Retinol metabolism in LLC-PK1 Cells. Characterization of retinoic acid synthesis by an established mammalian cell line

Specific assays, based on gas chromatography-mass spectrometry and high-performance liquid chromatography, were used to quantify the conversion of retinol and retinal into retinoic acid by the pig kidney cell line LLC-PK1. Retinoic acid synthesis was linear for 2-4 h as well as with graded amounts of either substrate to at least 50 microM. Retinoic acid concentrations increased through 6-8 h, but decreased thereafter because of substrate depletion (t1/2 of retinol = 13 h) and product metabolism (1/2 = 2.3 h). Retinoic acid metabolism was accelerated by treating cells with 100 nM retinoic acid for 10 h (t1/2 = 1.7 h) and was inhibited by the antimycotic imidazole ketoconazole. Feedback inhibition was not indicated since retinoic acid up to 100 nM did not inhibit its own synthesis. Retinol dehydrogenation was rate-limiting. The reduction and dehydrogenation of retinal were 4-8-fold and 30-60-fold faster, respectively. Greater than 95% of retinol was converted into metabolites other than retinoic acid, whereas the major metabolite of retinal was retinoic acid. The synthetic retinoid 13-cis-N-ethylretinamide inhibited retinoic acid synthesis, but 4-hydroxylphenylretinamide did not. 4'-(9-Acridinylamino)methanesulfon-m-anisidide, an inhibitor of aldehyde oxidase, and ethanol did not inhibit retinoic acid synthesis. 4-Methylpyrazole was a weak inhibitor: disulfiram was a potent inhibitor. These data indicate that retinol dehydrogenase is a sulfhydryl group-dependent enzyme, distinct from ethanol dehydrogenase. Homogenates of LLC-PK1 cells converted retinol into retinoic acid and retinyl palmitate and hydrolyzed retinyl palmitate. This report suggests that substrate availability, relative to enzyme activity/amount, is a primary determinant of the rate of retinoic acid synthesis, identifies inhibitors of retinoic acid synthesis, and places retinoic acid synthesis into perspective with several other known pathways of retinoid metabolism.     

Biogenesis of retinoic acid from beta-carotene. Differences between the metabolism of beta-carotene and retinal

The ability of beta-carotene to serve as precursor to retinoic acid was examined in vitro with cytosol prepared from rat tissues. The rate of retinoic acid synthesis from 10 microM beta-carotene ranged from 120 to 224 pmol/h/mg of protein with intestinal cytosol, and from 344 to 488 pmol/h/mg of protein with cytosols prepared from kidney, lung, testes, and liver. Retinol generated during beta-carotene metabolism was not the major substrate for retinoic acid synthesis. At low substrate concentrations (2.5 microM), the rates of retinoic acid synthesis in intestinal cytosol from beta-carotene or retinol were equivalent, and at higher concentrations (10 microM) the rates of retinoic acid synthesis from beta-carotene or retinol in intestine, testes, lung, and kidney were comparable. Thus, beta-carotene metabolism may be an important source of retinoic acid in retinoid target tissues, particularly in species such as humans that are capable of accumulating high concentrations of tissue carotenoids. Retinal, considered an initial retinoid product of beta-carotene metabolism, was not detected as a product of beta-carotene metabolism in vitro. A ratio of retinol and retinoic acid different from that observed during beta-carotene metabolism in vitro was observed with incubations of retinal under identical conditions. These data indicated that beta-carotene metabolism is not merely a simple process of producing retinal and releasing it into solution to be metabolized independently.     

Efficacy validation of synthesized retinol derivatives In vitro: stability,toxicity, and activity

Retinol (vitamin A) is used as an antiwrinkle agent in the cosmetics industry. However, its photo-instability makes it unsuitable for use in general cosmetic formulations. To improve the photo-stability of retinol, three derivatives (3, 4, and 5) were synthesized and their biological activities were analyzed. 1H NMR and HPLC analysis indicated that derivatives 3 and 5 were much more stable than retinol under our sunlight exposure conditions. When human adult fibroblasts were treated, the IC(50) of derivative 3 was 96 microM, which is similar to that of retinol, as determined by the MTT assay. Derivatives 4 and 5 were 2.5 and 8 times more toxic than retinol, respectively. At 1 microM treatment, like retinol, derivatives 3 and 4 were specifically active for RARalpha out of six retinoid receptors (RAR/RXRalpha, beta, gamma). Dose-dependent analysis confirmed that derivative 4 was as active as retinol and the other two derivatives were less active for RARalpha. The effect of our derivatives on the expression of collagenase, an indicator of wrinkle formation, was measured using the transient co-expression of c-Jun and RT-PCR in HaCaT cells. Collagenase promoter activity, which is increased by c-Jun expression, was reduced 42% by retinol treatment. The other derivatives inhibited collagenase promoter activity similarly. These results were further confirmed by RT-PCR analysis of the collagenase gene. Taken together, our results suggest that retinol derivative 3 is a promising antiwrinkle agent based on its higher photo-stability, lower RARalpha activity (possibly indicating reduced side effects), and similar effect on collagenase expression.     

Vitamin A effects on UMR 106 osteosarcoma cells are not mediated by specific cytosolic receptors.

Retinol and retinoic acid at 20 microM altered cell morphology and inhibited cell proliferation of UMR 106 osteosarcoma cells in culture. No specific cytosolic binding proteins for retinol could be detected.     

The biosynthesis of retinoic acid from retinol by rat tissues in vitro

This report shows that a spectrum of vitamin A-dependent tissues can produce retinoic acid by synthesis in situ, indicates that cellular retinol and retinoic acid binding proteins are not obligatory to retinoic acid synthesis, and provides initial characterization of retinoic acid synthesis by rat tissues. Retinoic acid synthesis from retinol was detected in homogenates of rat testes, liver, lung, kidney, and small intestinal mucosa, but not spleen. Zinc did not stimulate the conversion of retinol into retinoic acid by liver homogenates. Retinoic acid synthesis was localized in cytosol of liver and kidney, where its rate of synthesis from retinol was fourfold (liver) and sevenfold (kidney) slower than from retinal. The synthesis of retinoic acid from retinol required NAD and was not supported by NADP. NADH (0.5 mM) reduced retinoic acid synthesis from retinol, supported by NAD (2 mM), by 50-70%, but was fivefold less potent in reducing retinoic acid synthesis from retinal. Dithiothreitol enhanced the conversion of retinol, but not retinal, into retinoic acid. EDTA inhibited the conversion of retinol into retinoic acid slightly (13%, liver; 29%, kidney). A high ethanol concentration (100 mM), relative to retinoid substrate (10 microM), inhibited retinoic acid synthesis from retinol (liver, 54%; kidney, 30%) and from retinal (30%, liver; 9%, kidney). 4'-(9-Acridinylamino)methansulfon-m-anisidine, an inhibitor of aldehyde oxidase, and disulfiram, a sulfhydryl-group crosslinking agent, were potent inhibitors of retinoic acid synthesis at 10 microM or less, and seemed equipotent in liver and kidney. 4-Methylpyrazole, an inhibitor of ethanol metabolism, also inhibited retinoic acid synthesis from retinol, but was less potent than the former two inhibitors, and affected liver to a greater extent than kidney, particularly with retinal as substrate.     

Vitamin A and bone formation. Different responses to retinol and retinoic acid of chick bone cells in organ culture

The response of bone cells in organ culture to retinol and retinoic acid was studied. Both stimulated incorporation of [3H]thymidine into DNA by 16-day embryonic chick calvaria, but the time-course of the responses differed; the peak responses to retinol and retinoic acid occurred at about 18 h and 48 h, respectively. Although retinol inhibited chick bone collagen synthesis retinoic acid had no effect, but it did stimulate non-collagenous protein synthesis, whereas the effect on the latter of retinol was, if anything, inhibitory. When present with retinol, retinoic acid was able to attenuate the inhibitory effect of the former on chick bone collagen synthesis, but preincubation with retinoic acid had no such effect. In neonatal murine calvarial cultures, retinoic acid inhibited collagen synthesis selectively in the same manner as did retinol. The ability of chick osteoblasts to respond differently to retinol and retinoic acid suggests that both forms of the vitamin may have a role in bone formation and that their intracellular models of action may differ although the attenuation response indicates there may be some interaction between the two.     

Characterization of the effect of retinol on Plasmodium falciparum in vitro

A preliminary study from our laboratory found retinol (vitamin A alcohol) to have in vitro activity against Plasmodium falciparum at concentrations close to those in normal human serum (1-3 microM). To characterize the antimalarial potential of retinol in more detail, the 3D7 and K1 laboratory strains of P. falciparum were maintained in continuous culture and [3H]hypoxanthine incorporation and microscopy were used to assess the effect of retinol against asexual stages of the parasite life-cycle. Losses of retinol and retinol-associated hemolysis were also quantified in the in vitro culture system. There were retinol losses of >50% but no hemolysis was observed with added retinol concentrations up to 100 microM. All stages of parasite development showed comparable sensitivity to retinol including merozoite invasion (range of mean IC50 values 10.1-21.4 microM after adjustment for losses). Retinol pre-treatment of uninfected RBC did not inhibit merozoite invasion. Retinol treatment was associated with increased vacuolization within the parasite food vacuole and evidence of parasite membrane rupture. These appearances were similar to those seen with quinoline and artemisinin compounds. Although these data do not support a role for acute retinol supplementation in the treatment of falciparum malaria, they add to knowledge regarding potential antimalarial therapies and justify assessment of more potent synthetic retinoids and their metabolites.     

Microsomes convert retinol and retinal into retinoic acid and interfere in the conversions catalyzed by cytosol

Rat liver microsomes converted retinol into retinal and retinoic acid. The production of retinal was observed over a range of substrate concentrations (10-100 microM), but retinoic acid was detected only at retinol concentrations of 50 microM or higher. At 50 microM retinol, the rate of microsomal retinal production was 2-fold greater than that of cytosol, but the rate of retinoic acid synthesis was 4-fold less than that of cytosol. Retinal was also converted into retinoic acid by rat liver microsomes, but at a rate 2-5% of that catalyzed by cytosol. Microsomes also interfered with the conversion of retinol and retinal into retinoic acid by rat liver cytosol. A 50% decrease in the cytosolic rates of retinoic acid production from retinol or retinal was caused by microsomal to cytosolic protein ratios of 0.1 and 0.5, respectively. Under the incubation conditions, which included NAD in the medium, addition of microsomes to cytosol did not affect the elimination half-life of retinol or retinoic acid, but did decrease the elimination half-life of retinal by 2-fold. These data show that retinal synthesis from retinol does not necessarily reflect retinoic acid synthesis and suggest that liver microsomes sequester free retinol and convert it into retinal primarily for elimination, rather than to serve as substrate for cytosolic retinoic acid synthesis.     

Effect of retinol and retinoic acid on permeability,electrical resistance and phase transition of lipid bilayers

Retinol and retinoic acid have been incorporated into the artificial membrane systems, planar bimolecular lipid membranes and liposomes, and their effects on several membrane parameters have been measured. 1. Retinol and retinoic acid increased the permeability of egg lecithin liposomes to K⁺, I^? and glucose when incorporated into the membranes at levels as low as 0.5 membrane mol%. Retinoic acid influenced permeability more than did retinol for each of the solutes tested. 2. Retinol and retinoic acid both decreased the electrical resistance of egg lecithin-planar bimolecular lipid membranes from 0.5 to 8 membrane mol%. Retinoic acid effected a larger change than did retinol. 3. Retinol and retinoic acid increased the permeability of dimyristoylphosphatidylcholine and dipalmitoylphosphatidylcholine liposomes to water at 1.0 and 3.0 membrane mol%. A larger effect on water permeability was measured for retinoic acid than for retinol. 4. Retinol and retinoic acid at 1.0 and 3.0 membrane mol% were shown to lower the phase-transition temperature of liposomes composed of dimyristoylphosphatidylcholine or dipalmitoylphosphatidylcholine. Phase-transition temperatures were monitored by abrupt changes in water permeability and liposome size associated with the transition. Retinoic acid lowered the phase-transition temperature of dimyristoylphosphatidylcholine liposomes more than did retinol, while both retinoids had almost the same effect on dipalmitoylphosphatidylcholine liposomes.     

Vitamin A and bone formation. Effect of an excess of retinol on bone collagen synthesis in vitro.

The influence of an excess of retinol on bone formation was studied by using cultures of embryonic-chick calvaria. Retinol decreased collagen synthesis in a dose-dependent manner, non-collagenous protein synthesis being relatively unaffected. Collagen synthesis was significantly inhibited after 24 h of culture with retinol and was progressively decreased, compared with control cultures containing no retinol, as the period of culture was increased. The effect of retinol on collagen synthesis could be reversed by incubation of calvaria for further periods in retinol-free medium. Incorporation of [3H]thymidine and [3H]uridine into DNA and RNA respectively was not altered by culturing calvaria with retinol for 22 h. These latter findings, and the selectivity for collagen synthesis, all suggested that the effect observed was not a cell-toxicity phenomenon. The effect of retinol on collagen synthesis by chick calvarial osteoblasts was probably direct and not mediated by osteoclasts, since a negligible number of the latter cells is present in chick calvaria. In cultures of neonatal murine calvaria, which contain many osteoclasts, retinol similarly inhibited synthesis of collagen, but not of non-collagenous protein; the concentrations of retinol necessary to produce the response were similar to those required to stimulate bone resorption in vitro.     

Separable binding proteins for retinoic acid and retinol in bovine retina.

Binding proteins for retinoic acid and retinol were separated from a supernatant prepared from bovine retina. Fraction IV from DEAE-cellulose chromatography bound exogenous [3H] retinoic acid which could not be effectively displaced by retinol, retinal, retinyl acetate or palmitate, but which was readily displaced with excess retinoic acid. [3H] Retinol was bound by fraction V from DEAE-cellulose chromatography and was not displaced by retinal, retinoic acid, retinyl acetate or retinyl palmitate, but was readily displaced by excess retinol. Unlike bovine serum retinol-binding protein, neither intracellular binding protein formed a complex with purified human serum prealbumin. The supernatant from bovine retinas was estimated to contain five times more retinoic acid binding than retinol binder.     

Effects of Vitamin A and Its Analogs on Nonenzymatic Lipid Peroxidation in Rat Brain Mitochondria

Vitamin A (retinol) and some of its analogs exhibited varying degrees of inhibition on induced iron and ascorbic acid lipid peroxidation of rat brain mitochondria. Malonyldialdehyde production was used as an index of the extent of in vitro lipid peroxidation. The fat-soluble vitamins retinol, retinol acetate, retinoic acid, retinol palmitate, and retinal at concentrations between 0.1 and 10.0 mmol/L inhibited brain lipid peroxidation. Retinol and retinol acetate were the most effective inhibitors. It is concluded from this study that retinol and its analogs can be considered as potential antioxidant factors, more potent than some of the well-known antioxidants such as alpha-tocopherol and butylated hydroxytoluene.     

Separable binding proteins for retinoic acid and retinol in bovine retina

Binding proteins for retinoic acid and retinol were separated from a supernatant prepared from bovine retina. Fraction IV from DEAE-cellulose chromatography bound exogenous [³H] retinoic acid which could not be effectively displayed by retinol, retinal, retinyl acetate or palmitate, but which was readily displaced with excess retinoic acid. [³H] Retinol was bound by fraction V from DEAE-cellulose chromatography and was not displaced by retinal, retinoic acid, retinyl acetate or retinyl palmitate, but was readily displaced by excess retinol. Unlike bovine serum retinol-binding protein, neither intracellular binding protein formed a complex with purified human serum prealbumin. The supernatant from bovine retinas was estimated to contain five times more retinoic acid binding than retinol binder.     

Retinoic acid synthesis by cytosol from the alcohol dehydrogenase negative deermouse

Cytosolic alcohol dehydrogenase in the deermouse is coded by a single genetic locus and a strain of the deermouse which is alcohol dehydrogenase negative exists. These two strains of the deermouse were used to extend insight into the role of cytosolic alcohol dehydrogenases in the conversion of retinol into retinoic acid. Retinoic acid synthesis from physiological concentrations of retinol (7.5 microM) with cytosol from the alcohol dehydrogenase negative deermouse was 13% (liver), 14% (kidney), 60% (testes), 78% (lung), and 100% (small intestinal mucosa) of that observed with cytosol from the positive deermouse. The rates in the negative strain ranged from 0.3 to 0.7 nmol/h/mg protein: sufficient to fulfill cellular needs for retinoic acid. Ten millimolar 4-methylpyrazole inhibited retinoic acid synthesis 92, 94, 26, and 30% in kidney, liver, lung, and testes of the positive deermouse, respectively, but only 50, 30, 0, and 0% in the same tissues from the negative deermouse. Ethanol (300 mM) did not inhibit retinoic acid synthesis in kidney cytosol from the negative strain. Therefore multiple cytosolic dehydrogenases, including alcohol dehydrogenases, contribute to retinol metabolism in vitro. The only enzyme(s) likely to be physiologically significant to retinoic acid synthesis in vivo, however, is the class of dehydrogenase, distinct from ethanol dehydrogenase, that is common to both the positive and the negative deermouse. This conclusion is supported by the data described above, the kinetics of retinoic acid synthesis and retinal reduction in kidney cytosol from the negative deermouse, and the very existence of the alcohol dehydrogenase negative deermouse. This work also shows that microsomes inhibit the cytosolic conversion of retinol into retinoic acid and that the synthesis of retinal, a retinoid that has no known function outside of the eye, does not reflect the ability or capacity of a sample to synthesize retinoic acid.     

A commentary on the inhibition by retinoids of leukotriene B4 production in leukocytes

The order of potency of retinoids as inhibitors of A23187-induced production of leukotriene B4 (LTB4) in human polymorphonuclear leukocytes (PMN) was retinoic acid greater than retinal greater than retinol. However, the conversion of exogenous arachidonate (AA) to LTB4 by PMN homogenates was inhibited in the rank order retinol greater than retinal much greater than retinoic acid. The agreement between active concentrations of retinol in these two systems is consistent with this compound acting directly to inhibit AA metabolism: this is not so for the other retinoids. The order of potency for inhibition of phorbol dibutyrate (PDBu)-stimulated superoxide (O-2) production in HL60 granulocytes was retinol greater than retinoic acid much greater than retinal (inactive); neither retinol nor retinal displaced [3H]PDBu from HL60 cells. We conclude that inhibition of LTB4 production by retinoic acid and retinal is neither through inhibition of AA metabolism nor through inhibition of protein kinase C.     

Retinol supplementation in murine Plasmodium berghei malaria: effects on tissue levels, parasitaemia and lipid peroxidation

Reduced plasma retinol concentrations occur in human malaria but the benefits of supplementation remain uncertain. We assessed the in vivo efficacy of retinol administration, and its effect on lipid peroxidation, in a Plasmodium berghei murine model. Animals received vehicle (n=17) or retinol (i) before P. berghei inoculation (four doses), (ii) at parasitaemia 10-15% (three to four doses) or (iii) before and after inoculation (six to seven doses; n=15 in each group), with euthanasia on day 8 post-inoculation or when the parasitaemia exceeded 50%. Multiple-dose pre-inoculation retinol reduced endpoint parasitaemia by 24% (P=0.001 versus controls). A reduction of 18% (P=0.042) was observed when retinol was given to parasitaemic animals. Retinol was ineffective when given both before and after infection (11% reduction; P=0.47). Although retinol supplementation did not change plasma retinol concentrations, liver retinol content increased and correlated inversely with endpoint parasitaemia (r=-0.45, P=0.001). Malaria infection augmented concentrations of the free radical lipid peroxidation end-product F(2)-isoprostanes in plasma, erythrocytes and liver by 1.8-, 2.8- and 4.9-fold, respectively, but retinol supplementation had no effect on these increases. Consistent with some human malaria studies, prophylactic retinol reduces P. berghei parasitaemia. This effect relates to augmentation of tissue retinol stores rather than to retinol-associated changes in oxidant status.     

Genetic dissection of retinoid dehydrogenases

Biochemical studies indicate that alcohol dehydrogenase (ADH) metabolizes retinol to retinal, and that aldehyde dehydrogenase (ALDH) metabolizes retinal to retinoic acid, a molecule essential for growth and development. Summarized herein are several genetic studies supporting in vivo functions for ADH and ALDH in retinoic acid synthesis. Gene targeting was used to create knockout mice for either Adh1 or Adh4. Both knockout mice were viable and fertile without obvious defects. However, when wild-type and Adh4 knockout mice were subjected to vitamin A deficiency during gestation, the survival rate at birth was 3.3-fold lower for Adh4 knockout mice. When adult mice were examined for production of retinoic acid following retinol administration, Adh1 knockout mice exhibited 10-fold lower retinoic acid levels in liver compared with wild-type, whereas Adh4 knockout mice differed from wild-type by less than 2-fold. Thus, Adh1 plays a major role in the metabolism of a large dose of retinol to retinoic acid in adults, whereas Adh4 plays a role in maintaining sufficient retinol metabolism for development during retinol deficiency. ALDHs were examined by overexpression studies in frog embryos. Injection of mRNAs for either mouse Raldh1 or Raldh2 stimulated retinoic acid synthesis in frog embryos at the blastula stage when retinoic acid is normally undetectable. Overexpression of human ALDH2, human ALDH3, and mouse Aldh-pb did not stimulate retinoic acid production. In addition, Raldh2 knockout mice exhibit embryonic lethality with defects in retinoid-dependent tissues. Overall, these studies provide genetic evidence that Adh1, Adh4, Raldh1, and Raldh2 encode retinoid dehydrogenases involved in retinoic acid synthesis in vivo.     

DNA repair and replication in human fibroblasts treated with (+/-)-r-7,t-8-dihydroxy-t-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene

DNA repair and replication were examined in diploid human fibroblasts after treatment with (+/-)-r-7,t-8-dihydroxy-t-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE-I). Unscheduled DNA synthesis exhibited a linear response to BPDE-I concentrations up to 1.5 microM and a saturation plateau after higher concentrations. Maximal unscheduled DNA synthesis was observed in the first hour after treatment with synthesis diminishing progressively thereafter. Half-maximal unscheduled DNA synthesis was seen within 4-6 h after treatment with 0.7 microM BPDE-I. DNA replication was inhibited by BPDE-I in a dose- and time-dependent fashion. The mechanisms of this inhibition were characterized by velocity sedimentation of pulse-labeled nascent DNA in alkaline sucrose gradients. Very low concentrations of BPDE-I (0.03 and 0.07 microM) were found to inhibit replicon initiation by up to 50% within 30-60 min after treatment. Recovery of initiation following these low concentrations was evident within 3 h after treatment. Higher concentrations of carcinogen inhibited DNA synthesis in active replicons. This effect was manifested by a reduction in incorporation of precursor into replication intermediates of greater than 1 X 10(7) Da with the concurrent production of abnormally small nascent DNA. When viewed 45 min after treatment with 0.17 microM BPDE-I the combination of these two effects partially masked the inhibition of replicon initiation. However, even after treatment with 0.33 microM BPDE-I an effect on initiation was evident. These results reveal a pattern of response to BPDE-I that is quite similar to that produced by 254 nm radiation.