C S Silverman-Jones

Biosynthetic studies on mannolipids and mannoproteins of normal and vitamin A-depleted hamster livers

The incorporation of [1-14C]mannose into hamster liver glycolipids and glycoproteins was studied in normal and vitamin A-depleted hamsters. Severly (25% weight loss) and mildly (no weight loss) deficient animals were compared to vitamin A-fed controls. The incorporation of [14C]mannose into glycolipids and glycoproteins decreased in mild and severe vitamin A deficiency by 63-90% compared to vitamin A-fed animals. These results were essentially the same whether expressed per g of wet liver, per DNA or per protein. The size of the pools of mannose, glucose and galactose and their specific radioactivity in liver were determined by gas-liquid chromatography of the boronates of the hexitols (Eisenberg, Jr, F. (1972) Methods Enzymol. XXVIIIB, 168-178) in normal and vitamin A-deficient conditions. It was found that the amount of free hexoses per g of liver was very similar in normal and vitamin A-deficient conditions. The specific radioactivities for mannose and glucose were greater in vitamin A deficiency, thus excluding the possibility that the observed severe decrease in glycopeptide and glycolipid synthesis is a reflection of a similar decrease in the specific radioactivity of the precursor pools. Quantitation of mannose in glycoprotein showed a 79% decrease in vitamin A deficiency. Specific radioactivity of mannose in glycoproteins, 20 min after injection of the label, was 187 dpm/mug of mannose in the normal and 48 kpm/mug of mannose in the vitamin A-deficient livers. It is concluded that vitamin A is necessary for the biosynthesis of liver mannose-containing glycoproteins and glycolipids.

RECENT DEVELOPMENTS IN STUDIES ON BIOLOGICAL FUNCTIONS OF VITAMIN A IN NORMAL AND TRANSFORMED TISSUES

Abstract A biochemical pathway of phosphorylation and glycosylation of vitamin A has recently been found in hepatic, intestinal and epidermal tissues. More recent work suggests that mannosylretinylphosphate functions as a donor of mannose to membrane glycoconjugates. These reactions might ultimately explain the effects of vitamin A deficiency and some of the effects of excess vitamin A on biological systems. Studies of the effect of retinoids on cellular in vitro systems showed an increase in the adhesive properties of spontaneously-transformed mouse fibroblasts in culture (Balb/c 3T12–3 cells). These cells are usually detached from the culture dish surface in an EDTA adhesion assay. After culturing in presence of 3.3 x 10-6 to 3.3 x 10-5M retinol or retinoic acid the cells are no longer lifted from the plate and their morphology and adhesion resemble those of normal fibroblasts. This phenomenon of increased adhesion is observed as early as two days after exposure to the retinoid and it is readily reversible upon culturing in medium without exogenous retinoid. A variety of retinoids was tested in the adhesion assay. The most active compounds were retinol, retinyl-phosphate, retinoic acid, 5,6-epoxyretinoic acid and the TMMP and DACP derivatives of retinoic acid. All these compounds possess biological activity in other systems. Anhydroretinol, perhydromonoeneretinol, the phenyl derivative of retinoic acid, which do not have biological activity in other systems, did not increase adhesion of 3T12 cells. Other polyprenoid compounds without vitamin A activity were also tested in this assay. Dolichol, dolichylphosphate juvenile hormone, abscisic acid, β-ionone, dibutyryl cyclic adenosine monophosphate and sodium butyrate did not induce adhesion. The mechanism by which retinol and retinoic acid increase the adhesive properties of 3T12 cells was investigated. Cyclic adenosine monophosphate and guanosine monophosphate levels were not significantly altered by retinoid treatment at least at 6, 24, 48 and 72 hours after treatment with 3.3 x 10-5M retinoic acid, when most of the cells remain attached. Retinoic acid stimulated the incorporation of (2-3H) mannose into glycoproteins of 3T12 cells. (11, 123H and carboxyl-14C)Retinoic acid was incorporated into a compound (Metabolite I) which had chromatographic properties of a glycosylretinylphosphate. The synthesis of this compound was time-dependent and was not carried out by formalin -fixed 3T12 cells. Mild alkaline conditions which release anhydroretinol from retinylphosphate, also cleaved Metabolite I to yield a product with the polarity of a hydrocarbon, but slightly more polar than anhydroretinol. It is suggested that retinoic acid can be reduced to an alcohol, probably after metabolic modification. It is further suggested that such “retinol-like” compound would follow the same route of phosphorylation and glycosylation as shown for retinol in other systems. Microsomes from 3T12 cells were active as the intact cells in synthesizing mannosylretinylphosphate and dolichyl mannosylphosphate. Exogenous retinylphosphate specifically stimulated the synthesis of mannosylretinylphosphate. Thus it appears that vitamin A is involved in glycosyl transfer reactions in the 3T12 system, as well as in normal membranes. It remains to be established whether the observed increased adhesion is the result of such involvement. A novel reaction for retinol was found in 3T12 cells. Up to 55% of exogenously supplied retinol was converted to the hydrocarbon anhydroretinol in 48 hours. The same reaction was also carried out by microsomes from 3T12 cells, which converted 7% of retinol to anhydroretinol in 30 minutes at 37°C. This reaction may well represent a detoxification mechanism for the transformed cell.

Recent studies on the involvement of retinyl phosphate as a carrier of mannose in biological membranes

Rat liver microsomes synthesized [14C]mannosylretinylphosphate and dolichyl [14C]mannosylphosphate from guanosinedisphosphate [14C]mannose, retinylphosphate and dolichylphosphate. Two distinct enzyme activities were shown to be responsible for the biosynthesis of the two mannolipids. A higher affinity mannosyl transferase (EA I), responsible for dolichylmannosylphosphate synthesis, displayed a Km for GDP-mannose of 1.7 microM; while a lower affinity enzyme (EA II), responsible for mannosylretinylphosphate synthesis, displayed a Km for GDP-mannose of 12.5 microM. These Km values were unaffected by the addition of either dolichylphosphate for EA II, or retinylphosphate for EA I. The same Km values were found before and after solubilization of the enzyme activity with 1% Triton X-100. Differential solubilization of EA I and EA II was demonstrated, utilizing different concentrations of Triton X-100. Triple-labeled mannosylretinylphosphate was prepared from [3H]retinylphosphate, retinyl[32P]phosphate and GDP-[14C]mannose from incubations containing rat liver microsomes. This compound was shown to donate [14C]mannose to endogenous acceptors of rat liver microsomes.

Separation of mannosylretinylphosphate from dolichylmannosylphosphate by chromatography on columns of DEAE-Sephacel

Abstract A one-step column chromatographic procedure on DEAE-Sephacel allows the separation of mannosylretinylphosphate from dolichylmannosylphosphate with minimal breakdown of the mannosylretinylphosphate. Using this procedure, subcellular fractions of rat liver were shown to be active in synthesizing both mannolipids from GDP-[ 14 C]mannose in the absence or presence of exogenous retinylphosphate.

Retinoic acid treatment of fibroblasts causes a rapid decrease in [3H]inositol uptake

NIH 3T3 fibroblasts treated with all-trans-retinoic acid (RA) showed a dramatic decrease in the uptake of [3H]inositol compared to solvent-treated controls. The onset of RA-induced inhibition of [3H]inositol uptake was rapid with a 10-15% decrease occurring after 2-3 h of RA exposure and 60-70% reduction after 16 h of RA treatment. A progressive dose-dependent decrease in inositol uptake was found as the concentration of RA increased from 10(-8) to 10(-5) M and the effect was fully reversible within 48 h after RA removal. The Vmax and Kt for the controls were 10 nmol/2.5 x 10(6) cells/2 h and 51 microM; and for RA-treated cells the values were 4 nmol/2.5 x 10(6) cells/2 h and 52 microM. The decreased [3H]inositol uptake was not due to a change in the affinity (Kt) of the transporter for the inositol but to a decrease in the Vmax. The maximal effect on inositol uptake was dependent on RA treatment of the cells after they reached saturation density or if made quiescent by serum starvation. RA was the most active of the different retinoids examined in the order RA greater than 13-cis-RA = retinyl acetate greater than all-trans-retinol greater than 5,6-dihydroxyretinoic acid methyl ester greater than N-4-hydroxyphenyl retinamide. In contrast to this effect on inositol, the uptake of fucose, mannose, galactose, and glucose was either not affected or enhanced (for mannose and fucose) by RA treatment. RA inhibition of inositol uptake was also observed in 3T3-Swiss and Balb/3T3 cells but not in two virally transformed 3T3 cell lines. Phlorizin, amiloride, and monensin inhibited inositol uptake by 66, 74, and 58%, respectively, and this inhibition was additive when the cells were treated with RA as well as these inhibitors. A decreased incorporation of [3H]inositol into polyphosphoinositides was also observed in RA-treated cells but not to the same extent as for [3H]inositol uptake. In conclusion, RA treatment of 3T3 fibroblasts decreases the uptake of [3H]inositol by up to 70% within 8 to 10 h at near physiological concentrations in a reversible and specific manner.

METABOLITES OF RETINOL IN CULTURED HUMAN DERMAL FIBROBLASTS

Human individual variations in responsiveness to and metabolism of biologically active materials have been most extensively studied in cultured dermal fibroblasts. The effects of retinoids on mouse (Exp. Cell. Res. 119: 289-299, 1979; J. Lip. Res. 20: 357-362, 1979) and rat (Nature 276: 272-274, 1978) fibroblasts have been well characterized, especially in terms of parameters relating to malignant transformation. Spontaneously transformed mouse fibroblasts (3T12-3 cells) have been shown to actively metabolize retinol to the hydrocarbon anhydroretinol as well as retinyl esters (J. Lip. Res. 20: 357-362, 1979). Previous studies of cultured human dermal fibroblasts have demonstrated differential responses to retinol by cells from patients with Hurlers Syndrome (J. Exp. Med. 124: 1181-1198, 1966). Cellular retinoic acid binding has recently been identified in human fibroblasts (Lacroix, A., Anderson, G.D., and Lippman, M.E. Poster Paper, this volume), but two independent studies have failed to find in human fibroblasts cellular retinol-binding protein (Invest. Ophthalmal. 15: 1017-1022, 1976; Lacroix, A., et al. Poster Paper, this volume). Accordingly the present study was designed to identify the metabolites of retinol in dermal fibroblasts from four individuals. Cells were cultured with [15]4C]retinol in Hams F-12 medium containing 15% FCS for 72 hr. Cell pellets were extracted with absolute methanol and the lipid extract analyzed by thin-layer chromatography on silica gel in toluene/chloroformlmethanol (4/1/1) yielding three main peaks of radioactivity: one at ROO, constituted less than lo%, one in the area of retinol constituted * 40% and a main peak in the area of retinyl esters 50% of total). The identity of the various metabolities was verified by an improved reverse phase HPLC procedure (Anal. Biochem. 102: 243-248, 1980). Retinyl fatty acid esters represented 40-50% of the total lipid-extracted radioactivity. Retinol, the second most abundant species, represented 30-40% of the total. A third peak of highly polar metabolite(s) eluted in the area of mannosylretinylphosphate represented * 4-5% of the radioactivity. The 13-cis-retinol isomer of retinol detected by HPLC accounted for 4-5% of the label. No form of retinoic acid was detected. Comparative studies of the metabolism of retinoids by fibroblasts from patients with genetically determined skin cancer are now in progress.