Activation of nuclear hormone receptor peroxisome ... - Nature

B R I E F C O M M U N I C AT I O N S work was supported by grants from the German Research Council (SPP1047 and FOR 449/T2) and from Biosynexus to A.P. COMPETING INTERESTS STATEMENT The authors declare competing financial interests (see the Nature Medicine website for details).

© 2004 Nature Publishing Group http://www.nature.com/naturemedicine

Received 16 November 2003; accepted 5 January 2004 Published online at http://www.nature.com/naturemedicine/

1. Archer, G.L. Clin. Infect. Dis. 26, 1179–1181 (1998). 2. von Eiff, C., Becker, K., Machka, K., Stammer, H. & Peters, G. N. Engl. J. Med. 344, 11–16 (2001). 3. Peacock, S.J., de Silva, I. & Lowy, F.D. Trends Microbiol. 9, 605–610 (2001).

Activation of nuclear hormone receptor peroxisome proliferator–activated receptor-δ accelerates intestinal adenoma growth Rajnish A Gupta1, Dingzhi Wang1, Sharada Katkuri1, Haibin Wang2, Sudhansu K Dey2, & Raymond N DuBois1,2,3 We treated Apcmin mice, which are predisposed to intestinal polyposis, with a selective synthetic agonist of peroxisome proliferator–activated receptor-δ (PPAR-δ). Exposure of Apcmin mice to the PPAR-δ ligand GW501516 resulted in a significant increase in the number and size of intestinal polyps. The most prominent effect was on polyp size; mice treated with the PPAR-δ activator had a fivefold increase in the number of polyps larger than 2 mm. Our results implicate PPAR-δ in the regulation of intestinal adenoma growth. PPARs are ligand-activated transcription factors that belong to the nuclear hormone receptor superfamily1. Evidence supports a role for the PPAR-δ/β subtype in embryo implantation2 and development3, epidermal maturation and wound healing4,5, and regulation of fatty acid metabolism6, as well as in repressing the atherogenic inflammatory response7. Recent data suggest that PPAR-δ is also important in colorectal cancer (CRC) and is overexpressed in most CRCs8,9. PPAR-δ expression and/or activity is increased after loss of the adenomatous polyposis coli (APC) tumor suppressor gene or after K-Ras activation. The cyclooxygenase-2 metabolite prostacyclin also increases PPAR-δ activity in CRC cells8. Data in support of a direct role for PPAR-δ in CRC are conflicting. Somatic deletion of both PPARD alleles from an established CRC cell line results in decreased tumor growth10. In contrast, PPAR-δ was shown to be dispensable for polyp formation in Apcmin mice3, although this study was limited to a small number (n = 3) of Ppard–/–Apcmin mice. Notably, previous studies did not test whether PPAR-δ activation with a high-affinity agonist influ-

4. Endl, J., Seidl, H.P., Fiedler, F. & Schleifer, K.H. Arch. Microbiol. 135, 215–223 (1983). 5. Soldo, B., Lazarevic, V. & Karamata, D. Microbiology 148, 2079–2087 (2002). 6. Novick, R.P. Methods Enzymol. 204, 587–636 (1991). 7. Kokai-Kun, J.F., Walsh, S.M., Chanturiya, T. & Mond, J.J. Antimicrob. Agents Chemother. 47, 1589–1597 (2003). 8. Prince, G.A., Jenson, A.B., Horswood, R.L., Camargo, E. & Chanock, R.M. Am. J. Pathol. 93, 771–791 (1978). 9. Niewiesk, S. & Prince, G. Lab. Anim. 36, 357–372 (2002). 10. Peschel, A. et al. J. Biol. Chem. 274, 8405–8410 (1999). 11. Aly, R., Shinefield, H.R., Litz, C. & Maibach, H.I. J. Infect. Dis. 141, 463–465 (1980). 12. Cole, A.M. et al. J. Immunol. 169, 6985–6991 (2002). 13. Ganz, T. Nat. Rev. Immunol. 3, 710–720 (2003). 14. Peschel, A. Trends Microbiol. 10, 179–186 (2002). 15. Palaniyar, N., Nadesalingam, J. & Reid, K.B. Immunobiology 205, 575–594 (2002).

ences intestinal adenomatous polyp growth. We therefore treated Apcmin mice with the PPAR-δ synthetic ligand GW501516. GW501516 was shown to be a PPAR-δ subtype-selective ligand using combinatorial chemistry and structure-based drug design11. We evaluated this compound for PPAR-δ activation and receptor selectivity in HCT116 CRC cells using a mammalian two-hybrid transfection assay. Exposure of cells to physiological concentrations of GW501516 resulted in a dose-dependent increase in PPAR-δ–GAL-4 transactivation (Fig. 1a). Doses of GW501516 up to 10 µM did not transactivate either PPAR-γ or PPAR-α, confirming that GW501516 is a high-affinity, PPAR-δ-selective agonist. A large percentage of human colorectal polyps have inactivating mutations in the APC gene. Apcmin mice develop multiple intestinal polyps, providing a useful model system for our studies. Expression of PPAR-δ in adenomas and normal tissue from the small and large intestines was determined at the RNA (data not shown) and protein (Fig. 1b) levels. PPAR-δ was found mainly in intestinal epithelial cells in both the normal intestine and adenomas. To test the effects of PPAR-δ activation on polyp growth, we treated Apcmin mice with either vehicle or 10 mg/kg of GW501516. Treatment was limited to 6 weeks because of the development of rectal bleeding and signs of anemia in the ligand-treated mice. Consistent with published reports, the control Apcmin mice developed an average of 30 small intestine polyps and 1.4 colonic polyps. In contrast, GW501516 treatment led to a twofold increase in polyp number in the small intestine, with no change in the large bowel (Table 1). Two previous studies reported that treatment of Apcmin mice with a PPAR-γ agonist increased polyp formation12,13, raising the concern that GW501516 could be promoting polyposis by cross-activation of PPAR-γ rather than through direct activation of PPAR-δ. In the previous studies, however, the PPAR-γ ligand affected polyp size only in the colon, not in the small intestine. Our results with GW501516 are exactly the opposite: an increase in polyp size and number in the small intestine but not in the colon. Thus, it is unlikely that the phenotype seen with GW501516 is caused by PPAR-γ activation. Adenoma size is an independent risk factor for progression to CRC14. To determine whether PPAR-δ influences adenoma size, data were stratified for polyp size. Notably, mice treated with the PPAR-δ agonist showed a fivefold increase in polyps larger than 2 mm (Table 1), suggesting that PPAR-δ activation primarily

Departments of 1Medicine, 2Pediatrics and Cell and Developmental Biology and 3Cancer Biology, The Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-6838, USA. Correspondence should be addressed to R.N.D. ([email protected]). Published online 1 February 2004; doi:10.1038/nm993

NATURE MEDICINE VOLUME 10 | NUMBER 3 | MARCH 2004

245

B R I E F C O M M U N I C AT I O N S


a

c

b

d

Figure 1 Effect of a PPAR-δ agonist on intestinal polyposis. (a) GW501516 is a high-affinity PPARδ selective agonist. HCT116 colon carcinoma cells were transiently transfected with UAS-tkluc, pRL-SV40 and the indicated PPAR ligandbinding domain–GAL-4 fusion constructs. Data (n = 3) represent fold activation (luciferase activity) over cells treated with vehicle (0.1% DMSO) (mean ± s.e.m.). (b) Immunolocalization of PPAR-δ in intestinal tract of C57BL/6J Apcmin mice. Sections of small (SI) and large (LI) intestines were immunostained with rabbit antibody to mouse PPAR-δ (5 µg/ml)2. Shown are representative images of intestinal epithelia showing nuclear localization of PPAR-δ (reddish brown) in normal villi (N) and polyps (P). Preneutralization of antibody by excess antigenic peptide resulted in no positive staining (control, C). Scale bar, 100 µm. (c) Histology of tumors from vehicle- and GW501516-treated mice. Five-week-old C57BL/6J Apcmin mice (Jackson Laboratory). At 6 weeks of age, mice were gavaged five times a week for 6 weeks with 300 µl of 0.5% carboxymethylcellulose solution (vehicle) containing GW501516 (courtesy of CareX), or with 300 µl of vehicle alone as control. Tumor sample identifications were masked, and number, diameter and distribution of tumors were recorded. Histological analysis was performed in a blinded manner. Representative H&E-stained micrographs are shown. Scale bar, 100 µm. (d) GW501516 protects against growth factor withdrawal–induced apoptosis in wild-type, but not Ppard–/–, HCT116 cells. Cells were plated in serum-containing media overnight, followed by 24-h replacement with serum-free media containing vehicle (0.1% DMSO) or increasing concentrations of GW501516. After 24 h, cells were analyzed for apoptosis using a TUNEL-based assay (Roche). Data are presented as percent apoptosis of vehicle treated cells (n = 3; mean ± s.e.m.).

affects the rate of polyp growth rather than initiating polyp forma- stimulating cell proliferation. In contrast, pretreatment of wildtion. There is genetic evidence to support this notion. A previous type HCT116 cells with GW501516 significantly suppressed apopstudy tested the role of PPAR-δ in intestinal polyposis by generat- tosis in a dose-dependent manner (Fig. 1d). This antiapoptotic ing Ppard–/–Apcmin mice3. Nevertheless, although there was no dif- effect of GW501516 was not seen in Ppard–/– cells, strongly sugference in the average number of polyps between Ppard+/+ and gesting that the effects were the result of PPAR-δ activation. Ppard–/–Apcmin mice, a noticeable decrease was observed in the Collectively, these results argue that PPAR-δ stimulates intestinal adenoma development and size by activating antiapoptotic pathnumber of large polyps (>2 mm) in mice lacking PPAR-δ. The histology from both groups was typical of min adenomas ways in intestinal epithelial cells. Activating ligands of PPAR-δ are in the later stages of developwith carcinoma in situ, and there was no evidence for invasion beyond the basement membrane. In contrast, polyps in our ment as drugs for the treatment of diseases linked to energy imbalGW501516-treated mice were much larger, with a slightly higher ance, including dyslipidemia syndromes, morbid obesity and atherosclerosis. Our results raise concern that certain people (with degree of dysplasia (Fig. 1c). To understand the mechanism(s) by which PPAR-δ promotes APC loss) consuming PPAR-δ agonists could be at higher risk for adenomatous polyp growth in the intestine, we evaluated the CRC. Patients with familial adenomatous polyposis who have effects of PPAR-δ activation in human CRC cells. To eliminate germline APC mutations may be at particular risk for accelerated nonspecific effects of GW501516 unrelated to PPAR-δ activation, we examined wild- Table 1 Effect of GW501516 on tumor number and size in intestinal tracts of Apcmin mice type and Ppard–/– HCT116 cells10. Colon Intestine Exposure of wild-type or Ppard–/– HCT116 3 mm Total cells to increasing concentrations of 1.4 ± 0.4 6.0 ± 0.8 9.8 ± 0.9 11.1 ± 0.9 a 1.9 ± 0.8b 0 29.7 ± 2.2a GW501516 had no effect on cell prolifera- Control a b Experiment 1.7 ± 0.4 8.5 ± 1.6 13.6 ± 2.6 23.5 ± 3.5 9.8 ± 1.8 1 ± 0.4 56.4 ± 7.4a tion (data not shown), suggesting that PPAR-δ does not promote polyp size by aP < 0.01 by Student t-test, compared with controls; n = 10 mice per group.

246

VOLUME 10 | NUMBER 3 | MARCH 2004 NATURE MEDICINE


B R I E F C O M M U N I C AT I O N S polyposis and invasive carcinoma if given a PPAR-δ agonist for other indications. In addition, most sporadic colonic adenomas are initiated by somatic mutations in the APC gene. Because larger adenomas are at increased risk for progression to malignancy, patients with pre-existing polyps consuming a PPAR-δ agonist could be at higher risk for developing CRC. In summary, our results provide the first evidence that PPAR-δ activation promotes the growth of intestinal adenomas. However, these results must be verified in Ppard–/–Apcmin mice. PPAR-δ may be an attractive target for the development of small molecule antagonists as chemopreventive and/or chemotherapeutic agents for CRC. Our results should also raise caution for the pharmaceutical industry and others developing PPAR-δ agonists for clinical use until their proneoplastic effects are carefully examined in an appropriate model system. ACKNOWLEDGMENTS Mice were housed in accordance with institutional and NIH guidelines. This work is supported in part by US Public Health Services grants RO1DK 47279 and P0-CA-77839 to R.N.D. and HD 33994 to S.K.D. R.N.D. is a Hortense B. Ingram Professor of Molecular Oncology. R.N.D. (R37-DK47297) and S.K.D. (R37-HD12304) are recipients of National Institutes of Health MERIT awards.

NATURE MEDICINE VOLUME 10 | NUMBER 3 | MARCH 2004

We thank the T.J. Martell Foundation and the National Colorectal Cancer Research Alliance for their generous support to R.N.D. Histological analysis was performed by K. Washington (pathologist). COMPETING INTERESTS STATEMENT The authors declare that they have no competing financial interests. Received 19 October 2003; accepted 15 January 2004 Published online at http://www.nature.com/naturemedicine/ 1. Willson, T.M., Brown, P.J., Sternbach, D.D. & Henke, B.R. J. Med. Chem. 43, 527–550 (2000). 2. Lim, H. et al. Genes Dev. 13, 1561–1574 (1999). 3. Barak, Y. et al. Proc. Natl. Acad. Sci. USA 99, 303–308 (2002). 4. Tan, N.S. et al. Genes Dev. 15, 3263–3277 (2001). 5. Michalik, L. et al. J. Cell Biol. 154, 799–814 (2001). 6. Wang, Y.X. et al. Cell 113, 159–170 (2003). 7. Lee, C.H. et al. Science 302, 453–457 (2003). 8. Gupta, R.A. et al. Proc. Natl. Acad. Sci. USA 97, 13275–13280 (2000). 9. He, T.C., Chan, T.A., Vogelstein, B. & Kinzler, K.W. Cell 99, 335–345 (1999). 10. Park, B.H., Vogelstein, B. & Kinzler, K.W. Proc. Natl. Acad. Sci. USA 98, 2598–2603 (2001). 11. Oliver, W.R. et al. Proc. Natl. Acad. Sci. USA 98, 5306–5311 (2001). 12. Saez, E. et al. Nat. Med. 4, 1058–1061 (1998). 13. Lefebvre, A.M. et al. Nat. Med. 4, 1053–1057 (1998). 14. Kim, E.C. & Lance, P. Gastroenterol. Clin. North Am. 26, 1–17 (1997).

247