84

All-trans retinoic acid (ATRA) is the primary biologically active retinoid and has been shown to exert its effects primarily through nuclear retinoid receptor proteins. Fenretinide (N-[4-Hydroxyphenyl]retinamide or 4HPR) is a synthetic retinoid with antitumor and chemopreventive activities. N-[4-Methoxyphenyl]retinamide (4MPR) is the major metabolite of 4HPR in patients. Using a series of retinoid receptor knockout cell lines we demonstrated two distinct effects of 4HPR and 4MPR. The first is a rapid induction of apoptosis caused by both phenylretinamides at a concentration of 10μM, and the second is a delayed induction of differentiation at a lower concentration (1μM) observed only with 4HPR. In the present study we compared the chemopreventive effects of 4HPR, 4MPR and ATRA (receptor dependent activity only) in the DMBA/TPA 2-stage mouse skin carcinogenesis model. All three agents suppressed tumor formation, with ATRA being the most potent, implicating retinoid receptor activation in this effect. The chemopreventive effect of 4HPR was similar to 4MPR, suggesting that although 4HPR can activate retinoid receptors in cell lines, its action in the 2-stage model is primarily retinoid receptor independent. Interestingly, 4HPR and 4MPR act at early steps in skin carcinogenesis while ATRA acts to suppress tumor promotion. We next generated gene expression profiles of mouse skin treated with each retinoid at early stages of the 2-stage carcinogenesis protocol, using Affymetrix U74Av2 and 430 2.0 microarrays. The genes identified have been grouped into two categories. 1.) Genes co-regulated by 4HPR and 4MPR in TPA treated skin and 2.) Genes oppositely regulated (counteregulated) by ATRA and TPA. In category 1, out of the 12,422 genes represented on the U74Av2 array, 354 and 465 genes were significantly altered by 4HPR and 4MPR relative to TPA treatment alone, respectively (|fold change| >= 1.2, where P<0.05 for detecting fold change). Amongst these, 161 genes were regulated similarly by both 4HPR and 4MPR in mouse skin. In category 2, 463 genes were identified as either upregulated or downregulated by TPA (p>=0.05). Approximately 60% of these were counteregulated when ATRA was coadministered with TPA. We have since expanded the range of genes tested in this category to the full transcriptome using the 430 2.0 array, which contains probes representing 39,000 genes. We propose that 4HPR and other phenylretinamides exert therapeutic and chemopreventive effects during early steps in skin carcinogenesis through retinoid receptor independent mechanisms, while ATRA acts through the retinoid receptors to suppress the tumor promoting effects of TPA. Biometric analysis of microrray data sets will be discussed in the context of the biological effects of each of the retinoids. Supported by funds from the Feist-Weiller Cancer Center, NIEHS P30 ES07784 and NCI CA78560.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]