Preventing tumor neovascularisation is one of the strategies recently developed to limit the dissemination of cancer cells and apparition of metastases. Although these approaches could improve the existing treatments, a number of unexpected negative effects have been reported, mainly linked to the hypoxic condition and the subsequent induction of the pro-oncogenic hypoxia inducible factor(s) resulting from cancer cells’ oxygen starvation. Here, we checked in vivo on colon cancer cells an alternative approach. It is based on treatment with myo-inositol trispyrophosphate (ITPP), a molecule that leads to increased oxygenation of tumors. We provide evidence that ITPP increases the survival of mice in a model of carcinomatosis of human colon cancer cells implanted into the peritoneal cavity. ITPP also reduced the growth of subcutaneous colon cancer cells xenografted in nu/nu mice. In the subcutaneous tumors, ITPP stimulated the expression of the homeobox gene Cdx2 that is crucial for intestinal differentiation and that also has an anti-tumoral function. On this basis, human colon cancer cells were cultured in vitro in hypoxic conditions. Hypoxia was shown to decrease the level of Cdx2 protein, mRNA and the activity of the Cdx2 promoter. This decline was unrelated to the activation of HIF1α and HIF2α by hypoxia. However, it resulted from the activation of a phosphatidylinositol 3-kinases-like mitogen-activated protein kinase pathway, as assessed by the fact that LY294002 and U0126 restored high Cdx2 expression in hypoxia. Corroborating these results, U0126 recapitulated the increase of Cdx2 triggered by ITPP in subcutaneous colon tumor xenografts. The present study provides evidence that a chemical compound that increases oxygen pressure can antagonize the hypoxic setting and reduce the growth of human colon tumors implanted in nu/nu mice.
Hypoxia is an aggravating factor in cancer that stimulates angiogenesis and consequently the intake of nutrients for tumor growth, and that also opens routes for invasive cells to disseminate out of the primary tumor to form metastases. On this basis, chemotherapy with drugs targeting the vascular endothelial growth factor pathway provides some benefit for the patients. However, a number of unexpected limitations have been encountered, in particular related to the fact that inhibiting vessel formation in tumors starves malignant cells from nutrients and oxygen, which subsequently activates the hypoxia response pathway(s) and the pro-oncogenic hypoxia inducible factor-1 (HIF1).1 We have previously described myo-inositol trispyrophosphate (ITPP), a molecule that increases the release of bound dioxygen from haemoglobin in vitro and improves oxygen tension under hypoxic conditions in vivo.2 This molecule ameliorates the exercise capacity of transgenic mice with severe heartfailure3 and inhibits angiogenesis in a chorioallantoid membrane model.4 Of note, when used on cancer models, ITPP was shown to reduce the growth of xenografted glioma and leads to the eradication of early liver tumors,4, 5 opening the possibility that increasing the oxygen load could represent a contrasting alternative to anti-vascular endothelial growth factor therapy in cancer. Colorectal cancer is the third cause of death by cancer and one of the leading types of cancer in which anti-angiogenic therapy is being evaluated. Here, we report beneficial context-dependent effects of ITPP in colon cancer models. We further show that the intestine-specific homeobox gene, Cdx2, which is crucial for homeostasis of the gut epithelium6 and also exhibits anti-tumor activity,7, 8 is a downregulated target of hypoxia in colon cancer cells, while being upregulated by ITPP.
Results and discussion
To address whether ITPP has any effect on colon tumour growth, we first used the intraperitoneal carcinomatosis model of HT29 cells implanted in the abdominal cavity of athymic nu/nu mice.9 Two weeks after the injection of 107 cells, a series of 10 mice were treated weekly with ITPP at 1.5 or 2 mg/g body weight or saline buffer. As a positive control for drug therapy, one series of mice was treated with Capecitabine at 0.25 mg/g body weight. The mortality was evaluated during 9 weeks of treatment and surviving mice were then euthanized and autopsied. Both doses of ITPP reduced the mortality as compared to NaCl, slightly better than Capecitabine (Figure 1a). At the end of the treatment two mice survived in the NaCl group, whereas five, five and four mice survived in the groups treated, respectively, with ITPP 1.5 mg/g, ITPP 2 mg/g and Capecitabine 0.25 mg/g. One of the two NaCl-treated mice was tumor free, but this ratio rose to 3/4 in Capecitabine-treated mice and even to 4/5 and 5/5 in mice treated with ITPP at 1.5 and 2 mg/g. These data indicate that ITPP antagonizes colon cancer growth in the model of intraperitoneal carcinomatosis.
Next, 2 × 106 HT29 cells were engrafted under the skin of nu/nu mice to follow up the tumor size during 5 weeks in the model of subcutaneous tumors. ITPP was injected weekly from the time of grafting or starting 2 weeks after cell implantation. Oxygen tension was monitored to assess the activity of ITPP. In untreated animals, it was 30 mm Hg in muscles but only 0.2–2 mm Hg in tumors, signalling their hypoxic setting; 2 h after ITPP administration, oxygen tension remained at 30–34 mm Hg in muscles; however, it strongly increased up to 32 mm Hg in the tumors and persisted above 25 mm Hg 2 days after ITPP administration. Together, this confirmed that ITPP increased the oxygen release inside tumors. As shown in Figure 1b, ITPP significantly reduced the growth of the subcutaneous HT29 tumors, leading to a twofold reduction of the tumor volume. These results are in line with those obtained with the intraperitoneal carcinomatosis model. To evaluate the effect of ITPP on other colon cancer cell lines than HT29, SW480 and HCT116 cells were grafted under the skin of nu/nu mice. As for HT29 cells, SW480 tumor growth was reduced by ITPP; however, the drug was much less efficient on HCT116 tumors (Figure 1c). Interestingly, unlike HT29 and SW480 cells in which the hypoxia-induced factor HIF1α is extremely low in normoxia, HCT116 cells exhibit a significant constitutive expression of HIF1α already in normoxia (Figure 1d). Thus, ITPP treatment, which increases oxygen tension in tumors, reduces tumor growth, but not in a cell line in which HIF1α is already expressed in normoxia.
Histological examination of HT29 tumors grown for 5 weeks showed a tendency towards more differentiation under ITTP treatment. The intestine-specific homeobox gene Cdx2 is a major regulator of intestinal differentiation6, 10 whose expression is altered in human colorectal cancers. Cdx2 reduction in colon cancer results in accelerated tumor progression and increased tumor cell migration and dissemination.7, 8 We have previously reported that although Cdx2 is very low in HT29 cells cultured in vitro, these cells are competent for the expression of this homeobox gene, as it can be turned on when they are grafted subcutaneously in nu/nu mice.11 These studies have emphasized the relevance of cell interactions between cancer cells and subjacent fibroblasts to stimulate Cdx2,11 an effect invoving extracellular matrix proteins like laminins.12, 13 However, until now the expression of Cdx2 was not analysed with regard to the oxygenation status. We therefore addressed whether the ITPP treatment has any effect on Cdx2 expression in tumors grafted for 3, 4 and 5 weeks in nu/nu mice. Immunohistochemistry was used to estimate the proportion of cell nuclei expressing the Cdx2 protein and RNA levels were quantified by reverse transcription–qPCR. The results illustrated in Figure 1e show that ITPP accelerates the rise of Cdx2 expression in xenografted HT29 tumors.
The above results obtained in vivo suggest that Cdx2 expression might be downregulated by hypoxia. We therefore, addressed this issue in cell lines cultured in vitro. For this purpose, SW480 and Caco2-TC7 cells, two human colon cancer cell lines that endogenously express this homeobox gene, were cultured under hypoxic conditions (3% O2). Figures 2a and b show that hypoxia caused a decrease of Cdx2 mRNA and Cdx2 protein. An even stronger effect was observed when cells were cultured in 1% oxygen, but this setting compromised cell survival. Hypoxia exerted its effect on gene transcription, as it reduced the activity of a reporter luciferase plasmid containing a 9.3-kb genomic fragment of the Cdx2 promoter14 (Figure 2c). HIF1α and HIF2α are major transcription factors known to mediate many of the effects of hypoxia. Knockingdown HIF1α or HIF2α or both HIF1α and HIF2α together using specific small interfering RNA did not prevent the decrease of Cdx2 in SW480 cells cultured in hypoxia (Figure 2d), although the efficacy of the knockdown was assessed by the inhibition of HIF1α or HIF2α production by the respective small interfering RNA in cells cultured in hypoxia (Figure 2d) and by the resulting inhibition of three known transcriptional targets of the HIF pathway: DDIT4, NDRG1 and SLC2A3 (Figure 2f). Conversely, when SW480 cells were cultured in normoxia and transfected with increasing amounts of HIF1α-encoding plasmid to saturate the physiological degradation pathway, the resulting overexpression of HIF1α failed to decrease Cdx2 levels (Figure 3e). Together, these data indicate that the expression of Cdx2 is a new target of hypoxia in colorectal cancer cells and that its downregulation by hypoxia is HIF-independent.
We used pharmacological drugs to decipher the pathway(s) involved in the hypoxia-induced decrease of Cdx2. SW480 cells were cultured in normoxia for 24 h, then the drugs were added and the cells were placed in hypoxia for 24 h in the continuous presence of drugs. Cdx2 mRNA was quantified by reverse transcription–qPCR. The inhibitors were Wortmanin and LY294002 for pan-phosphatidylinositol 3-kinase pathways, Sorafenib for RAF, U0126 for mitogen-activated protein kinase, Tetrabromobenzotriazole for Casein Kinase II, rapamycin for mTOR, Tricostatin-A for histone deacetylases and irinotecan for topoisomerase-1; we also used Forskolin to activate adenylate cyclase and the antioxidant N-acetyl-L-cysteine against ROS. Figure 3a indicates that LY294002 and U0126 prevented the Cdx2 decrease triggered by hypoxia, whereas all other compounds were without effect. Adding Sorafenib, Irinotecan or N-acetyl-L-cysteine did not improve the effect of LY294002 or U0126 on Cdx2. In addition, combining LY294009 and U0126 had the same effect as each inhibitor independently, suggesting that both target the same pathway.
The effect of LY294009 and U0126 on the hypoxia-induced downregulation of Cdx2 is consistent with the fact that both Akt and Erk are activated in HT29 and SW480 cells placed in hypoxia (Figures 3b and c). On the basis of these data obtained in cell lines cultured in vitro, we wondered whether the MEK inhibitor U0126 could also recapitulate the effect of ITPP on Cdx2 expression in xenografted tumors. For this purpose, HT29 cells were grafted subcutaneously in nu/nu mice and the mice were treated either with U0126 at 25 μmol/g of body weight or with saline buffer as control. As illustrated in Figure 3d, inhibiting MEK by U0126 in these hypoxic tumors significantly stimulated Cdx2 expression, as shown previously with the ITPP treatment in vivo.
In conclusion, this study provides evidence that a chemical compound that increases tumor oxygen tension improves the hypoxic setting and reduces the growth of human colon tumors implanted in nu/nu mice, except in a cell line where HIF1α is already expressed in normoxia. Although hypoxia has many more effects, we report here for the first time that it downregulates the intestine-specific homeobox gene Cdx2. As this gene exerts an anti-tumor function through multiple molecular mechanisms including the transcriptional stimulation of a Wnt signalling antagonist, Mucdhl15 and the transcription-independent reduction of NHEJ DNA repair activity,16 its restoration is expected to participate in the beneficial outcome resulting from ITPP treatment. Thus, restoring normoxia in tumors could be an alternative way to anti-vascular endothelial growth factor therapies that deserves further investigations and evaluation in preclinical models of cancer, in particular in combination with cytotoxic drugs currently used in chemotherapy.
Loges S, Mazzone M, Hohensinner P, Carmeliet P . Silencing or fueling metastasis with VEGF inhibitors: antiangiogenesis revisited. Cancer Cell 2009; 15: 167–170.
Kieda C, Greferath R, Crola da Silva C, Fylaktakidou KC, Lehn JM, Nicolau C . Suppression of hypoxia-induced Hif-1alpha and of angiogenesis in endothelial cells by myo-inositol trispyrophosphate-treated erythrocytes. Proc Natl Acad Sci USA 2006; 103: 15576–15581.
Biolo A, Greferath R, Siwik DA, Qin F, Valsky E, Fylaktakidou KC et al. Enhanced exercise capacity in mice with severe heart failure treated with an allosteric effector of hemoglobin, myo-inositol trispyrophosphate. Proc Natl Acad Sci USA 2009; 106: 1926–1929.
Sihn G, Walter T, Klein JC, Queguiner I, Iwao H, Nicolau C et al. Anti-angiogenic properties of myo-inositol trispyrophosphate in ovo and growth reduction of implanted glioma. FEBS Lett 2007; 581: 962–966.
Aprahamian M, Bour G, Akladios CY, Fylaktakidou K, Greferath R, Soler L et al. Myo-inositoltrispyrophosphate treatment leads to hif-1alpha suppression and eradication of early hepatoma tumors in rats. Chembiochem 2011; 12: 777–783.
Stringer EJ, Duluc I, Saandi T, Davidson I, Bialecka M, Sato T et al. Cdx2 determines the fate of postnatal endoderm. Development 2012; 139: 465–474.
Bonhomme C, Duluc I, Martin E, Chawengsaksophak K, Chenard MP, Kedinger M et al. The Cdx2 homeobox gene has a tumor suppressor function in the distal colon in addition to a homeotic role during gut development. Gut 2003; 52: 1465–1472.
Gross I, Duluc I, Benameur T, Calon A, Martin E, Brabletz T et al. The intestine-specific homeobox gene Cdx2 decreases mobility and antagonizes dissemination of colon cancer cells. Oncogene 2008; 27: 107–115.
Pearson JW, Fitzgerald D, Willingham MC, Wiltrout RH, Pastan I, Longo DL . Chemoimmunotoxin against a human colon tumor (HT29) xenografted in nude mice. Cancer Res 1989; 49: 3562–3567.
Verzi MP, Shin H, He HH, Sulahian R, Meyer CA, Montgomery RK et al. Differentiation-specific histone modifications reveal dynamic chromatin interactions and partners for the intestinal transcription factor CDX2. Dev Cell 2010; 19: 713–726.
Benahmed F, Gross I, Guenot D, Jehan F, Martin E, Domon-Dell C et al. The microenvironment controls CDX2 homeobox gene expression in colorectal cancer cells. Am J Pathol 2007; 170: 733–744.
Lorentz O, Duluc I, De Arcangelis A, Simon-Assmann P, Kedinger M, Freund JN . Key role of the Cdx2 homeobox gene in extracellular matrix-mediated intestinal cell differentiation. J Cell Biol 1997; 139: 1553–1569.
Turk N, Gross I, Gendry P, Stutzmann J, Freund JN, Kedinger M et al. Laminin isoforms: biological roles and effects on the intracellular distribution of nuclear proteins in intestinal epithelial cells. Exp Cell Res 2005; 303: 494–503.
Benahmed F, Gross I, Gaunt SJ, Beck F, Jehan F, Domon-Dell C et al. Multiple regulatory regions control the complex expression pattern of the mouse Cdx2 homeobox gene. Gastroenterology 2008; 135: 1238–1247.
Hinkel I, Duluc I, Martin E, Guenot D, Freund JN, Gross I . Cdx2 controls expression of the protocadherin Mucdhl, an inhibitor of growth and β-catenin activity in colon cancer cells. Gastroenterology 2012; 142: 875–885.
Renouf B, Soret C, Saandi T, Delalande F, Martin E, Vanier M et al. Cdx2 homeoprotein inhibits non-homologous end joining in colon cancer but not in leukemia cells. Nucl Acids Res 2012; 40: 3456–3469.
Pencreach E, Guerin E, Nicolet C, Lelong-Rebel I, Voegeli AC, Oudet P et al. Marked activity of irinotecan and rapamycin combination toward colon cancer cells in vivo and in vitro is mediated through cooperative modulation of the mammalian target of rapamycin/hypoxia-inducible factor-1alpha axis. Clin Cancer Res 2009; 15: 1297–1307.
Gross I, Lhermitte B, Domon-Dell C, Duluc I, Martin E, Gaiddon C et al. Phosphorylation of the homeotic tumor suppressor Cdx2 mediates its ubiquitin-dependent proteasome degradation. Oncogene 2005; 24: 7955–7963.
Marampon F, Bossi G, Ciccarelli C, Di Rocco A, Sacchi A, Pestell RG et al. MEK/ERK inhibitor U0126 affects in vitro and in vivo growth of embryonal rhabdomyosarcoma. Mol Cancer Ther 2009; 8: 543–551.
This work was supported by the INSERM (France), NormOxys Inc (USA) and the Ligue contre la Cancer du Haut-Rhin (France). Ms Saandi was supported by the AICR (UK) and the Département of Mayotte (France).
Drs Nicolau and Lehn have received compensation as members of the scientific advisory board and own stock in NormOxys Inc, which holds the patents on the applications of inositol trispyrrophosphate. Drs Greferath and Tufa have consulted for NormOxys Inc and have received compensations. Ms Derbal-Wolfrom was a recipient of a funding by NormOxys Inc. Ms Martin and Drs Aprahamian, Pencreach, Choquet, Duluc and Freund declare no potential conflict of interest.
About this article
Cite this article
Derbal-Wolfrom, L., Pencreach, E., Saandi, T. et al. Increasing the oxygen load by treatment with myo-inositol trispyrophosphate reduces growth of colon cancer and modulates the intestine homeobox gene Cdx2. Oncogene 32, 4313–4318 (2013). https://doi.org/10.1038/onc.2012.445
- colon cancer
- MAP kinase
- homeobox gene
Interprofessional and interdisciplinary collaboration for early phase oncological clinical trials in academia—Myo-inositoltrispyrophophate as model
Pharmacological Research (2020)
Impact of myo -inositol trispyrophosphate (ITPP) on tumour oxygenation and response to irradiation in rodent tumour models
Journal of Cellular and Molecular Medicine (2019)
Quantum Blue Reduces the Severity of Woody Breast Myopathy via Modulation of Oxygen Homeostasis-Related Genes in Broiler Chickens
Frontiers in Physiology (2019)
Frontiers in Endocrinology (2019)