187:85–93. 46. Fukata M, Hernandez Y, Conduah D, Cohen J, Chen A, Breglio K, et al. 35. Huang CY, Kuo WT, Huang YC, Lee TC, Yu LC. Resistance to hypoxia- Innate immune signaling by Toll-like receptor-4 (TLR4) shapes the inflam- induced necroptosis is conferred by glycolytic pyruvate scavenging of matory microenvironment in colitis-associated tumors. Inflamm Bowel mitochondrial superoxide in colorectal cancer cells. Cell Death Dis Dis 2009;15:997–1006. 2013;4:e622. 47. HernandezY,SotolongoJ,BreglioK,ConduahD,ChenA,XuR,etal. 36. Hormann N, Brandao I, Jackel S, Ens N, Lillich M, Walter U, et al. Gut The role of prostaglandin E2 (PGE 2) in toll-like receptor 4 (TLR4)- microbial colonization orchestrates TLR2 expression, signaling and epi- mediated colitis-associated neoplasia. BMC Gastroenterol 2010; thelial proliferation in the small intestinal mucosa. PLoS One 2014;9: 10:82. e113080. 48. Tang XY, Zhu YQ, Wei B, Wang H. Expression and functional research of 37. Hirai T, Chida K. Protein kinase Czeta (PKCzeta): Activation mechanisms TLR4 in human colon carcinoma. Am J Med Sci 2010;339:319–26. and cellular functions. J Biochem 2003;133:1–7. 49. Nenci A, Becker C, Wullaert A, Gareus R, van LG, Danese S, et al. Epithelial 38. Nishikawa K, Toker A, Johannes FJ, Songyang Z, Cantley LC. Determination NEMO links innate immunity to chronic intestinal inflammation. Nature of the specific substrate sequence motifs of protein kinase C isozymes. J Biol 2007;446:557–61. Chem 1997;272:952–60. 50. Edelblum KL, Washington MK, Koyama T, Robine S, Baccarini M, Polk DB. 39. Ranganathan S, Wang Y, Kern FG, Qu Z, Li R. Activation loop phosphor- Raf protects against colitis by promoting mouse colon epithelial cell ylation-independent kinase activity of human protein kinase C zeta. survival through NF-kappaB. Gastroenterology 2008;135:539–51. Proteins 2007;67:709–19. 51. Greten FR, Eckmann L, Greten TF, Park JM, Li ZW, Egan LJ, et al. IKKbeta 40. Martin-Villa JM, Ferre-Lopez S, Lopez-Suarez JC, Corell A, Perez-Blas M, links inflammation and tumorigenesis in a mouse model of colitis-asso- rnaiz-Villena A. Cell surface phenotype and ultramicroscopic analysis of ciated cancer. Cell 2004;118:285–96. purified human enterocytes: A possible antigen-presenting cell in the 52. Scheeren FA, Kuo AH, van Weele LJ, Cai S, Glykofridis I, Sikandar SS, et al. A intestine. Tissue Antigens 1997;50:586–92. cell-intrinsic role for TLR2-MYD88 in intestinal and breast epithelia and 41. Schroder NW, Morath S, Alexander C, Hamann L, Hartung T, Zah- oncogenesis. Nat Cell Biol 2014;16:1238–48. ringer U, et al. Lipoteichoic acid (LTA) of Streptococcus pneumoniae 53. Yang KJ, Shin S, Piao L, Shin E, Li Y, Park KA, et al. Regulation of 3- and Staphylococcus aureus activates immune cells via Toll-like recep- phosphoinositide-dependent protein kinase-1 (PDK1) by Src involves tor (TLR)-2, lipopolysaccharide-binding protein (LBP), and CD14, tyrosine phosphorylation of PDK1 and Src homology 2 domain binding. whereas TLR-4 and MD-2 are not involved. J Biol Chem 2003;278: J Biol Chem 2008;283:1480–91. 15587–94. 54. Hsu RY, Chan CH, Spicer JD, Rousseau MC, Giannias B, Rousseau S, et al. 42. Xin C, Zhang H, Liu Z. miR-154 suppresses colorectal cancer cell growth LPS-induced TLR4 signaling in human colorectal cancer cells increases and motility by targeting TLR2. Mol Cell Biochem 2014;387:271–7. beta1 integrin-mediated cell adhesion and liver metastasis. Cancer Res 43. Dziarski R, Tapping RI, Tobias PS. Binding of bacterial peptidoglycan to 2011;71:1989–98. CD14. J Biol Chem 1998;273:8680–90. 55. Rousseau MC, Hsu RY, Spicer JD, McDonald B, Chan CH, Perera RM, et al. 44. Nilsen NJ, Deininger S, Nonstad U, Skjeldal F, Husebye H, Rodionov D, et al. Lipopolysaccharide-induced toll-like receptor 4 signaling enhances the Cellular trafficking of lipoteichoic acid and Toll-like receptor 2 in relation to migratory ability of human esophageal cancer cells in a selectin-dependent signaling: Role of CD14 and CD36. J Leukoc Biol 2008;84:280–91. manner. Surgery 2013;154:69–77.

www.aacrjournals.org Cancer Res; 76(16) August 15, 2016 4695

Eritoran Suppresses Colon Cancer by Altering a Functional Balance in Toll-like Receptors That Bind Lipopolysaccharide

Wei-Ting Kuo, Tsung-Chun Lee and Linda Chia-Hui Yu

Cancer Res 2016;76:4684-4695. Published OnlineFirst June 21, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-16-0172

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2016/06/21/0008-5472.CAN-16-0172.DC1

Cited articles This article cites 53 articles, 13 of which you can access for free at: http://cancerres.aacrjournals.org/content/76/16/4684.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/76/16/4684.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/76/16/4684. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.