Retinol metabolism and lecithin:retinol acyltransferase levels are reduced in cultured human prostate cancer cells and tissue specimens

Recent studies from our laboratory have indicated that the metabolism of vitamin A (retinol) to retinyl esters, carried out primarily by the enzyme lecithin:retinol acyltransferase (LRAT), is greatly reduced in human carcinoma cell lines of the oral cavity, skin, breast, and kidney as compared with their normal epithelial counterparts. These studies suggest that human carcinoma cells are retinoid-deficient relative to normal epithelial cells. In this study, we examined the metabolism of [(3)H]retinol and [(3)H]retinoic acid (RA) in human prostate cancer lines and in primary cultures of human prostate epithelial cells. Normal cells esterified all of the [(3)H]retinol added to the cultures. In contrast, all seven prostate cancer cell lines and four primary cultures derived from prostatic adenocarcinomas metabolized only trace amounts of [(3)H]retinol to [(3)H]retinyl esters. Correlated with this relative lack of esterification of [(3)H]retinol by the cancer cells was loss of expression of LRAT protein, whereas normal cells expressed abundant levels of LRAT protein by Western analysis. The metabolism of [(3)H]RA was also examined in these prostatic cells. Two of the prostate cancer tumor lines, DU 145 and PJ-1, exhibited rapid metabolism of [(3)H]RA; in contrast, the other tumor lines or primary cultures metabolized [(3)H]RA at a much slower rate. We also found that the immortalization of normal human prostatic epithelial cells by SV40 T antigen led to a reduction in LRAT protein expression and esterification of [(3)H]retinol. Further transformation to tumorigenicity with the ras oncogene resulted in loss of detectable LRAT expression. Finally, we analyzed LRAT protein expression in tissue sections from six prostatectomy specimens by immunohistochemistry. LRAT protein was predominantly expressed in the basal cells of normal prostatic epithelium, whereas its expression was lost in prostate cancer. Collectively, these data implicate aberrant retinoid metabolism in the process of prostatic carcinogenesis.     

Metabolism of retinol and retinoic acid in N-methyl-N-nitrosourea-induced mammary carcinomas in rats

This study was conducted to examine the in vivo uptake and metabolism of natural retinoids by N-methyl-N-nitrosourea-induced mammary carcinomas. In this study, endogenous retinol and retinyl esters were present in normal mammary epithelial cells, but were undetectable in N-methyl-N-nitrosourea-induced mammary carcinomas in rats as determined by high-pressure liquid chromatography. No differences were found in plasma levels of retinol, in liver retinyl esters, or total content of vitamin A between tumor-bearing and control animals. Administered labeled retinol was taken up and esterified by normal mammary epithelial cells. Tumor-bearing rats were given injections i.p. of either [3H]retinol or [3H]retinoic acid. Radioactivity increased progressively with time in liver and other tissues except in breast tumor, where the uptake fluctuated over the 8 days after the injection of [3H]retinol; in mammary tumors practically no metabolism of [3H]retinol occurred, while in other tissues extensive esterification was detectable. In contrast, in animals given injections of [3H]retinoic acid, the uptake and metabolism of the label in the breast tumors paralleled with those found in other tissues. Neither the activity of acyl coenzyme A:retinol acyl transferase nor the activity of retinyl ester hydrolase was altered in the mammary tumor compared to the normal mammary gland. On the other hand, a significant decrease in the retinal oxidase activity was found in tumor tissue compared to normal mammary tissue. Since no esterification of [3H]retinol occurred in vivo despite the presence of acyl coenzyme A:retinol acyl transferase activity, it is possible that a specific defect in the cellular uptake of retinol may exist in N-methyl-N-nitrosourea-induced mammary carcinomas.     

Inhibition of retinoid toxicity by rat liver S9 fraction in cultured mammalian cells

The effects of retinol (Rol), retinyl acetate (RAc) or retinoic acid (RA) on growth of Chinese hamster V79 cells were studied. In addition, the effect of Rol on cell growth in the human lymphoid cell line RPMI no. 1788 was examined. Cultures were treated with retinoid alone, with retinoid plus S9 mix and with retinoid plus the supernatant of S9 fraction. Treatment of Rol, RAc or RA alone inhibited the growth of both cell lines as evidenced by a dose-dependent decrease of viable cell counts. Concurrent treatment of a retinoid with the well-known metabolic activation system S9 mix containing the rat liver S9 fraction plus enzyme cofactors resulted in a total elimination of the toxic effect of a retinoid. For instance, V79 cells treated with Rol, RAc or RA alone at the highest dose of 32 micrograms/ml, and human lymphoid cells treated with Rol at the highest dose of 24 micrograms/ml resulted in killing of over 90% of the cells, while addition of S9 mix to the cultures treated with such high doses of a retinoid showed no reduction of viable cells as compared with the controls. The supernatant of S9 fraction after high speed centrifugation also had dose-dependent protective effects against the toxicity of retinol in V79 cells. Finally, experiments using sucrose gradient centrifugation and [3H] Rol suggests that binding of a retinoid to its specific binding protein in the S9 mixture and supernatant greatly decreases or abolished the toxicity of free retinoid in cultured mammalian cells.     

Differential expression of the enzyme that esterifies retinol, lecithin:retinol acyltransferase, in subtypes of human renal cancer and normal kidney.

PURPOSE: Retinoids, a group of compounds, including vitamin A (retinol), and related metabolites, have been shown to regulate the growth and differentiation of many types of cells. IFN-alpha and either 13-cis-retinoic acid or liposomal all-trans retinoic acid have been used in the treatment of patients with metastatic renal cell carcinoma. We knew that samples from renal cell carcinomas contained greatly reduced levels of retinol and retinyl esters relative to samples from normal human kidney. This prompted us to examine the levels of LRAT (lecithin:retinol acyltransferase) protein in various subtypes of human kidney cancers relative to normal human kidney by immunohistochemistry. EXPERIMENTAL DESIGN: We examined 31 partial or radical nephrectomy specimens diagnosed with kidney tumors between 1997 and 1998. Representative paraffin-embedded tissue blocks from each tumor, with each containing adjacent nonneoplastic renal parenchyma, were used for immunohistochemical analysis with affinity purified antibodies to human LRAT protein. RESULTS: LRAT protein was detected at high levels in the epithelial cells in the tubules and the lining of Bowman's capsule in the glomeruli of normal, nonneoplastic kidney sections. Among the 31 tumors, there were 13 cases of conventional (clear cell) renal cell carcinoma (RCC; including 2 multilocular cystic RCCs), 7 papillary RCC, 6 chromophobe RCC, 1 RCC, unclassified, and 4 renal oncocytoma. All tumors showed diffuse immunoreactivity for LRAT. In each case, the staining was uniform throughout the tumor, with only minimal variation in the staining intensity between different areas. All 4 renal oncocytomas, 2 of 6 chromophobe RCCs, 1 conventional (clear cell) carcinoma, 1 RCC, unclassified, and 2 conventional RCCs, which were of the multilocular cystic-type stained strongly (3+) for LRAT. In contrast, the remaining conventional RCCs and the papillary RCC samples stained much less intensely for LRAT. Of the 10 tumors that stained 3+ for LRAT in the study, 9 were either benign tumors or tumors with low malignant potential. CONCLUSIONS: These data show that LRAT expression is higher in renal tumors with an indolent biological behavior. Additional studies will ascertain if LRAT possesses any prognostic or therapeutic role in renal cancer.     

Human melanomas of fibroblast and epithelial morphology differ widely in their ability to synthesize retinyl esters

Reduced retinyl ester synthesis has been associated with several forms of cancer; we therefore proposed studying melanoma development from the perspective of this biochemical pathway. Cultures of human melanoma cells with fibroblastoid morphology showed negligible retinyl ester synthesis; in sharp contrast, those with epithelioid morphology were capable of retinol esterification. Further, isolated proliferating epidermal melanocytes (HFSC/2) esterified retinol, whereas proliferating normal skin fibroblasts (F:CCD-1121.Sk) did not. A primary site cutaneous melanoma and its metastatic match (both of epithelioid morphology) were capable of retinol esterification, while a matched fibroblastoid tumor pair did not synthesize retinyl esters; nevertheless, LRAT (lecithin:retinol acyltransferase) protein was found in microsomal fractions from all four tumors. A mutation screen in the LRAT coding region and adjacent intronic sequences revealed several novel mutations in these melanomas as well as in HFSC/2 and F:CCD-1121.Sk cells: a single nucleotide polymorphism in exon 1(37A-->G), a silent mutation in exon 2a (188 A-->G/186 G-->A), and an insertion in the 5'UTR (9-10insC). CRBP-1 basal expression was present in the HFSC/2, and in both sets of matched tumor pairs; however, steady-state levels in the fibroblastoid melanoma pair were one-third that found in the epithelioid matched tumor pair. Co-culture of human primary site epithelioid melanoma with proliferating normal human skin fibroblasts abrogated retinol esterification within 96 h and increased the expression of the active form of TGFbeta-1 by 2.4-fold. A concomitant 3.2-fold downregulation of CRBP-1 expression took place. This is the first study to (1) demonstrate an association between retinyl ester synthesis and cutaneous melanoma morphological phenotypes; (2) suggest the existence of a soluble, diffusible inhibitor of the retinol esterification pathway; (3) report the ability of the isolated, proliferating human epidermal melanocyte to esterify retinol; and (4) provide evidence of DNA variants in the coding region of LRAT.     

Secretion of N-(4-hydroxyphenyl) retinamide-retinol-binding protein from liver parenchymal cells: Evidence for reduced affinity of the complex for transthyretin

The synthetic retinoid 4-HPR has been shown to markedly lower the plasma concentration of both retinol and RBP in rats and humans. We have studied the effect of 4-HPR on the secretion of retinol-RBP from liver cells in vivo and in vitro. In rats maintained with a normal diet, a vitamin A-deficient diet or a normal diet supplemented with 4-HPR, chylomicrons [³H]retinyl esters were rapidly cleared from the plasma. The secretion of chylomicron-derived [³H]retinol from tissues to the circulation, however, was different. In control rats, the lymph-derived [³H]retinol peaked after about 2 hr, whereas 4-HPR treatment effectively reduced this peak of [³H]retinol. Our results suggest that 4-HPR inhibits secretion of retinol-RBP from the liver. Therefore, we decided to study the effect of 4-HPR on the secretion of RBP using the human hepatoma cell line HepG2. Retinol and 4-HPR were found to induce the secretion of RBP. The medium from cells treated with 4-HPR was immunoprecipitated with antibodies against human RBP. HPLC analysis of the precipitated RBP revealed the presence of 4-HPR. When the medium from cells incubated with either 4-HPR or retinol was applied to a TTR affinity column, we found that RBP from cells incubated with 4-HPR had a considerably reduced affinity for TTR. We conclude that 4-HPR binds RBP and thereby induces secretion of RBP in HepG2 cells, and that the secreted 4-HPR-RBP complex has a reduced affinity for TTR. This observation may explain the 4-HPR-induced reduction of plasma retinol and RBP observed in in vivo studies. Int. J. Cancer 71:654-659, 1997. © 1997 Wiley-Liss, Inc.     

Retinoid metabolism in the prostate: effects of administration of the synthetic retinoid N-(4-hydroxyphenyl)retinamide

We have carried out a series of complementary in vivo and in vitro studies to better understand the metabolism of vitamin A by the prostate gland. Male Sprague-Dawley rats were fed either a control diet sufficient in vitamin A [CON group; 0.8 microg retinol equivalents (RE)/g diet] or a CON diet supplemented with the synthetic retinoid N-(4-hydroxyphenyl)retinamide (4-HPR; CON+/-4HPR group; 1,173 )microg of 4-HPR/g diet). After an i.v. injection of a physiological radiolabeled dose of retinol, the vitamin A content and radioactivity of plasma and a number of tissues, including the prostate glands, were monitored for time periods ranging between 30 min and 41 days. On the basis of the results of these vitamin A turnover studies, we developed tissue subsystem models to describe vitamin A dynamics in the prostates of both the CON and CON+4HPR groups. There was a gradual decrease in the vitamin A content of the prostates of the 4-HPR-treated group as compared with the control, such that by the end of the study period, the CON+4HPR group averaged 0.166 +/- 0.0827 (mean +/- SD) REs, whereas the CON group was 0.732 +/- 0.190 REs. The fraction of vitamin A exiting the prostate each day was not significantly different in the CON as compared with the CON+4HPR group [0.149 +/- 0.103 versus 0.155 +/- 0.191 h(-1) (mean +/- FSD), respectively]; however, the average amount of vitamin A turning over from the CON+4HPR group prostates (0.0885 /microg/day) was nearly three times less than that of the CON group (0.243 microg/day). To obtain more detailed information on the mechanisms that might be involved in the changes in vitamin A kinetics observed in our in vivo studies, we used both a normal human prostate cell line (PrEC) and a human prostate adenocarcinoma cell line (LNCaP) to monitor in vitro retinol and 4-HPR dynamics. Cells were treated with 4-HPR for different time periods up to 48 h (PrEC) or 96 h (LNCaP). Retinol in the media was taken up readily by both PrEC and LNCaP cells, and there was conversion of retinol to the major storage esters of vitamin A, retinyl palmitate and retinyl stearate, as well as several minor retinyl esters, in a pattern indicative of normal retinoid esterification activity. Although 4-HPR was taken up readily and over time accumulated in both cell lines, conversion of 4-HPR to its major metabolite, N-[4-methoxyphenyl]retinamide, as well as several other metabolites of 4-HPR was apparent only in the LNCaP cells. Our findings would suggest that a study design that includes appropriately designed complementary in vivo and in vitro experimental systems represents a useful approach to better understanding possible mechanisms involved in basic retinoid functioning and interactions in the prostate as well as in other organs and related tissue culture systems.     

Increased sensitivity of human keratinocytes immortalized by human papillomavirus type 16 DNA to growth control by retinoids.

Human papillomavirus (HPV) type 16 (HPV16) is associated with a large percentage of cervical malignancies, and HPV16 DNA can immortalize human keratinocytes in vitro. The transforming ability of the virus resides primarily in the open reading frames E6 and E7. Retinoids are potent modulators of growth and differentiation of keratinocytes and have been shown to reverse cervical lesions resulting from HPV infection. We compared the sensitivity of normal human foreskin keratinocytes (HKc) and four immortalized HKc lines, independently obtained by transfection of different normal HKc strains with HPV16 DNA (HKc/HPV16), to growth control by retinoic acid (RA). All the HKc/HPV16 lines were 10- to 100-fold more sensitive than normal HKc to growth inhibition by RA in both clonal and mass culture growth assays. The precursor to RA, retinol, was also found to be a more potent inhibitor of growth of HKc/HPV16 than normal HKc, while beta-carotene did not inhibit growth of either normal HKc or HKc/HPV16. In addition, HKc/HPV16 lines were more sensitive than normal HKc to modulation of keratin expression by RA and retinol. No differences were observed in the rate of uptake of [3H]RA or [3H]retinol between normal HKc and HKc/HPV16. Dot blot analysis of RNA extracted from HKc/HPV16 cultured in the absence or in the presence of 10(-7) M RA showed that the expression of the HPV16 open reading frames E6 and E7 is reduced 2- to 4-fold by RA. In addition, Northern blot analysis demonstrated that RA inhibition of E6 and E7 expression was both dose and time dependent. Overall, these results suggest that the increased sensitivity of the HKc/HPV16 lines to growth control by RA may be mediated by an inhibition of the expression of HPV16 gene products which are required for the maintenance of continuous growth.     

Inhibition by retinoids of in vitro invasive ability of human lung carcinoma cells

The effect of retinoid treatment of A549 human lung carcinoma cells on in vitro cell invasion using the human amnion basement membrane (BM) was investigated. A 2-day retinoid pretreatment of the cells resulted in a significant reduction in their invasive ability. The most effective retinoid, retinol acetate, inhibited cell migration through the BM and degradation of [3H] proline labeled BM components by 50% at noncytotoxic concentrations of 0.09, and 3 micrograms/ml, respectively. Inhibition by retinol acetate of A549 cell invasive potential was accompanied by a significant decrease in type IV collagenase activity and no change in transglutaminase activity.     

Biotransformation and protein binding of N-(4-hydroxyphenyl)retinamide in murine mammary epithelial cells

The biotransformation of N-(4-hydroxyphenyl)retinamide (HPR), and interactions of parent compound and/or metabolites with the cellular retinoid binding proteins (CRBPs) and cellular retinoic acid binding proteins (CRABPs) were examined in murine mammary tumor virus (MuMTV)-induced murine mammary tumor cells (GR-3A) grown in monolayer cell culture. Soluble fractions (cytosols) obtained from the extracts of GR-3A cells after high speed centrifugation were found to contain proteins of approx. 15,000 daltons which bound retinol and retinoic acid, but did not bind HPR or HPR metabolites. Moreover, HPLC analysis of GR-3A cell extracts demonstrated that [3H]retinol and [3H]retinoic acid were not detected in cells that had been exposed to [3H]HPR for 48 h. These findings, that under in vitro conditions there is no appreciable enzymatic hydrolysis of HPR to retinoic acid or conversion to retinol, suggests that the metabolism and cytological effects of HPR may be distinct from those of retinol or retinoic acid within murine mammary epithelial cells.     

Pharmacologic disposition of chemopreventive retinoids (review)

The disposition of natural and synthetic retinoids, some of which effectively prevent cancer in carcinogen-dosed animals, varies with the structure of the compound. Disappearance of retinoids from serum may be either linear or exponential, and there may be a prolonged terminal elimination phase. Metabolism of these compounds is extensive, with little or no absorbed retinoid being excreted unchanged. Compounds giving rise to retinol can be stored as retinyl esters in the liver. In contrast, retinoids with a terminal carboxyl group are not stored but are metabolized on the ring and/or on the sidechain prior to excretion, which occurs often after conjugation of the metabolites with glucuronic acid. Isomerization of the double bonds in the sidechain may also occur. The metabolites of natural retinoids do not appear to possess more biological activity than the parent compounds, leading to the conclusion that their metabolism is, in general, degradative.     

Photoreaction,Phototoxicity, and Photocarcinogenicity of Retinoids

Abstract

Sunlight is a human carcinogen. Many retinoid-containing cosmetics are used to protect damages caused by sunlight irradiation. Since retinol is thermally unstable and retinyl palmitate (RP) is relatively more stable, RP is also widely used as an ingredient in cosmetic formulations. In general, little is known about the photodecomposition of retinoids and the toxicity of retinoids and their photodecomposition products on the skin's responses to sunlight. This review focuses on the update information on photoreactions, phototoxicity, and photocarcinogenicity of the natural retinoids including retinol, retinal, retinoid acid (RA), retinyl acetate, and RP.     

Assay of retinoids in biological samples by reverse-phase high-pressure liquid chromatography

A separation procedure for retinoids based on reversephase high-pressure liquid chromatography with solvent mixtures of acetonitrile and water is described. The method may be applied to the screening of synthetic retinoids, which have potential for use in the prevention of cancer. It is easily adapted to a variety of biological samples and can be applied to other conventional retinoid assays in liver and plasma, detecting as little as 1 nmol retinyl esters and less than 0.3 nmol retinol per g tissue. The one-step chromatography results in separation and simultaneous determination of many of the synthetic retinoids and all of the natural retinoids, including the retinyl esters that are separated into their major fatty acid components. The method has been applied to the analysis of retinoid levels in the liver and intestine of vitamin A-deficient hamsters following a p.o. dose (0.5 mg/day for 2 days) of retinyl acetate or of a synthetic vitamin A analog and is predictive of the degree to which various synthetic retinoids can be converted to retinol and stored in the liver as retinyl esters. Because of its speed, excellent recoveries, and high resolution, the method offers significant advantages over previous, more lengthy procedures.     

Overview of current knowledge of metabolism of vitamin A and carotenoids

Current knowledge about the metabolism of naturally occurring retinoids and carotenoids was summarized. Dietary provitamin A carotenoids are largely converted to retinol (vitamin A) during intestinal absorption in the mucosal cell. In humans, a limited amount of carotenoids can be absorbed intact, along with retinyl esters (newly synthesized or from dietary vitamin A), mainly via lymph chylomicrons. Carotenoids are stored in several tissues, particularly liver and fat. They are transported in plasma by lipoproteins (density less than 1.21 g/ml), particularly by the low-density lipoproteins. Plasma carotenoids are usually a mixture of compounds with and without provitamin A activity; beta-carotene is about 20-25% of the total. Newly absorbed vitamin A is stored in the liver as retinyl esters. Storage involves both the hepatic parenchymal cells and the nonparenchymal stellate cells. Vitamin A is mobilized from liver stores and transported in plasma as retinol bound to a specific transport protein, retinol-binding protein (RBP). Retinol mobilization is highly regulated by factors that control the rates of RBP synthesis and secretion. Much is known now about the chemical structure, metabolism, and biologic roles of RBP, RBP delivers retinol to peripheral target tissues; delivery may involve cell surface receptors for RBP. Tissues of rats, humans, and other species contain soluble binding proteins with specificity for either retinol (cellular retinol-binding protein) or retinoic acid (cellular retinoic acid-binding protein). These intracellular proteins have been purified from several tissues and partly characterized. From both immunoassay and immunocytochemical studies, information is available about their tissue distribution and levels. Retinoic acid is mainly absorbed through the portal system and transported in plasma as the anion bound to serum albumin. Nonspecific and unregulated delivery of retinoids to biologic membranes apparently leads to vitamin A (retinoid) toxicity.     

Effects of dietary retinoids upon growth and differentiation of tumors derived from several murine embryonal carcinoma cell lines

We have examined the effects of dietary retinoids upon the growth and differentiation of seven embryonal carcinoma lines in mice. The control diet contained 4000 IU/mg retinyl palmitate; the other diets contained 2 x 10(5) IU/mg retinyl palmitate, 50 mg/kg all-trans-retinoic acid (RA), 100 mg/kg RA, and no retinoid. The RA-containing diets had little influence on tumor latency or incidence but did suppress growth of many of the tumors. Decreased tumor mass was, in most but not all instances, accompanied by an increased proportion of differentiated cells. Increased differentiation was most commonly quantitative rather than qualitative; i.e., there was a larger proportion of the same types of differentiated cells seen in tumors from the control diet group rather than an increase in the spectrum of cell types observed. Notably, tumors from two differentiation-defective embryonal carcinoma lines were refractory to both the differentiation-inducing and growth-suppressing properties of dietary RA. Taken together, our results suggest that dietary RA can reduce teratocarcinoma growth in part by promoting differentiation but that other mechanisms are likely to be involved. The therapeutic benefits that we observed with dietary RA were compromised by adverse effects, including failure of the mice to gain weight as effectively as those on the control diet. The effects of elevated levels of retinyl palmitate, or its omission from the diet, were much less striking than that of RA. Both modifications tended to decrease tumor latency but had little effect, if any, upon ultimate tumor mass. Elimination of retinoid from the diet failed to significantly reduce degree of differentiation in tumors which normally differentiate extensively in animals on retinoid-containing diets. Excess retinyl palmitate led to a marginal increase in differentiation in F9 tumors and a statistically significant increase in differentiation in OC15-S1 tumors. Tumors from other embryonal carcinoma lines did not contain elevated levels of differentiated cells. The interpretation of these results is complicated by our observations that although our dietary alteration in levels of palmitate were dramatic, they resulted in much more modest differences in circulating retinoid levels when compared with mice on the control diet.     

Vitamin A contents of rat intestinal epithelium and jejunal mucinous adenocarcinoma.

Concentrations of retinol and tetinyl esters were assayed in rat intestinal mucosa and in chemically induced transplanted mucinous adenocarcinoma of the jejunum. Lipid extract from the tissues was chromatographed on deactivated alumina to isolate retinol and retinyl esters, which were determined by specific spectrofluorometry. Normal intestinal mucosa tissue contains 556 ng of retinol equivalents as retinyl esters and 303 ng of free retinol/g of wet tissue. The concentration of retinyl esters in the intestinal mucosa from rats carrying the transplanted tumor was 341 ng/g wet tissue; no free retinol was detected in the small intestinal epithelium of these rats. Liver tissue from the tumor-bearing rats contained 157 microng of retinol equivalents as retinyl esters and 136 microng of free retinol/g of wet tissue. The concentration of vitamin A per cell in the adenocarcinoma tissue was about 20 times less than that in intestinal epithelium.