Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Dec 1.
Published in final edited form as: Mol Cancer Res. 2010 Nov 11;8(12):1601–1609. doi: 10.1158/1541-7786.MCR-10-0101

The MIF homolog, D-dopachrome tautomerase (D-DT), promotes COX-2 expression through β-catenin-dependent and independent mechanisms

Dan Xin 1, Beatriz E Rendon 2, Ming Zhao 2, Millicent Winner 2, Arlixer McGhee Coleman 3, Robert A Mitchell 1,2,3,*
PMCID: PMC3075601  NIHMSID: NIHMS252067  PMID: 21071513

Abstract

The cytokine/growth factor, macrophage migration inhibitory factor (MIF) contributes to pathologies associated with immune, inflammatory and neoplastic disease processes. Several studies have demonstrated an important contributing role for MIF-dependent cyclooxygenase-2 (COX-2) expression in the progression of these disorders. We now report that the MIF homolog, D-dopachrome tautomerase (D-DT), is both sufficient and necessary for maximal COX-2 expression in colorectal adenocarcinoma cell lines. D-DT-dependent COX-2 transcription is mediated in part by β-catenin protein stabilization and subsequent transcription. Also contributing to D-DTs regulation of COX-2 expression are the activities of both c-jun-N-terminal kinase (JNK) and the MIF-interacting protein, Jab1/CSN5. Interestingly, D-DT-dependent β-catenin stabilization was found to be regulated by COX-2 expression suggesting the existence of an amplification loop between COX-2 and β-catenin-mediated-transcription in these cells. Because both COX-2 and β-catenin-mediated transcription are important contributors to colorectal cancer (CRC) disease maintenance and progression, these findings suggest a unique and novel regulatory role for MIF family members in CRC pathogenesis.

Keywords: c-Jun, colorectal cancer, CSN5, JNK, MIF-2, PGE2, Wnt

Introduction

The development of colorectal cancer arises from the sequential accumulation of mutations or deletions in the coding sequence of a number of tumor-suppressor genes and oncogenes (1). One of the most commonly mutated tumor suppressors in CRC is the adenomatous polyposis coli (APC) gene which normally controls the levels and activity of Wnt-dependent transcription, through β-catenin phosphorylation/degradation, of pro-tumorigenic gene products (1, 2). One such gene product of β-catenin/TCF-dependent transcription is cyclooxygenase 2 (COX-2), a well described target of the Wnt pathway that is generally accepted as playing a contributory role in colorectal adenocarcinoma initiation and progression (3-5). COX-2 is one of two isoforms of prostaglandin H2 synthase (PGHS) and acts to catalyze the synthesis of eicosanoids and prostaglandins from arachidonic acid. COX-2 promotes CRC tumorigenesis and neoplastic maintenance through its metabolites' effects on angiogenesis, apoptosis and tumor cell invasiveness (6-8). Accordingly, selective COX-2 inhibitors suppress the growth of tumor cells in vitro and tumor growth and maintenance in vivo (9, 10). However, clinical studies in humans reveal that COX-2 antagonists also induce phenotypic changes in human vascular smooth muscle cells that increase the risk of myocardial infarction and other thrombotic cardiovascular events (11).

The pro-inflammatory and mitogenic cytokine, macrophage migration inhibitory factor (MIF) has been found to be an important endogenous mediator of COX-2 expression in a number of different cell types and is necessary for several of MIF's pro-inflammatory and pro-tumorigenic activities (12-15). Unlike other cytokines, MIF also has the unique ability to catalyze a non-physiologic enzymatic reaction (16). MIF converts D-Dopachrome – a stereoisomer of dopachrome not present in mammals - into 5, 6-dihydroxyindole-2-carboxylic acid. The only known MIF homolog, D-dopachrome tautomerase (D-DT), retains this tautomerase activity but also de-carboxylates the D-dopachrome substrate to give a 5, 6-dihydroxyindole product (17). While D-DT retains only 38% identity and 49% homology to MIF, the tertiary structure of D-DT is remarkably similar (18). Despite these intriguing similarities to the well studied MIF, there are virtually no reports on the biologic function(s) D-DT. Early studies describing the enzymatic activity and molecular cloning reveal that D-DT is relatively highly expressed in heart, brain, spleen, lung, skeletal muscle, kidney and testes while liver expression appears to be the highest (17). Although no prior studies investigating D-DT report on its expression in the colon, findings described herein indicate that D-DT is highly expressed in two human colorectal cancer cell lines.

We recently demonstrated that D-DT cooperates with MIF in dictating the steady state expression of the pro-angiogenic growth factors, VEGF and IL-8, in non-small cell lung cancer (NSCLC) cell lines (19). Angiogenic growth factor expression mediated by endogenous D-DT relies upon a c-Jun-N-terminal kinase (JNK)/AP-1-dependent signaling pathway. In the present study, we investigated the contribution and mechanism of D-DT to COX-2 expression in colorectal adenocarcinoma cells. We report herein that both β-catenin-dependent and β-catenin-independent pathways are utilized by D-DT in regulating COX-2 expression in human colorectal adenocarcinoma cells.

Materials and Methods

Cells and reagents

HCT-116, HT-29 and HeLa cell lines were purchased from ATCC and grown in Dulbecco's Modified Eagle Medium (DMEM) supplemented 10% heat inactivated FBS, L-glutamate and Gentamycin. Myc-tagged wildtype and mutant human β-catenin mammalian expression constructs were kindly provided by Dr. Frank McCormick (University of California, San Francisco) (20). Dr. Curtis C. Harris (National Cancer Institute, NIH), kindly provided us with wildtype, mutant and deletion human COX-2 promoter luciferase constructs. Top-Flash and Fop-Flash reporter plasmids were from Promega (Madison, WI). Antibodies used for immunoblotting include polyclonal and monoclonal antibodies directed against MIF (Santa Cruz Biotechnology, Santa Cruz, CA and R&D Systems, Minneapolis, MN, respectively), D-DT (19), V5 (Sigma), β-catenin (BD Transduction Laboratories, San Jose, CA), COX-2 (Cayman Chemical, Ann Arbor, MI), β-actin (Sigma), CSN5 (Bethyl Laboratories, Montgomery, TX) and α-tubulin (Sigma).

RNA Interference

shRNA design software from Dharmacon siDESIGN Center (www.dharmacon.com/sidesign/) was used to design shRNA sequences. MIF, D-DT and CSN5 oligos (MIF: 5′-CCTTCTGGTGGGGAGAAAT-3′; D-DT#2: 5′-GCCAGGACCGGATACTTAT-3′; Jab1/CSN5: 5′-GCTCAGAGTATCGATGAAA-3′) were ordered from Dharmacon (Thermo Scientific, Lafayette, CO). Commercially available shRNA directed against β-catenin was from Santa Cruz Biotechnology (Santa Cruz, CA). shRNA oligos were transfected into cells using Oligofectamine following manufacturer's directions (Invitrogen, Carlsbad, CA) (19, 21). As negative controls, both a commercially available control shRNA (Dharmacon) and a scrambled shRNA based on the sequence of D-DT shRNA were used accordingly (both referred to as nonspecific, Scr).

Quantitative PCR

For total RNA isolation, the RNeasy Mini Kit (Qiagen, Valencia, CA) was used following the manufacturer's protocol followed by cDNA synthesis using Omniscript RT (Qiagen, Madison, WI). Quantitative PCR was carried out on a DNA Engine Opticon Thermocycler (BioRad, Hercules, CA) using Takara PCR mix (Takara Bio Inc, Shiga, Japan), 0.3 μM forward and reverse primers (Invitrogen) and SYBr Green (Molecular Probes) diluted to a ratio of 1:25,000. Primers used were:

  • COX-2: Forward 5′-CTAGAGCCCTTCCTCCTGTG-3′;

  • Reverse 5′-GGGGATCAGGGATGAACTTT-3′

  • D-DT: Forward 5′-AGAACCGCTCCTACAGCAAG-3′

  • Reverse 5′-TAGGCGAAGGTGGAGTTGTT-3′

  • β-actin: Forward 5'-CAAGGCCAACCGCGAGAAGA-3′

  • Reverse 5′-GGATAGCACAGCCTGGATAG-3′

Relative expression levels of mRNAs were determined using the delta CT method. The ΔΔCT was calculated as the difference between the normalized CT values (ΔCT) of the treatment and the control samples: ΔΔCT = ΔCT treatment – ΔCT control. ΔΔCT was then converted to fold change by the following formula: fold change = 2–ΔΔCT.

Western Blotting

For whole cell extracts, cells were lysed in 1×RIPA buffer, harvested by scraping and homogenized with a 23 gauge needle on a 1 ml syringe. Protein concentrations were determined by the DC Protein Assay kit (BioRad, Hercules, CA). Equal concentrations of protein from indicated lysates were run on pre-cast 10% or 4-20% SDS-Polyacrylamide gels (BioRad) followed by transfer to PVDF membranes (Millipore, Charlottesville, VA). Membranes were blocked in blocking buffer (5% nonfat dried milk, 0.2% Tween-20, 1% Goat Serum in TBS), incubated with primary antibody in the same solution, washed 3 × 5 minutes in wash buffer (TBS + Tween-20) incubated with secondary antibody at a 1:8000 dilution in blocking buffer, washed again and visualized with ECL reagent (Pierce). TIFF images of scanned blots were analyzed by Scion Image (Scion Corp., Frederick, MD).

Luciferase Assay

4 × 104 cells/ml were plated in a 24 well plate and transfected with shRNA the following day for an additional 48 hours. Cells were then transiently co-transfected with 0.125 μg/well of Top-Flash or Fop-Flash luciferase promoter plasmid together with 0.0125 μg/well Renilla pRL-null plasmid (Promega) using Lipofectamine (Invitrogen) transfection reagent. After 24 hrs, Firefly and Renilla luciferase activities were measured by the Dual Luciferase Reporter Assay System (Promega, Madison, WI) on a TD-20/20 luminometer (Turner Designs).

Adenovirus

Adenovirus for D-DT was prepared as previously described (19). β-catenin and constitutively active MKK7 recombinant adenoviruses were purchased from Vector Biolabs (San Diego, CA) and were used to infect cells at ~ 5 × 107 virus particles/ml. COX-2 adenovirus was generated using the Gateway cloning system (Invitrogen). Briefly, human COX-2 was PCR amplified and TOPO cloned into the pENTR D-TOPO plasmid. LR Recombinase (Invitrogen) was used to shuttle inserts into the pAd/CMV/V5-DEST destination vector and subclones were confirmed by sequencing. Adenoviral vectors were digested with PacI, ethanol precipitated and transfected into 293A adenoviral packaging cells using Lipofectamine (Invitrogen). After amplifying viral supernatants, virus was purified using Virabind purification columns (Cell Biolabs, San Diego, CA) and tested for expression efficiency vs. toxicity.

Statistical Analyses

Results were expressed as mean ± SEM. Data comparisons were derived by unpaired, two-tailed T tests using GraphPad Prism 4.1 statistical program. P values < 0.05 were considered significant.

Results

COX-2 regulation by D-DT

Because the macrophage migration inhibitory factor (MIF) homolog, D-dopachrome tautomerase (D-DT), cooperates with MIF in regulating the steady state expression of angiogenic growth factors in lung adenocarcinoma cells, we investigated whether D-DT similarly regulates COX-2 expression as has been reported for MIF (12-14). Specifically, we examined the potential role for D-DT-dependent expression of COX-2 in human colorectal carcinoma cancer (CRC) cell lines. HT-29, a grade II human colon adenocarcinoma cell line (mutant APC) and HCT-116, human colorectal carcinoma cell line (22) (wildtype APC) were transfected with scrambled, D-DT or MIF shRNA oligos as described in the Materials and Methods section. As shown in Fig. 1A, shRNA-mediated knockdown of either MIF or D-DT in HT-29 and HCT-116 cells results in significant reductions in COX-2 expression accompanying ~ 70 – 80% reductions in MIF and D-DT protein levels (19) and data not shown.

FIGURE 1.

FIGURE 1

Open in a new tab

Regulation of COX-2 expression by D-DT A, Human colorectal carcinoma cell lines, HCT116 and HT-29 were plated in 6-well plates for 24 hrs followed by oligotransfection of 50 nM scrambled, non-specific shRNA oligos, 50 nM D-DT-specific shRNA oligos or 50 nM MIF-specific shRNA oligos as described (19, 21). After 48 hours, lysates were prepared and equal amounts of protein were analyzed by immunoblotting for COX-2 and α-Tubulin (loading control). Scion Image was used for densitometry and COX-2/Tubulin densitometry values are shown. Data is representative of 4 independent experiments. B, HT-29 cells were transfected with a D-DT overexpression construct in a dose-dependent manner (Lane 1, 4 μg pCDNA3.1 vector alone; Lane 2, 1 μg pCDNA3.1/D-DT; Lane 3, 2 μg pCDNA3.1/D-DT; lane 4, 4 μg pCDNA3.1/D-DT. 36 hours later, equal concentrations of protein lysates were analyzed by immunoblotting for COX-2, D-DT, V5-epitope tag and β-actin. Data is representative of 2 independent experiments. C, Different concentrations of pCDNA3.1/D-DT expression construct (0 ug, 1 ug or 2 ug pCDNA3.1/D-DT and 2 μg, 1 μg or 0 μg pCDNA3.1 vector alone to control for plasmid concentration) were transfected into HT-29 cells. 48 hr later total RNA was isolated and reverse transcribed to cDNA. Real time PCR was used to determine the relative quantities of mRNA in the indicated samples. Data shown represents the ΔCt of the average of duplicate reactions for each condition between target mRNA (D-DT and COX-2) and β-actin and is representative of two independent experiments.

To investigate whether D-DT is sufficient to promote COX-2 expression, HT-29 cells were transfected with varying concentrations of a D-DT mammalian expression construct or vector control alone and COX-2 protein levels were evaluated. As shown in Fig. 1B, COX-2 protein levels increased proportionally to increasing D-DT in D-DT transfected cells vs. control plasmid transfected cells alone. Furthermore, D-DT over-expression resulted in increased transcription from a COX-2 promoter-luciferase reporter plasmid (Fig. 1C). Combined, these results suggest that COX-2 is a transcriptional target for D-dopachrome tautomerase.

D-DT promotes β-catenin stabilization and transcriptional activity

Because β-catenin-mediated transcription contributes to COX-2 expression (3, 4) we next tested whether D-DT regulates β-catenin levels and/or its transcriptional activity. As shown in Fig. 2A, whole cell lysates from cells rendered deficient in D-DT by shRNA show reduced protein levels of both COX-2 and β-catenin. Moreover, D-DT-deficient cells had ~ 50% less endogenous β-catenin/TCF-dependent transcription as assessed by a β-catenin/TCF reporter plasmid (Fig. 2B – first bars in series). Importantly, ectopic over-expression of β-catenin in control shRNA-transfected cells resulted in a ~ 2.5 fold increase in ectopically expressed β-catenin/TCF reporter plasmid activity while D-DT-deficient cells had only a nominal increase in β-catenin-dependent transcription (Fig. 2B – middle bars in series). Interestingly, overexpression of a phosphorylation defective β-catenin construct resulted in a similar increase in ectopic β-catenin-dependent transcription in D-DT containing cells as expected but D-DT-deficient cells were largely resistant suggesting that the defect in ectopically expressed b-catenin-mediated transcription observed in D-DT-knockdown cells is independent of APC/GSK3β-mediated phosphorylation (Fig. 2B – last bars in series).

FIGURE 2.

FIGURE 2

Open in a new tab

Reduced β-catenin expression and activity in D-DT-deficient cells A, HT-29 cells were plated in 6-well plates for 24 hours followed by oligotransfection with 50 nM scrambled shRNA oligos or 50 nM D-DT-specific shRNA oligos for an additional 48 hours. Lysates were prepared and equal amounts of protein were analyzed by immunoblotting for COX-2, β-catenin and β-actin (loading control). B, 24 hours after transfection with Scr or D-DT shRNA, cells were co-transfected with the β-catenin/TCF reporter plasmids pTOP-flash/pRL-null (Renilla) alone or together with vector control (pCDNA3.1), wildtype (pCDNA3.1/β-catenin) or phosphorylation mutant β-catenin (pCDNA3.1/β-cateninmut) expression constructs. 24 hrs later, Firefly and Renilla luciferase activities were measured by the Dual Luciferase Reporter Assay System. Results are expressed as relative light units (RLUs) after normalizing ratios of luciferase/Renilla luciferase from triplicate samples. Parallel transfections with the β-catenin/TCF non-responsive reporter control construct, pFOP-flash, revealed equally negligible RLUs in all cells regardless of D-DT or β-catenin status (not shown). C, Cells were treated as in A) and 24 hours after shRNA transfection, D-DT adenovirus (+ Ad-D-DT) was used to infect D-DT shRNA transfected cells and GFP adenovirus was used for the remainder of the cells (− Ad-D-DT) for an additional 48 hours. Lysates were prepared and equal amounts of protein were analyzed by Western blot for COX-2, β-catenin and D-DT. D, Cells were treated with 20 μg/ml cycloheximide for the indicated times 48 hours after transfection with Scr or D-DT shRNA (top panel) or 24 hours after infection with Ad-GFP (− Ad-D-DT) or Ad-D-DT (+ Ad-D-DT). At the indicated times following treatment with cycloheximide, cells were lysed and equal amounts of protein from each sample were analyzed by immunoblotting. All results are representative of at least 3 independent experiments.

To rule out the possibility that the reduced COX-2 and β-catenin protein levels in D-DT-deficient cells was independent of the loss of D-DT, we re-introduced D-DT by adenoviral infection. Please note that adenoviral-mediated expression of D-DT – as opposed to adding recombinant D-DT extracellularly - was used throughout these studies in order to recapitulate the D-DT expression in CRC cells with respect to both intra- and extracellular compartments. Because D-DT's homolog, MIF, has functional activities both inside and outside of cells (23), we sought to ensure that if D-DT has similar functional properties it would be able to carry out both by expressing it as it would be found normally in cells. As shown in Fig. 2C, both COX-2 and β-catenin protein levels were largely restored to control shRNA transfected cell levels in D-DT knockdown cells infected with adenoviral D-DT. These results validate D-DT as being responsible for the effects of D-DT shRNA on COX-2 and β-catenin expression.

The ability of β-catenin to contribute to nuclear TCF-dependent transcription depends, in large part, on the relative ability of E3-ubiquitin ligase complexes to recognize and ubiquitylate cytoplasmic β-catenin thus leading to its degradation (24). To investigate whether decreased β-catenin levels and transcriptional activity observed in D-DT-deficient cells was due to decreased β-catenin protein stability, we determined the relative degradation of β-catenin in D-DT over-expressing and deficient cells. Following the addition of the protein translation inhibitor, cycloheximide, to scrambled or D-DT shRNA transfected HT-29 cells, total β-catenin protein levels were assessed by immunoblotting. As shown in Fig. 2D (top panel), β-catenin expression goes down dramatically faster in D-DT-deficient cells than in control cells. Accordingly, cells infected with D-DT adenovirus displayed slightly higher β-catenin stability compared to GFP infected cells (Fig. 2D, bottom panel), suggesting that endogenous D-DT levels are, in fact, important for maintaining β-catenin stability. It should be noted that the ability of ectopically expressed D-DT to stabilize β-catenin and induce COX-2 expression is generally modest suggesting that endogenous D-DT levels are near their contribution threshold.

We next determined whether D-DT-dependent modulation of β-catenin stability contributes to COX-2 expression in CRC cell lines. Because both endogenous and ectopically expressed D-DT (albeit modestly) stabilizes β-catenin protein levels, we over-expressed increasing amounts of β-catenin in the presence or absence of ectopically-expressed D-DT. As shown in Fig. 3A, HT-29 cells were transfected with empty vector or D-DT along with increasing amounts of β-catenin, and COX-2 expression was evaluated. While β-catenin transfection, in the absence of ectopically expressed D-DT, was sufficient for enhancing COX-2 expression over basal levels, the combination of D-DT with β-catenin significantly increased COX-2 levels above that seen with ectopically expressed β-catenin alone. This result supports our hypothesis that D-DT acts to stabilize β-catenin thereby increasing its pool of transcriptionally active nuclear β-catenin.

FIGURE 3.

FIGURE 3

Open in a new tab

D-DT-mediated β-catenin stabilization contributes to, but is not wholly responsible for, D-DT-dependent COX-2 expression A, Cells were plated in 6-well plates and were transfected the following day with pCDNA3.1/D-DT or pCDNA3.1/β-catenin expression constructs, alone or in combination. For all plasmids, one “+” correlates to 1 μg construct transfected into cells. The total amount of DNA transfected into cells was equalized by co-transfecting vector control, pcDNA3.1. 48 hours post-transfection, cell lysates were assessed by immunoblotting for COX-2 and β-catenin. Densitometry of COX-2 and Tubulin were determined and graphed as COX-2/Tubulin. B, Cells were transfected with Scr or β-catenin shRNA oligos and were infected with Ad-GFP or Ad-D-DT 24 hours later as indicated. 48 hours post-infection, cells were lysed and assessed by immunoblotting for COX-2, β-catenin, D-DT and α-tubulin (loading control). Densitometry of COX-2 western blots from 2 independent experiments was determined and fold COX-2/Tubulin expression is shown.

To further investigate whether D-DT-mediated regulation of enhanced β-catenin expression contributes to D-DT-induced COX-2, D-DT was over-expressed in the presence or absence of β-catenin and COX-2 induction was determined. As shown in Fig. 3B, adenoviral-delivered D-DT increased both COX-2 and β-catenin levels in control shRNA-transfected cells as expected. Interestingly, shRNA knockdown of β-catenin in D-DT overexpressing cells resulted in a reduction of D-DT-induced COX-2 expression by only ~ 50%. This finding indicated to us that β-catenin may not be the only point of control in D-DT-dependent COX-2 expression in CRC cells.

Regulation of JNK/c-jun pathway activation by D-DT

Our laboratory recently reported that D-DT is an integral regulator of the c-jun-N-terminal kinase (JNK) signaling pathway leading to c-jun phosphorylation and angiogenic growth factor expression in lung adenocarcinoma cells (19). Because c-jun/AP-1 transcription is a well described pathway leading to COX-2 transcription (25-27), we next sought to investigate whether D-DT-dependent c-jun activation contributes to, along with β-catenin, D-DT-induced COX-2 expression. From Fig. 4A and consistent with our earlier findings (19), D-DT-deficient cells have reduced levels of phosphorylated c-jun (Fig. 4A) while cells infected with Ad-D-DT results in a substantial increase in c-jun phosphorylation compared to cells infected with Ad-GFP (Fig. 4B). Importantly, cells treated with the JNK small molecule inhibitor, SP600125, abolished both steady state (Ad-GFP infected) and Ad-D-DT-induced COX-2 expression (Fig. 4B). Combined, these findings strongly suggest that phosphorylated c-jun is a central participant in CRC COX-2 expression.

FIGURE 4.

FIGURE 4

Open in a new tab

JNK/c-Jun are necessary for COX-2 expression in CRC cells and D-DT contributes to JNK/c-Jun activation pathway A, Cells were oligo-transfected with Scr and D-DT shRNA oligos for 48 hours, lysed and immunoblotted for phospho-c-Jun, total c-jun and α-tubulin. B, Cells were infected with Ad-GFP or Ad-D-DT adenovirus for 24 hours and then treated with 40 μM SP600125 for an additional 6 hours. Cells were harvested, lysed and immunoblotted for phospho-c-Jun, COX-2 and β-actin. Densitometry of COX-2/β-actin is shown below. C, HT-29 cells were transfected with Scr or D-DT shRNA and 24 hours later cells were infected with Ad-GFP or constitutively active MKK7 adenovirus (Ad-CA-MKK7) as indicated. After 48 hours, cells were harvested, lysed and immunoblotted for phospho-c-Jun, COX-2, MKK7, D-DT and α-tubulin. Densitometries of COX-2 and tubulin were determined from two independent experiments and graphed as COX-2/Tubulin.

To further investigate the contribution of c-jun activation in D-DT-mediated COX-2 expression, we over-expressed a constitutively active mutant of MKK7 (CA MKK7), the upstream kinase of c-jun-N-terminal kinase (JNK), in D-DT-deficient cells. As shown in Fig. 4D, overexpression of CA MKK7 results in increases in both phosphorylated c-jun and COX-2 expression in control shRNA-transfected cells. Interestingly, CA MKK7 increased c-jun phosphorylation levels – although at slightly lower levels than that observed in control cells - but had no effect on rescuing defective COX-2 expression in D-DT-deficient cells. These results suggested to us that phosphorylated c-jun-mediated transcription – likely by the AP-1 transcription factor complex - was somehow defective in D-DT-deficient cells.

Kleeman and colleague showed that the D-DT homolog, MIF, regulates AP-1-mediated transcriptional activity by functionally regulating the MIF binding protein, Jab1 (Jun activation domain binding protein-1)/CSN5 (28). In an effort to determine whether Jab1/CSN5 might be implicated in the apparent requirement for D-DT in CA MKK-7-induced COX-2 expression – even though c-jun is adequately phosphorylated – we investigated the relative requirements for Jab1/CSN5 in steady state CRC COX-2 expression. As shown in Figure 5A, shRNA knockdown of CSN5 results in a dramatic reduction of COX-2 expression in CRC cells mirroring the effects of JNK inhibition (Fig. 5B). We next sought to determine if there was a requirement for Jab1/CSN5 in D-DT-induced COX-2 expression as would be predicted if Jab1/CSN5 is, in fact, necessary for AP-1 transcriptional activation of COX-2. As shown in Figure 5B, D-DT stimulates COX-2 expression in control shRNA transfected cells but CSN5-deficient cells are entirely refractory to D-DT-induced COX-2 expression despite a robust stimulation of c-jun phosphorylation (Fig. 5B, top panel). Similarly, but not unexpectedly, CA MKK7 was also unable to promote COX-2 expression in Jab1/CSN5-deficient cells (Fig. 5C). Combined, these results suggest that there is a functional requirement for Jab1/CSN5 in D-DT-induced AP-1 activity on COX-2 transcriptional expression. Although beyond the scope of these studies, experiments are currently underway to evaluate the binding and transactivation properties of AP-1 on the human COX-2 promoter in the context of D-DT and/or Jab1/CSN5 manipulation.

FIGURE 5.

FIGURE 5

Open in a new tab

CSN5 participates in the regulation of COX-2 and β-catenin. A, Nonspecific control (Scr) and CSN5-specific shRNA-transfected cells were harvested, lysed and immunoblotted for COX-2, CSN5 and α-tubulin. B, 24 hours after transfection with Scr or CSN5 shRNA oligos, cells were infected by either Ad-GFP or Ad-D-DT for an additional 48 hours. Cells were then harvested, lysed and immunoblotted for phospho-JNK, phospho-c-Jun, COX-2, CSN5, D-DT and α-tubulin. C, Cells were treated as in B) but were infected with Ad-GFP and Ad-CA-MKK7 adenoviruses. 48 hours after infection, cells were harvested, lysed and immunoblotted for phospho-c-Jun, COX-2, CSN5, MKK7 and α-tubulin. D, Cells were oligo-transfected with Scr or D-DT shRNA oligos and 24 hours later, cells were infected with Ad-GFP or Ad-COX-2 adenovirus as indicated (for 0 μl Ad-COX-2 cells were treated with Ad-GFP at 15 μl and Ad-GFP was co-infected with Ad-COX-2 at 5 and 10 μl to give 15 μl total adenovirus for each sample infected). 48 hrs after infection, cells were lysed and immunoblotted for β-catenin, COX-2, D-DT and α-tubulin. Densitometry of COX-2/Tubulin is shown. All data is representative of 2 independent experiments.

Finally, in an effort to evaluate how D-DT influences β-catenin stability, we re-expressed COX-2 using an adenoviral delivery system. The rationale was that prostaglandin E2 (PGE2), the primary product of COX-2-mediated arachidonic acid metabolism, is an important physiological regulator of β-catenin stabilization in colorectal cancers (29). Consistent with this and other studies demonstrating a role for COX-2/PGE2 in promoting β-catenin stability (30, 31), Ad-COX-2 increased β-catenin expression in both D-DT competent and D-DT-deficient cells (Fig. 5D). Intriguingly, we also observed a dose-dependent increase in endogenous (Fig. 5D, 2nd panel – lower band) COX-2 when Ad-COX-2 was expressed in D-DT competent and D-DT-deficient cells (Fig. 5D, 2nd panel – upper band). Combined, these results suggest that an amplification loop exists between COX-2-derived PGE2 and β-catenin/TCF-dependent COX-2 transcription in CRC cells and that D-DT expression is necessary for maximally promoting the COX-2 expression component of this pathway (Fig. 6).

FIGURE 6.

FIGURE 6

Open in a new tab

Proposed scheme of D-DT-mediated COX-2 expression in CRC cell lines.

Discussion

Prior studies from our laboratory revealed that MIF and D-DT cooperatively regulate JNK activation leading to AP-1-mediated VEGF and CXCL-8 expression (19). The combined effects of MIF and D-DT on angiogenic growth factor expression was found to provide nearly 90% of the neoangiogenic potential of those lung adenocarcinoma cell lines tested when assessed by HUVEC tube formation and migration assays (19). Importantly, the additive effects of MIF and D-DT in NSCLC was dependent upon the MIF and D-DT cell surface receptor, CD74 (19). Although not shown here, both HT-29 and HCT-116 cells used in this study express CD74 at moderately high levels consistent with prior studies demonstrating elevated CD74 expression in a large percentage (~ 85%) of human colorectal adenomas (32). Because COX-2 is an important transcriptional target of MIF (12, 13) and is phenotypically important for colorectal cancer maintenance and progression (33-35), we sought to determine if D-DT and MIF similarly regulate COX-2 expression in colorectal adenocarcinoma cells. We now demonstrate that, like MIF, D-DT is both sufficient and necessary for maximal COX-2 expression in CRC cell lines. Importantly, our results indicate – for the first time – an important regulatory role for D-DT in controlling the stability and transcriptional activity of β-catenin. Although studies are currently underway to elucidate this pathway in more detail and to validate these findings in a mouse model of spontaneously arising CRC (36) our current results indicate that loss of β-catenin stabilization in D-DT-deficient cells is overcome by ectopic re-introduction of COX-2. Coinciding with the apparent rescue of β-catenin expression in Ad-COX-2-expressing D-DT-deficient cells is the rescue of endogenous COX-2 indicating the existence of an amplification loop in these cells. This is telling because other rescue attempts, including overexpressing both wildtype and phosphorylation mutants of β-catenin (Fig. 2C and data not shown), CSN5 (not shown) and CA MKK7 (Fig. 4C), were all unsuccessful at restoring endogenous COX-2 expression. The lack of success with these rescue attempts was not surprising given the fact that both β-catenin and MKK7-induced AP-1 activation appear to require D-DT for maximal activities (β-catenin through enhanced stabilization and MKK7/AP-1 through regulation of CSN5). Given the results, we hypothesize that rescue of endogenous COX-2 by Ad-COX-2 is through PGE2-mediated re-stabilization of β-catenin and subsequent β-catenin/TCF-mediated COX-2 transcription. This hypothesis further predicts that the loss of β-catenin stabilization observed in D-DT-deficiency (Fig. 2) is due either to: 1) the loss of COX-2/PGE2 because of defective JNK/c-jun/AP-1 activity (Figs. 4 and 5), or; 2) D-DT-dependent regulation of β-catenin stability functions through the same pathway as that of PGE2. While the data suggests that Ad-COX-2-mediated β-catenin stabilization is responsible for the rescue of endogenous COX-2 in D-DT-deficient cells, we can't rule out the possibility that PGE2 also activates c-jun phosphorylation as has been suggested (37, 38). Recent studies reveal that PGE2 is an important endogenous regulator of β-catenin stabilization and Wnt signaling/transcription (29, 30). Studies are currently underway to evaluate the relative contribution of PGE2 to D-DT-dependent β-catenin stabilization. If our data does reveal a role for PGE2 in β-catenin stabilization in these cells, this would suggest that D-DT participates in an amplification loop consisting of β-catenin-dependent COX-2 transcription and COX-2-dependent β-catenin stabilization (Fig. 6). Nevertheless, our current results indicate that D-DT, and likely MIF, are necessary for maximal signaling/transcription through this pathway.

The fact that MIF family members facilitate both JNK/c-jun pathway activation and β-catenin stabilization suggest that COX-2 may be only one of a number of important gene products associated with colorectal carcinoma disease maintenance and progression. Studies by the Behrens' group demonstrate that these two pathways synergize to enhance intestinal tumorigenesis (39, 40). Specifically, JNK-phosphorylated c-jun interacts with β-catenin/TCF4 complexes on the promoters of Wnt and AP-1 transcriptional target genes and disruption of either pathway dramatically alters the tumor burden and the transcription of these target genes tumors from the ApcMin mouse model of intestinal tumorigenesis (39, 40). Given the apparent contributions of D-DT (and likely MIF) to both JNK/c-jun activation and β-catenin stabilization, it is not unreasonable to speculate that these MIF family members may promote CRC pathogenesis through this pathway.

Consistent with an important contributing role for MIF family members to intestinal neoplastic disease progression and pathogenesis, genetic ablation or small molecule inhibition of MIF alone results in significant reductions in intestinal tumorigenesis in mouse models of CRC (41, 42). Our results, coupled with earlier findings demonstrating potent additive effects of MIF and D-DT in contributing to non-small cell lung cancer angiogenesis (19), strongly suggest that the dual targeting of both MIF family members simultaneously may represent a potent and novel anti-cancer, immune or inflammatory disorder chemotherapeutic targeting strategy. Finally, due to the overlapping nature of biological activities between MIF and D-DT and the non-physiological activity of the catalysis for which D-DT was named (D-dopachrome tautomerase), we propose changing the name of D-DT to MIF-2.

Acknowledgments

Grant Support: This work was supported in part by NIH CA102285-S (A.M.C), NIH CA102285 (R.A.M), NIH CA129967 (R.A.M) and a grant from Philip Morris USA Inc. and Philip Morris International (R.A.M.).

Abbreviations

CSN5

COP9 signalosome subunit 5

CRC

colorectal cancer

COX-2

cyclooxygenase 2

D-DT

D-dopachrome tautomerase

JNK

c-jun-N-terminal kinase

PGE2

prostaglandin E2

TCF

T cell factor

References

  • 1.Phelps RA, Chidester S, Dehghanizadeh S, et al. A two-step model for colon adenoma initiation and progression caused by APC loss. Cell. 2009;137:623–634. doi: 10.1016/j.cell.2009.02.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Quyn AJ, Steele RJ, Carey FA, Nathke IS. Prognostic and therapeutic implications of Apc mutations in colorectal cancer. Surgeon. 2008;6:350–356. doi: 10.1016/s1479-666x(08)80007-8. [DOI] [PubMed] [Google Scholar]
  • 3.Wong NA, Pignatelli M. Beta-catenin--a linchpin in colorectal carcinogenesis? Am J Pathol. 2002;160:389–401. doi: 10.1016/s0002-9440(10)64856-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Araki Y, Okamura S, Hussain SP, et al. Regulation of cyclooxygenase-2 expression by the Wnt and Ras pathways. Cancer Res. 2003;63:728–734. [PubMed] [Google Scholar]
  • 5.Zhao S, Cai J, Bian H, Gui L, Zhao F. Synergistic inhibition effect of tumor growth by using celecoxib in combination with oxaliplatin. Cancer Invest. 2009;27:636–640. doi: 10.1080/07357900802672738. [DOI] [PubMed] [Google Scholar]
  • 6.Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell. 1998;93:705–716. doi: 10.1016/s0092-8674(00)81433-6. [DOI] [PubMed] [Google Scholar]
  • 7.Lev-Ari S, Strier L, Kazanov D, et al. Celecoxib and curcumin synergistically inhibit the growth of colorectal cancer cells. Clin Cancer Res. 2005;11:6738–6744. doi: 10.1158/1078-0432.CCR-05-0171. [DOI] [PubMed] [Google Scholar]
  • 8.Adam L, Mazumdar A, Sharma T, Jones TR, Kumar R. A three-dimensional and temporo-spatial model to study invasiveness of cancer cells by heregulin and prostaglandin E2. Cancer Res. 2001;61:81–87. [PubMed] [Google Scholar]
  • 9.Chun KS, Surh YJ. Signal transduction pathways regulating cyclooxygenase-2 expression: potential molecular targets for chemoprevention. Biochem Pharmacol. 2004;68:1089–1100. doi: 10.1016/j.bcp.2004.05.031. [DOI] [PubMed] [Google Scholar]
  • 10.Sheng H, Shao J, Kirkland SC, et al. Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest. 1997;99:2254–2259. doi: 10.1172/JCI119400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bertagnolli MM, Eagle CJ, Zauber AG, et al. Five-year efficacy and safety analysis of the Adenoma Prevention with Celecoxib Trial. Cancer Prev Res (Phila Pa) 2009;2:310–321. doi: 10.1158/1940-6207.CAPR-08-0206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Mitchell RA, Liao H, Chesney J, et al. Macrophage migration inhibitory factor (MIF) sustains macrophage proinflammatory function by inhibiting p53: regulatory role in the innate immune response. Proc Natl Acad Sci USA. 2002;99:345–350. doi: 10.1073/pnas.012511599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sampey AV, Hall PH, Mitchell RA, Metz CN, Morand EF. Regulation of synoviocyte phospholipase A2 and cyclooxygenase 2 by macrophage migration inhibitory factor. Arthritis Rheum. 2001;44:1273–1280. doi: 10.1002/1529-0131(200106)44:6<1273::AID-ART219>3.0.CO;2-8. [DOI] [PubMed] [Google Scholar]
  • 14.Carli C, Metz CN, Al-Abed Y, Naccache PH, Akoum A. Up-regulation of cyclooxygenase-2 expression and prostaglandin E2 production in human endometriotic cells by macrophage migration inhibitory factor: involvement of novel kinase signaling pathways 2009. Endocrinology. 2009;150:3128–3137. doi: 10.1210/en.2008-1088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Itoh A, Nishihira J, Makita H, Miyamoto K, Yamaguchi E, Nishimura M. Effects of IL-1beta, TNF-alpha, and macrophage migration inhibitory factor on prostacyclin synthesis in rat pulmonary artery smooth muscle cells. Respirology. 2003;8:467–472. doi: 10.1046/j.1440-1843.2003.00491.x. [DOI] [PubMed] [Google Scholar]
  • 16.Rosengren E, Bucala R, Aman P, et al. The immunoregulatory mediator macrophage migration inhibitory factor (MIF) catalyzes a tautomerization reaction. Mol Med. 1996;2:143–149. [PMC free article] [PubMed] [Google Scholar]
  • 17.Zhang M, Aman P, Grubb A, et al. Cloning and sequencing of a cDNA encoding rat D-dopachrome tautomerase. FEBS Lett. 1995;373:203–206. doi: 10.1016/0014-5793(95)01041-c. [DOI] [PubMed] [Google Scholar]
  • 18.Sugimoto H, Taniguchi M, Nakagawa A, Tanaka I, Suzuki M, Nishihira J. Crystal structure of human D-dopachrome tautomerase, a homologue of macrophage migration inhibitory factor, at 1.54 A resolution. Biochemistry. 1999;38:3268–3279. doi: 10.1021/bi982184o. [DOI] [PubMed] [Google Scholar]
  • 19.Coleman AM, Rendon BE, Zhao M, et al. Cooperative regulation of non-small cell lung carcinoma angiogenic potential by macrophage migration inhibitory factor and its homolog, D-dopachrome tautomerase. J Immunol. 2008;181:2330–2337. doi: 10.4049/jimmunol.181.4.2330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Tetsu O, McCormick F. Beta-catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature. 1999;398:422–426. doi: 10.1038/18884. [DOI] [PubMed] [Google Scholar]
  • 21.Rendon BE, Roger T, Teneng I, et al. Regulation of human lung adenocarcinoma cell migration and invasion by macrophage migration inhibitory factor. J Biol Chem. 2007;282:29910–29918. doi: 10.1074/jbc.M704898200. [DOI] [PubMed] [Google Scholar]
  • 22.Brattain MG, Fine WD, Khaled FM, Thompson J, Brattain DE. Heterogeneity of malignant cells from a human colonic carcinoma. Cancer Res. 1981;41:1751–1756. [PubMed] [Google Scholar]
  • 23.Rendon BE, Willer SS, Zundel W, Mitchell RA. Mechanisms of macrophage migration inhibitory factor (MIF)-dependent tumor microenvironmental adaptation. Exp Mol Pathol. 2009;86:180–5. doi: 10.1016/j.yexmp.2009.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.MacDonald BT, Tamai K, He X. Wnt/beta-catenin signaling: components, mechanisms, and diseases. Dev Cell. 2009;17:9–26. doi: 10.1016/j.devcel.2009.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zhang R, Chen HZ, Liu JJ, et al. SIRT1 suppresses activator protein-1 transcriptional activity and cyclooxygenase-2 expression in macrophages. J Biol Chem. 2009;285:7097–110. doi: 10.1074/jbc.M109.038604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Lee KM, Lee KW, Bode AM, Lee HJ, Dong Z. Tpl2 is a key mediator of arsenite-induced signal transduction. Cancer Res. 2009;69:8043–8049. doi: 10.1158/0008-5472.CAN-09-2316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Looby E, bdel-Latif MM, thie-Morales V, Duggan S, Long A, Kelleher D. Deoxycholate induces COX-2 expression via Erk1/2-, p38-MAPK and AP-1-dependent mechanisms in esophageal cancer cells. BMC Cancer. 2009;9:190. doi: 10.1186/1471-2407-9-190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Kleemann R, Hausser A, Geiger G, et al. Intracellular action of the cytokine MIF to modulate AP-1 activity and the cell cycle through Jab1. Nature. 2000;408:211–216. doi: 10.1038/35041591. [DOI] [PubMed] [Google Scholar]
  • 29.Castellone MD, Teramoto H, Williams BO, Druey KM, Gutkind JS. Prostaglandin E2 promotes colon cancer cell growth through a Gs-axin-beta-catenin signaling axis. Science. 2005;310:1504–1510. doi: 10.1126/science.1116221. [DOI] [PubMed] [Google Scholar]
  • 30.Goessling W, North TE, Loewer S, et al. Genetic interaction of PGE2 and Wnt signaling regulates developmental specification of stem cells and regeneration. Cell. 2009;136:1136–1147. doi: 10.1016/j.cell.2009.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lim K, Han C, Xu L, Isse K, Demetris AJ, Wu T. Cyclooxygenase-2-derived prostaglandin E2 activates beta-catenin in human cholangiocarcinoma cells: evidence for inhibition of these signaling pathways by omega 3 polyunsaturated fatty acids. Cancer Res. 2008;68:553–560. doi: 10.1158/0008-5472.CAN-07-2295. [DOI] [PubMed] [Google Scholar]
  • 32.Cuthbert RJ, Wilson JM, Scott N, Coletta PL, Hull MA. Differential CD74 (major histocompatibility complex Class II invariant chain) expression in mouse and human intestinal adenomas. Eur J Cancer. 2009;45:1654–1663. doi: 10.1016/j.ejca.2009.02.005. [DOI] [PubMed] [Google Scholar]
  • 33.Oshima M, Dinchuk JE, Kargman SL, et al. Suppression of intestinal polyposis in Apc delta716 knockout mice by inhibition of cyclooxygenase 2 (COX-2) Cell. 1996;87:803–809. doi: 10.1016/s0092-8674(00)81988-1. [DOI] [PubMed] [Google Scholar]
  • 34.Sheng H, Shao J, Kirkland SC, et al. Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest. 1997;99:2254–2259. doi: 10.1172/JCI119400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA. 1997;94:3336–3340. doi: 10.1073/pnas.94.7.3336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Dietrich WF, Lander ES, Smith JS, et al. Genetic identification of Mom-1, a major modifier locus affecting Min-induced intestinal neoplasia in the mouse. Cell. 1993;75:631–639. doi: 10.1016/0092-8674(93)90484-8. [DOI] [PubMed] [Google Scholar]
  • 37.Chen D, Reierstad S, Lin Z, et al. Prostaglandin E(2) induces breast cancer related aromatase promoters via activation of p38 and c-Jun NH(2)-terminal kinase in adipose fibroblasts. Cancer Res. 2007;67:8914–8922. doi: 10.1158/0008-5472.CAN-06-4751. [DOI] [PubMed] [Google Scholar]
  • 38.Black SA, Jr, Palamakumbura AH, Stan M, Trackman PC. Tissue-specific mechanisms for CCN2/CTGF persistence in fibrotic gingiva: interactions between cAMP and MAPK signaling pathways, and prostaglandin E2-EP3 receptor mediated activation of the c-JUN N-terminal kinase. J Biol Chem. 2007;282:15416–15429. doi: 10.1074/jbc.M610432200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Nateri AS, Spencer-Dene B, Behrens A. Interaction of phosphorylated c-Jun with TCF4 regulates intestinal cancer development. Nature. 2005;437:281–285. doi: 10.1038/nature03914. [DOI] [PubMed] [Google Scholar]
  • 40.Sancho R, Nateri AS, de Vinuesa AG, et al. JNK signalling modulates intestinal homeostasis and tumourigenesis in mice. EMBO J. 2009;28:1843–1854. doi: 10.1038/emboj.2009.153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Wilson JM, Coletta PL, Cuthbert RJ, et al. Macrophage migration inhibitory factor promotes intestinal tumorigenesis. Gastroenterology. 2005;129:1485–1503. doi: 10.1053/j.gastro.2005.07.061. [DOI] [PubMed] [Google Scholar]
  • 42.He XX, Chen K, Yang J, et al. Macrophage migration inhibitory factor promotes colorectal cancer. Mol Med. 2009;15:1–10. doi: 10.2119/molmed.2008.00107. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES