Supported by the Korea Healthcare technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A100277) and by the Science Research Center (Bone Metabolism Research Center) funded by the Korean Ministry of Education, Science and Technology (2009-0063267).
Potential conflict of interest: Nothing to report.
Chaetocin, an antibiotic produced by Chaetomium species fungi, was recently found to have antimyeloma activity. Here we examined whether chaetocin has anticancer activities against solid tumors. Chaetocin inhibited the growth of mouse and human hepatoma grafts in nude mice. Immunohistochemical analyses revealed that chaetocin inhibits hypoxia-inducible factor-1α (HIF-1α) expression and vessel formation in the tumors. Chaetocin also showed antiangiogenic anticancer activities in HIF-1α(+/+) fibrosarcoma grafted in mice, but not in HIF-1α(−/−) fibrosarcoma. Biochemical analyses showed that chaetocin down-regulated HIF-1α and the transcripts of HIF-1 target genes including vascular endothelial growth factor in hepatoma tissues and in various hepatoma cell lines. Based on the reported literature, unsuccessful efforts were made to determine the mechanism underlying the action of chaetocin. Unexpectedly, chaetocin was found to cause the accumulation of HIF-1α premessenger RNA (pre-mRNA) but to reduce mature mRNA levels in hepatoma cells and tissues. Such an effect of chaetocin was not observed in cell lines derived from normal cells, and was cell type-dependent even among cancer cell lines. Conclusions: Our results suggest that chaetocin could be developed as an anticancer agent to target HIF-1 in some cancers including hepatoma. It is also suggested that the HIF-1α pre-mRNA splicing is a novel therapeutic target for controlling HIF-1-mediated pathological processes. (HEPATOLOGY 2011;.)
Hepatocellular carcinoma (also called malignant hepatoma) is one of the most common malignant tumors and the third leading cause of cancer mortality worldwide.1 Despite many efforts to develop various classes of agents, systemic chemotherapy and hormone therapy have failed to significantly increase the survival of patients with advanced hepatoma. However, recent advances in the understanding of hepatoma progression have led to the development of novel molecularly targeted therapies.2 Because angiogenesis is pivotal for the development and progression of hepatoma, key molecules regulating angiogenesis are regarded as promising targets for treating hepatoma.3
Hypoxia inevitably develops in rapidly growing tumors and is an important microenvironment that forces changes in tumor behavior. In particular, hypoxia activates hypoxia-inducible factor-1α (HIF-1α), which promotes the progression of malignancy by stimulating angiogenesis and by augmenting the ability of tumors to survive.4, 5 The roles of HIF-1α have been extensively investigated in cancer patients and in tumor-bearing mice.6, 7 Consequently, HIF-1α is believed to be a valid target for the treatment of aggressive tumors, and many efforts have been made to identify suitable HIF-1α inhibitors.8
Chaetocin, which is produced by Chaetomium sp., is an antibiotic having the thiodioxopiperazine structure (a disulfide-bridged piperazine).9 Other thiodioxopiperazines are known to have antimicrobial, antiviral, immunosuppressive, and antiinflammatory activities,10, 11 but the biological activity of chaetocin has been reported in relatively few reports. In one such report, it was suggested that chaetocin inhibits the histone methyltransferase suv39H1.12 Recently, it was also demonstrated that chaetocin induces apoptosis of myeloma cells and retards the growth of myeloma xenografts.13 Mechanistically, it was proposed that chaetocin produces oxidative damage in myeloma cells by inhibiting the antioxidant enzyme thioredoxin reductase.14 However, little is known about the effects of chaetocin on solid tumors, and thus we tested the anticancer activity of chaetocin against solid tumors. Here we demonstrated that chaetocin has antiangiogenic and anticancer activities in hepatoma and fibrosarcoma grafts, and that these actions of chaetocin are due to HIF-1α down-regulation caused by the deregulation of HIF-1α premessenger RNA (pre-mRNA) splicing.
Hep3B hepatoma, HepG2 hepatoma, Huh7 hepatoma, A549 lung carcinoma, HCT116 colon carcinoma, MCF7 mammary carcinoma, and Hepa 1c1c-7 mouse hepatoma cell lines were obtained from the American Type Culture Collection (Manassas, VA), and von Hippel-Lindau (VHL)(+/+; −/−) renal cell carcinoma 4 (RCC4) from the European Collection of Cell Cultures (London, UK). HIF-1α(+/+; −/−) mouse embryonic fibroblast cell lines were kindly provided by Dr. Randall Johnson (University of California, San Diego). Cells were cultured in α-modified Eagle medium or in Dulbecco's Modified Eagle's medium, both of which were supplemented with 10% heat-inactivated fetal calf serum, in a 5% CO2 humidified atmosphere at 37°C. The oxygen tension in the chamber was either 20% (normoxic) or 1% (hypoxic). Chaetocin and other chemicals were administered to medium 1 hour before normoxic or hypoxic incubation.
Animals and Tumor Grafts.
Male nude mice (BALB/cAnNCrj-n/n) were purchased from Charles River Japan (Shin-Yokohama, Japan) and housed in a specific pathogen-free room. Mice (6 weeks old) were injected subcutaneously in a flask with 5 × 106 viable cancer cells. After tumor volumes reached 100-150 mm3, mice were treated with dimethyl sulfoxide (DMSO), chaetocin (0.25 mg/kg, intraperitoneally [i.p.]), and/or doxorubicin (1 mg/kg, i.p.) once a day. Tumor volumes were measured with a caliper and calculated using the equation volume = ab2/2, where “a” is the maximal width and “b” is maximal orthogonal width. All procedures were conducted in accordance with the guidelines of the Laboratory Animal Ethics Committee of Seoul National University.
All data were analyzed using Microsoft Excel 2007 or SPSS (v. 10.0) software and results are expressed as means and 95% confidence intervals or standard deviations. The two-sided Mann-Whitney U test was used to compare luciferase activities, vascular endothelial growth factor (VEGF) concentrations, and HIF-1α-positive cell, CD31-positive vessel, and Transferase-Mediated dUTP Nick-End Labeling (TUNEL)-positive cell numbers. Tumor sizes were compared using analysis of variance (ANOVA) followed by Duncan's multiple range test. Differences were considered significant for P < 0.05. All statistical tests were two-sided.
Chaetocin Has Antiangiogenic and Anticancer Activities Against Hepatoma Grafts.
To examine if chaetocin has anticancer activity against hepatoma, chaetocin was injected for 2 weeks into mice bearing Hepa 1c1c-7 tumors. Tumor growth was significantly retarded after chaetocin treatment (Fig. 1A). Interestingly, chaetocin did not induce massive cell death or deformation (Fig. 1B, upper), suggesting that cytotoxicity does not primarily underlie the anticancer effect. Moreover, neither histological changes nor apoptosis was observed in the livers of mice treated with chaetocin for 2 weeks (Fig. 1B, lower). Because a hypoxic microenvironment and angiogenesis are critical for tumor growth, we assessed HIF-1α expression and vascular density immunohistochemically. In chaetocin-treated tumors, HIF-1α-positive cells and CD31-positive threadlike vessels were noticeably reduced, but TUNEL-positive apoptotic cells increased (Fig. 1B, upper); the results are summarized in Fig. 1C. Furthermore, VEGF mRNA was down-regulated overall in chaetocin-treated tumors (Fig. 1D). These results strongly suggest that chaetocin inhibits tumor growth by deregulating HIF-1α-mediated angiogenesis. Chaetocin also inhibited the hypoxic inductions of HIF-1α protein and VEGF mRNA in Hepa 1c1c-7 cells cultured in vitro (Fig. 1E).
Anticancer Effect of Chaetocin Depends on HIF-1α.
To confirm that the anticancer effect of chaetocin is due to HIF-1α suppression, we injected HIF-1α(+/+) or (−/−) mouse embryonic fibroblast (MEF) cells into the flanks of nude mice to establish fibrosarcomas. Chaetocin inhibited the growth of HIF-1α(+/+) fibrosarcoma, but not HIF-1α(−/−) fibrosarcoma (Fig. 2A, Supporting Information Fig. 1A). In HIF-1α(+/+) tumors, HIF-1α expression and vascular formation were reduced and apoptosis was induced by chaetocin (Fig. 2B, Supporting Information Fig. 1B). Chaetocin attenuated the hypoxic induction of HIF-1α and VEGF in HIF-1α(+/+) MEF cells, but not in HIF-1α(−/−) MEF cells (Fig. 2C). These results indicate that the antiangiogenic and anticancer effects of chaetocin are due to its inhibition of HIF-1α. To determine whether chaetocin interferes with physiological responses to hypoxia, we analyzed erythropoietin (EPO) mRNA levels in the kidneys of mice that had been subjected to hypoxia (10% O2). Even after chaetocin treatment for 7 days, the hypoxic induction of renal EPO was not attenuated, which suggested that chaetocin has a tumor-specific action (Supporting Information Fig. 1C).
Chaetocin Inhibits HIF-1-Mediated Hypoxic Responses in Human Cancer Cells.
The hypoxic induction of HIF-1α was attenuated by chaetocin in human hepatoma cell lines (Fig. 3A). We also examined whether the HIF-2α isoform compensates for HIF-1α suppression by chaetocin. HIF-2α was also slightly suppressed by chaetocin (Fig. 3A), which suggests that HIF-1α inhibition is uncompensated. As compared with hepatoma cell lines, other cancer cells, such as, HCT116, MCF7, and A549, showed less or no response to chaetocin at 100 nM (its effective concentration in hepatoma cells) (Fig. 3B). A higher concentration (500 nM) of chaetocin was required to inhibit HIF-1α substantially in these cells (Supporting Information Fig. 2A), indicating that sensitivity to chaetocin may be cell type-dependent. To examine the effect of chaetocin on cell viability, we treated Hep3B and HepG2 cells with various doses of chaetocin in 20% or 1% O2 atmospheres for 24 or 48 hours. However, cell viabilities were unaffected by chaetocin in the concentration range that effectively inhibited HIF-1α (Supporting Information Fig. 3A), but when cells were subjected to severe hypoxia (0.1% O2 for 48 hours), chaetocin at ≥100 nM significantly reduced cell viabilities (Supporting Information Fig. 3B). EPO-enhancer and VEGF-promoter reporters were activated in hypoxia, which was inhibited by chaetocin (Fig. 3C, Supporting Information Fig. 4A). In Hep3B and HepG2, the hypoxic inductions of HIF-1 target mRNAs (VEGF, pyruvate dehydrogenase kinase 1 [PDK1], carbonic anhydrase 9 [CA9], and EPO) and VEGF protein were attenuated by chaetocin (Fig. 3D,E). Because HIF-1α facilitates adenosine triphosphate (ATP) production through glycolysis, we examined whether chaetocin blocks ATP generation in Hep3B cells. Chaetocin decreased intracellular ATP levels under both normoxic and hypoxic conditions (Supporting Information Fig. 4B). In addition, it attenuated the productions of pyruvate and lactate during hypoxia (Supporting Information Fig. 4C,D). These results suggest that chaetocin blocks glycolytic ATP production due to HIF-1α suppression. We also confirmed that chaetocins obtained from three different sources have similar effects on HIF-1 activity and HIF-1α expression (Supporting Information Fig. 5A,B).
Chaetocin Inhibits the De Novo Synthesis of HIF-1α Protein.
We next studied the mechanism by which chaetocin down-regulates HIF-1α. We first examined whether Suv39H1, oxidative stress, or thioredoxin reductase-1 is involved in HIF-1α suppression.12-14 However, HIF-1α expression and the chaetocin effect occurred regardless of these factors (Fig. 4A-C). A novel mechanism might underlie HIF-1α suppression by chaetocin and, thus, we investigated the mechanism stepwise. Initially, we examined if chaetocin destabilizes HIF-1α protein. As a consequence, the oxygen-dependent degradation of HIF-1α was not altered by chaetocin (Fig. 4D). Even after HIF-1α had been oxygen-independently stabilized by MG132 (proteasome inhibitor) or desferrioxamine (prolyl-hydroxylase domain protein [PHD] inhibitor), chaetocin suppressed HIF-1α (Fig. 4E, Supporting Information Fig. 6A). Given that pVHL and p53 facilitate HIF-1α degradation,15 we checked the effect of chaetocin in VHL-null RCC4 and in p53-null HCT116, but these efforts were in vain (Supporting Information Fig. 6B). We next checked the synthesis of HIF-1α protein in the presence of MG132 and found that it was attenuated by chaetocin (Fig. 4F). Because the PI3K-AKT-mTOR (mammalian target of rapamycin) pathway determines the translation of HIF-1α,16 we examined the effect of chaetocin on the pathway, but AKT and mTOR were not inactivated by chaetocin (Supporting Information Fig. 6C). These results hinted that chaetocin suppresses HIF-1α at the pretranslational level.
Chaetocin Down-Regulates HIF-1α mRNA.
In human hepatoma cells, HIF-1α mRNA was down-regulated by chaetocin at the concentrations which suppressed HIF-1α (Fig. 5A) as early as 4 hours after treatment (Fig. 5B). However, HIF-1α mRNA was not destabilized by chaetocin (Fig. 5C). Next, we investigated chromatin state at the promoter region of the HIF1A gene using chromatin-immunoprecipitation. The proximal promoter was predominantly associated with euchromatic histone H3 (methyl-K4 and acetyl-K9), but not with heterochromatic histone H3 (methyl-K9). Also, the promoter was targeted by transcription factors STAT3 and nuclear factor kappaB (NF-κB) and by RNA polymerase II (Fig. 5D).17, 18 This indicates that HIF-1α is actively transcribed. However, chaetocin did not affect the chromatin state, suggesting that chaetocin does not control the transcription of HIF-1α. We next analyzed HIF-1α pre-mRNA to evaluate the de novo synthesis of HIF-1α mRNA, and surprisingly found that chaetocin increases HIF-1α pre-mRNA levels (Fig. 5E). These paradoxical findings encouraged us to address the splicing of HIF-1α mRNA.
Chaetocin Deregulated the Pre-mRNA Splicing of HIF-1α.
We performed reverse-transcription polymerase chain reaction (RT-PCR) using two sets of primers to identify both spliced (shorter) and unspliced (longer) mRNAs simultaneously. Chaetocin inhibited the splicing in the same dose- and time-dependent manner as it down-regulated HIF-1α mRNA and protein (Fig. 6A,B). In contrast, the mRNAs of other genes (RIOK3, DNAJB1, and BRD2), whose pre-mRNAs have been reported to be unspliced by spliceosome inhibitors,19 were well spliced even in the presence of chaetocin (Fig. 6C), suggesting that splicing inhibition by chaetocin is a gene-dependent event. Furthermore, chaetocin did not inhibit HIF-1α pre-mRNA splicing effectively in noncancerous cells and other cancer cells (Fig. 6D,E). We also found that HIF-1α pre-mRNA was unspliced in five of six chaetocin-treated tumors (Fig. 6F). These results indicate that the in vivo effect of chaetocin on HIF-1α is attributed to the deregulation of HIF-1α pre-mRNA splicing.
Chaetocin Effectively, but Incompletely, Inhibited the Growths of Human Hepatoma Xenografts.
Although chaetocin retarded mouse hepatoma growth, it failed to inhibit tumor growth completely. As many anticancer regimens are based on drug combinations, we examined whether chaetocin could be used for combination therapy. In addition, we examined the effect of chaetocin against human hepatoma xenografts. After allowing Hur7 tumors to grow in mice, they were treated with DMSO, chaetocin, doxorubicin, or cotreated with chaetocin and doxorubicin. Mice were found to lose weight significantly after doxorubicin or combination treatment (Fig. 7A) and, thus, we decided to terminate all experiments on the 10th day. Chaetocin and doxorubicin both significantly retarded the growth of human hepatoma, but the combination treatment failed to attenuate tumor growth completely (Fig. 7A, Supporting Information Fig. 8). Furthermore, in hepatoma tissues, VEGF, and HIF-1α were down-regulated and HIF-1α pre-mRNA splicing was impaired (Fig. 7B,C). It was also found that doxorubicin inhibited VEGF expression without noticeably changing the level and splicing of HIF-1α mRNA, as has been demonstrated.20 These results further support the antiangiogenic and anticancer effects of chaetocin against hepatoma, and suggest that chaetocin does not potentiate the effects of cytotoxic anticancer agents. The potential combinatorial use of chaetocin needs to be reevaluated with other drugs using different treatment schedules.
In the present study, we found that chaetocin retards the in vivo growth of hepatoma and fibrosarcoma in an HIF-1α-dependent manner, and that it inhibits HIF-1α expression and vascular formation in tumors. Furthermore, chaetocin attenuated the HIF-1-mediated induction of hypoxic genes under culture conditions. Mechanistically, chaetocin was found to inhibit the synthesis of HIF-1α by deregulating the splicing of HIF-1α pre-mRNA, and in terms of this effect, hepatoma cells were found to be more sensitive to chaetocin than other cancer cells or immortalized normal cells. These results suggest that chaetocin has therapeutic potential for the control of solid tumors, including hepatoma. Furthermore, our findings suggest that HIF-1α pre-mRNA splicing should also be viewed as a therapeutic target.
The thiodioxopiperazine moiety of chaetocin has chirality opposite to that of chetomin. Chetomin has been reported to directly inhibit the interaction between HIF-1α and p300 and, thus, to repress HIF-1-driven gene expression.21 A recent report demonstrated that despite structural differences, three thiodioxopiperazines commonly inhibit the p300 binding in vitro and reduce VEGF secretion in HCT116 cells.22 However, as HIF-1α expression had not been determined, we examined whether chetomin, like chaetocin, down-regulates HIF-1α. Although chetomin repressed the transcriptional activity of HIF-1α, it had no effect on HIF-1α expression or pre-mRNA splicing (Supporting Information Fig. 7). These results indicate that chaetocin and chetomin inhibit HIF-1α in different ways. Indeed, we could not check the effect of chaetocin on p300-HIF-1α binding because HIF-1α disappeared. Nevertheless, because HIF-1α synthesis precedes p300-HIF-1α binding, the anticancer effect of chaetocin might be primarily due to HIF-1α suppression.
VEGF acts in a paracrine manner on endothelial cells to increase numbers of blood and lymphatic vessels, and also in an autocrine manner activates the VEGF receptor-mediated survival pathway. Therefore, antibodies and peptides that antagonize VEGF or its receptors have been developed as anticancer therapies.23, 24 We found that chaetocin inhibits VEGF production in hepatoma cells and grafts, and that vessels were poorly developed in chaetocin-treated tumors. These results suggest that the VEGF suppression underlies the antiangiogenic and anticancer action of chaetocin.
To correct ATP depletion and subsequent acidosis in hypoxia, HIF-1α facilitates ATP generation by up-regulating a number of glycolytic enzymes, but it inhibits oxidative phosphorylation by inducing PDK1, which blocks the trichloroacetic acid (TCA) cycle.25 HIF-1α also corrects acidosis by inducing CA9, which generates HCO.26 Accordingly, suppression of these metabolic genes by chaetocin may contribute to its cytotoxicity to hepatoma cells cultured under severe hypoxic conditions.
Many small molecules that inhibit HIF-1 have been reported in the literature. Some functionally inhibit HIF-1α by blocking its binding to p300 or DNA,21, 27 and others down-regulate HIF-1α by destabilizing it or by inhibiting its translation.28, 29 However, to the best of our knowledge, no agent has been previously reported to inhibit HIF-1α at the mRNA splicing level. Then, how does chaetocin inhibit HIF-1α pre-mRNA splicing? Spliceosome consists of small nuclear ribonucleoproteins and a host of associated proteins. In fact, 200 or more splicing factors have been found to interact with spliceosome. Because of its complexity, the splicing process is not well understood.30 Chaetocin did not affect the splicings of pre-mRNAs other than HIF-1α, which suggests that chaetocin targets some splicing factor(s) that specifically participate in HIF-1α pre-mRNA splicing, but the mechanism responsible for HIF-1α pre-mRNA splicing remains open.
Spliceosome has been viewed as a potential target for cancer therapy since pladienolide B and spliceostatin A were discovered. Both of these natural products impair pre-mRNA splicing by targeting the splicing factor SF3b, and consequently, inhibit tumor cell survival and growth.19, 30, 31 Chaetocin is a new example of the RNA process-targeting anticancer class. However, as compared to previously reported inhibitors, chaetocin has the merits of acting on specific cancer cells and genes and, thus, chaetocin offers the possibility of more selective antihepatoma therapy with fewer side effects.
We thank Dr. Eric Huang at the University of Utah and Dr. Randall Johnson at the University of California for kindly donating research materials.