Turnover of palmitate, arachidonate and glycerol in phospholipids of rat rod outer segments

Retinal de novo lipogenesis coordinates neurotrophic signaling to maintain vision

Membrane lipid composition is central to the highly specialized functions of neurological tissues. In the retina, abnormal lipid metabolism causes severe forms of blindness, often through poorly understood neuronal cell death. Here, we demonstrate that deleting the de novo lipogenic enzyme fatty acid synthase (FAS) from the neural retina, but not the vascular retina, results in progressive neurodegeneration and blindness with a temporal pattern resembling rodent models of retinitis pigmentosa. Blindness was not rescued by protection from light-evoked activity; by eating a diet enriched in palmitate, the product of the FAS reaction; or by treatment with the PPARα agonist fenofibrate. Vision loss was due to aberrant synaptic structure, blunted responsiveness to glial-derived neurotrophic factor and ciliary neurotrophic factor, and eventual apoptotic cell loss. This progressive neurodegeneration was associated with decreased membrane cholesterol content, as well as loss of discrete n-3 polyunsaturated fatty acid- and saturated fatty acid-containing phospholipid species within specialized membrane microdomains. Neurotrophic signaling was restored by exogenous cholesterol delivery. These findings implicate de novo lipogenesis in neurotrophin-dependent cell survival by maintaining retinal membrane configuration and lipid composition, and they suggest that ongoing lipogenesis may be required to prevent cell death in many forms of retinopathy.

Transport of phosphatidylcholine to Xenopus photoreceptor rod outer segments in the presence of tunicamycin

Study of the dynamics of membrane protein and phospholipid transport from the inner to the outer segment of vertebrate photoreceptors has shown an interesting dissociation of the two components under a number of experimental treatments which inhibit protein synthesis or transport. Under conditions which block the addition of opsin to outer segments, various lipids continue to be synthesized and transported to the outer segment in the presence of monensin, puromycin, brefeldin A, tunicamycin and several general metabolic inhibitors. In the current study, isolated retinas from adult Xenopus laevis were incubated with or without 20 micrograms mg-1 of tunicamycin in total darkness or light for 2-12 h in the presence of [3H]choline to study the dependence of phosphatidylcholine synthesis and transport on protein transport to the outer segment. Phosphatidylcholine is a major bulk lipid of outer segments, comprising close to one half of the phospholipid of outer segment phospholipids, and blocking choline uptake in retinas is known to cause photoreceptor degeneration. Biochemical analysis demonstrates that tunicamycin does not block the synthesis of phosphatidylcholine in photoreceptor inner segments or transport of radiolabelled phosphatidylcholine to outer segments during 6 h incubations with [3H]choline in light or total darkness. Light and electron microscopic autoradiography and morphometric analysis show that [3H]choline radiolabelled phospholipid does not accumulate in a band of newly formed basal discs in the outer segment or in the tubulo-vesicular structures which accumulate in the intersegmental space of tunicamycin-treated retinas. We conclude that transport of phosphatidylcholine can occur independently of opsin transport to the outer segment but whether this represents two separable components of a single pathway or involves two distinct routes of transport to the outer segment is still unresolved.

Tunicamycin does not inhibit transport of phosphatidylinositol to Xenopus rod outer segments

Tunicamycin inhibits the dolichol pathway for N-linked glycosylation of proteins, including photoreceptor opsin, and causes a buildup of tubulo-vesicular profiles in the intersegmental space between photoreceptor rod inner and outer segments associated with disruption of new disc assembly. We tested the hypothesis that a tunicamycin lesion in photoreceptors would block lipid transport into the outer segment. Adult Xenopus retinas were preincubated in dim red light with 20 micrograms ml-1 of tunicamycin for one hour followed by incubation in the light for 2-6 h with tunicamycin plus either [3H]mannose, [3H]leucine, [2-(3)H]glycerol or [3H]myo-inositol. Tunicamycin caused accumulation of tubulo-vesicular membranes in the intersegmental space and significantly reduced both [3H]leucine and [3H]mannose incorporation into the basal region of rod outer segments. However, tunicamycin had no effect on [3H]glycerol incorporation into the rod outer segment phospholipids. After 5 h incubation with [3H]glycerol, radiolabel in outer segment fractions was associated primarily with phosphatidylinositol in both control and tunicamycin treated retinas. Quantitative light microscope autoradiography of both [3H]glycerol and [3H]inositol labelled retinas showed diffuse labelling over the entire rod outer segment in both control and tunicamycin treated retinas with no accumulation of radioactivity in the basal discs of control retinas or in the tubulo-vesicular structures in the intersegmental space of tunicamycin treated retinas. Our results indicate that despite the morphological disruption and inhibition of glycoprotein transport to outer segments after tunicamycin treatment, transport of labelled phosphatidylinositol occurs normally. These data add to a growing body of evidence separating the lipid and protein transport pathways to the outer segment.

Phospholipids in Drosophila heads: effects of visual mutants and phototransduction manipulations

A procedure was developed to label phospholipids in Drosophila heads by feeding radioactive phosphate (32Pi). High-performance thin-layer chromatography showed label incorporation into various phospholipids. After 24 h of feeding, major phospholipids labeled were phosphatidylethanolamine (PE), 47%; phosphatidylcholine (PC), 24%; and phosphatidylinositol (PI), 12%. Drosophila heads have virtually no sphingomyelin as compared with mammalian tissues. Notable label was in ethanolamine plasmalogen, lysophosphatidylethanolamine, lysophosphatidylcholine and lysophosphatidylinositol. Less than 1% of the total label was in phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate. Other lipids labeled included phosphatidylserine, phosphatidic acid and some unidentified lipids. A time course (3-36 h) study revealed a gradual decrease in proportion of labeled PI, an increase in proportion of labeled PC and no obvious change in labeled PE. There were no significant differences in phospholipid labeling comparing the no receptor potential (norpA) visual mutant and wild type under light vs. dark conditions. However, overall 32P labeling was higher in the wild type fed in the light as compared to the dark and to norpA either in light or dark. This suggests that functional vision facilitates incorporation of label. Differences in phospholipid labeling were observed between young and aged flies, particularly in lysophospholipids and poly-PI, implicating phospholipase A2 function in recycling. v Manipulations such as the outer rhabdomeres absent and eyes absent mutants and carotenoid deprivation failed to yield notable differences in phospholipid labeling pattern, suggesting that phospholipids important to vision may constitute only a minor portion of the total labeled pool in the head.

Lipids of frog retinal pigment epithelium: comparison with rod outer segments, retina, plasma and red blood cells

The glycerolipid and fatty acid compositions of frog retinal pigment epithelium (RPE) were determined and compared with rod outer segments (ROS), retina, plasma, and red blood cells (RBC). The glycerolipid class composition of RPE was similar to RBC and ROS or retina, with phosphatidylcholine and phosphatidylethanolamine being the major components. The fatty acid composition of RPE differed substantially from that of plasma or RBC; the former contained much higher levels of C-20 and C-22 polyunsaturated fatty acids (PUFAs), such as 20:4n-6 and 22:6n-3, but less C-18 mono-, dienoic, and trienoic acids. The difference between RPE and ROS or retina with respect to fatty acid profile was also dramatic; RPE had relatively less 22:6n-3, but more 20:4n-6 and 18:2n-6, than ROS or retina. These results suggest that frog RPE cells may selectively take up C-20 and C-22 PUFAs from the circulation, but preferentially deliver 22:6n-3 to the ROS and retina. Fatty acid analyses show that 20:4n-6 and 22:6n-3 were unevenly distributed among RPE glycerolipids; phosphatidic acid, diglyceride, triglyceride, and phosphatidylserine are relatively more enriched in 22:6n-3 compared with 20:4n-6. This information might imply that these two PUFAs are metabolized differently inside the frog RPE cells.

Metabolism of linolenic acid and docosahexaenoic acid in rat retinas and rod outer segments

Docosahexaenoic acid (22:6 omega 3) is uniquely enriched in photoreceptor outer segment phospholipids, comprising up to one-half of the fatty acids of phosphatidylethanolamine and phosphatidylserine. The current study was designed to investigate the incorporation of 22:6 omega 3 into outer segment phospholipids over 12 days and to determine whether the retina contained the enzymes necessary for elongation and desaturation of the major dietary precursor of 22:6 omega 3, the essential fatty acid linolenic acid (18:3 omega 3). Sprague-Dawley rats were injected intravitreally with [14C]22:6 omega 3 or [14C]18:3 omega 3 and kept in cyclic light (12 hr light/12 hr dark) for 2 hr to 12 days. Phospholipids from rod outer segments and the remaining retinal debris were separated by two-dimensional thin-layer chromatography. [14C]22:6 omega 3 radioactivity was initially highest in phosphatidylcholine and rapidly decreased from 45% of total phospholipid labeling at 2 hr to 26% by 1 and 3 days in ROS, while phosphatidylethanolamine labeling increased from 49 to 68% by 3 days and phosphatidylserine labeling increased from 3 to 14% over 12 days. Phenacyl derivatives of total fatty acids were separated by HPLC. A substantial conversion of [14C]18:3 to [14C]20:5, [14C]22:5 and [14C]22:6 was noted after 1 days, with increasing conversion to [14C]22:6 over the 12-day period. When only one eye was injected with [14C]18:3 omega 3, negligible radioactive fatty acids were detected in the contralateral eye from 1 to 12 days post-injection demonstrating that conversion of 18:3 to 22:6 occurred primarily within the injected eye. All enzymes for elongation and desaturation of 18:3 to 22:6 appear to be present in the eye. However, the conversion of 22:5 to 22:6 by delta-4 desaturase is evidently rate-limiting and may affect phospholipid replacement during photoreceptor outer segment renewal if this pathway proves to be essential for the supply of 22:6 during disk membrane formation.

Phospholipid molecular species from isolated bovine rod outer segments incorporate exogenous fatty acids at different rates

The incorporation of radiolabeled palmitic (16:0), oleic (18:1), and docosahexaenoic (22:6) acids into different molecular species of membrane phospholipids was investigated in isolated bovine rod outer segments (ROS). Phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylserine (PS) were isolated, and their diacylglyceroacetate and diacylglycerobenzoate derivatives were prepared, separated by HPLC, quantified, and assayed for radioactivity. Maximal incorporation of fatty acids occurred within 15-30 min. The rate of incorporation of the fatty acids into PC was three to six times higher than it was into PS or PE. The rate of incorporation of 22:6 into the molecular species, 22:6-22:6, of PC was ten to 15 times higher than into that of PE or PS, and it was three to four times higher than the incorporation rates of 22:6 into the other 22:6-containing molecular species. Similarly, incorporation of 18:1 into 18:1-22:6 was ten to 30 times more rapid in PC than in PE and PS, but in both PE and PS, 18:1 was incorporated into 18:1-22:6 at a rate of 20 to 25 times higher than the incorporation into the other molecular species analysed. For PC, incorporation of 16:0 was most rapid into 16:0-16:0, but for PE and PS it was most rapid into 16:0-20:4; for all cases, incorporation of 16:0 into these molecular species was four to six times more rapid than into the other 16:0-containing molecular species. These results are further evidence for the presence within a membranous organelle, the ROS, of an active acylation-deacylation system that is selective with regard to phospholipid, molecular species of phospholipid, and fatty acid.

Metabolism of lipid molecular species in rat rod outer segments

We have studied the metabolism of selected diacyl molecular species of phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), phosphatidylinositol (PI) and diacylglycerol (DG) from rat rod outer segments (ROS). Rats were injected intravitreally with [2-3H]glycerol. At 1, 2, 3, 4 and 6 days post-injection, ROS phospholipids and DG were isolated by two-dimensional thin-layer chromatography (TLC), derivatized, and fractionated into molecular species by high-performance liquid chromatography (HPLC). Selected molecular species were quantitated and counted for radioactivity. We found the following. In PC and PE, the specific activities of 22:6-22:6, 18:1-22:6 and 16:0-22:6 were highest at day 1 and then decreased in a nearly exponential manner. In contrast, the specific activities of 18:0-22:6 and 18:0-20:4 were substantially lower than these three molecular species and changed little over the 6-day period. In PS, the specific activities of 22:6-22:6, 18:0-22:6 and 18:1-22:6 were similar and did not reach their maximum until the 3rd or 4th days. In PC, the specific activities of the five molecular species examined were two to three times higher at day 1 than the same species in PE and PS. In PI and DG, the major molecular species were 16:0-20:4 and 18:0-20:4. The specific activities of these two molecular species at day 1 were about ten times higher than those of 20:4-containing species in PE and PC, and showed the most rapid turnover of any of the molecular species examined in this study.(ABSTRACT TRUNCATED AT 250 WORDS)

The localization and timing of post-translational modifications of rat rhodopsin

Rat retinas were labeled by incubation with, or intravitreal injection of, [14C]leucine along with tritiated palmitic acid, glucosamine or galactose. At selected intervals, subcellular fractions were prepared on linear sucrose gradients and rhodopsin was extracted and purified by affinity chromatography and gel electrophoresis. Time courses revealed that leucine rapidly and transiently labeled the rhodopsin in the rough endoplasmic reticulum (RER), with a maximum at 1.5 hr post-injection. Subsequently, the rod outer segments (ROS) contained the labeled rhodopsin, with the ROS labeling maximally at 6-12 hr. Palmitate labeling followed the same pattern but was subject to a delay, presumably because of a large intracellular pool of the fatty acid. With palmitate the RER rhodopsin was not maximally labeled until 12 hr. The acylation of rhodopsin takes place in the RER sometime after the polypeptide has been translated but before transport to the Golgi. Glucosamine labeling was also delayed because of intracellular pools of the sugar or its metabolic derivatives. But because of secondary glycosylation in the Golgi, the rhodopsin in the ROS also labeled maximally with glucosamine at about 6 hr. Administration of [3H]galactose resulted in the labeling of rhodopsin both in vivo and in vitro, in part possibly because of its conversion to mannose and subsequent insertion into the core oligosaccharide on the RER. However, in the ROS the ratio of tritium, derived from [3H]galactose, to [14C]leucine decreased by a factor of 2 between 6 and 24 hr post-injection. Moreover, between 6 and 12 hr post-injection, labeled rhodopsin molecules in the ROS underwent a shift in mobility on gels indicative of trimming to a lower molecular weight. Thus some sugar residues may be added to the rhodopsin in the inner segment and removed in the ROS.

Phosphoinositide metabolism in frog rod outer segments

Previous studies have shown that vertebrate rod outer segments (ROS) have a light activated phospholipase C which hydrolyzes phosphatidylinositol-4,5-bisphosphonate (PIP2). Three different experimental approaches have been used to test the hypothesis that the phosphatidylinositol (PI) biosynthetic cycle is present in ROS and that PIP2 can be regenerated from DG independent of rod inner segments. In the first study, enzyme activities of the PI cycle were assayed simultaneously in the presence of CTP, myo-inositol and [gamma-32P]ATP using endogenous lipids as substrates. Under these conditions, broken (leaky) ROS prepared by continuous sucrose gradient centrifugation showed PI, PIP and DG kinase activities similar to those found in intact ROS and non-ROS membranes, whereas PI synthetase activity was much lower in the leaky ROS than in the other two fractions. The relative distribution of PI synthetase specific activity in the three membrane preparations was similar to that of the microsomal enzyme marker cytochrome c reductase. ROS prepared by discontinuous sucrose gradient centrifugation showed only 2-3% of whole homogenate PI synthetase or phosphatidyl: cytidyl transferase activities, and the distribution of activities was the same as for microsomal and mitochondrial marker enzymes. In the second study, whole retinas were incubated with myo-[2-3H]inositol or [2-3H]glycerol in vitro, and the time course of incorporation of radioactivity into PI and other phospholipids was determined for ROS and three other retinal fractions. Over a 10-hr period, the rate of incorporation of myo-[2-3H]inositol or [2-3H]glycerol into PI in ROS was lowest among the various retinal fractions. In the third study, chemical analysis of the molecular species composition of PI, DG and phosphatidic acid (PA) from ROS shows that PA is substantially different from PI and DG, the latter two being quite similar. These results are consistent with a precursor-product relationship between PI and DG, but not with the conversion of DG to PA or of PA to PI. Taken together, these three studies indicate that ROS do not have PI synthetase or phosphatidyl: cytidyl transferase activities, but do have DG, PI and PIP kinase activities. Thus, the PI in ROS lost through rapid turnover must be replaced with molecules derived from de novo synthesis in the inner segment of the photoreceptor cell.

Metabolic labeling of rod outer segment phospholipids in miniature poodles with progressive rod-cone degeneration (prcd)

The recessive genetic defect in miniature poodles which results in progressive rod-cone degeneration (prcd) has been investigated in an attempt to determine the biochemical abnormality involved. In the present study, the rod outer segments of young prcd affected miniature poodles and normal dogs have been compared with respect to the incorporation of intravitreally injected [3H]palmitic acid. [14C]linolenic acid, and [14C]docosahexaenoic acid into neutral lipids and phospholipids as well as [3H]palmitate and [14C]leucine into rhodopsin. In addition, 3 mm trephined punches of retinas were incubated with [3H]palmitic acid, [3H]arachidonic acid, [14C]linolenic acid, [3H]serine, [14C]glycerol and [14C]leucine. No difference in incorporation of labeled precursors into lipids or rhodospin was noted between prcd affected and normal retinas. Phosphatidyl choline appeared to function as a carrier of fatty acids to the rod outer segment where they were redistributed to other phospholipids. An interesting lack of conversion of the essential fatty acid linolenic acid to docosahexaenoic acid was noted in both normal and affected retinas. This conversion involves elongation and desaturation of linolenic acid and may take place primarily in extraretinal tissues such as the liver. This finding, in conjunction with a parallel study of plasma fatty acids which has shown significantly lower levels of docosahexaenoic acid in prcd affected poodles, points to a possible systemic defect in the metabolism or transport of docosahexaenoic acid, a fatty acid uniquely enriched in the photoreceptor outer segments.

Addition of the chromophore to rat rhodopsin is an early post-translational event

Rat retinas were labeled either by intravitreal injection of [14C]leucine or by incubation with [3H]-leucine or [35S]-methionine. Subcellular fractions were prepared on linear sucrose gradients and rhodopsin was extracted with detergent and purified by chromatography on ConA-Sepharose. A fraction enriched in rough endoplasmic reticulum (RER) and substantially free of rod outer segments (ROS) was found to contain a light-sensitive protein exhibiting the properties of rhodopsin on ConA-Sepharose or Agarose chromatography and on SDS-polyacrylamide gel electrophoresis, as well as immunologically. Intravitreal injection of [3H]-retinol also labeled the rhodopsin in the RER under conditions in which the rhodopsin in the ROS was not heavily labeled. Thus the chromophore appears to be attached to opsin shortly after the apoprotein is translated in the RER.

Light exposure stimulates arachidonic acid metabolism in intact rat retina and isolated rod outer segments

This paper presents evidence that short-term exposure to light increases synthesis of hydroxyeicosatetraenoic acid (HETE) and prostaglandin D2 (PGD2) and stimulates the uptake and metabolism of 20:4 in phospholipids and triacylglycerols in rat retina. There was a time-dependent increase in incorporation of 1-14C-20:4 into glycerolipids in both dark-adapted and light-exposed groups. Exposure to light for 15 or 30 min enhanced the acylation of 20:4 into phosphatidylcholine, phosphatidylinositol, phosphatidylethanolamine and triacylglycerols. In the light-exposed groups there was a large increase in the conversion of 20:4 to leukotriene B4, two diHETEs, 5-HETE, 15-HETE, and PGD2. The stimulation of HETE synthesis by light could be due to early stages of light-induced lipid peroxidation in visual cells. To examine this, we studied peroxidation of 20:4 in isolated rod outer segments (ROS). There was more oxidation of 20:4 in light-exposed ROS, as compared to ROS incubated in the dark. Vitamin E and nordihydroguaiaretic acid inhibited the light-induced formation of some of these products. The data indicate that photo-oxidation of 20:4 in ROS is accompanied by enzymatic oxygenation that is stimulated by light. Increased production of HETEs an PGD2 may be a consequence of the light-induced stimulation of the metabolism of 20:4 in membrane phospholipids in the retina.

Metabolic labeling of normal canine rod outer segment phospholipids in vivo and in vitro

Twenty-four hours after the intravitreal injection of [3H]palmitate and [14C]docosahexaenoate in dogs, the rod outer segment phospholipids are highly labeled. Palmitate is found predominantly in phosphatidylcholine (PC), with lesser amounts in phosphatidylethanolamine (PE) and very little in either phosphatidylserine (PS) or phosphatidylinositol (PI). Docosahexaenoate most heavily labeled PE followed by PC, with lesser amounts in PS and very little in PI. Two-hour incubations of 3 mm trephine buttons removed from dog retinas produced very similar patterns of labeling with palmitate and docosahexaenoate. In vitro incubation of retina buttons with [3H]arachidonate produced heavy labeling of PI, with much less in PC and very little in either PS or PE. [3H]Glycerol labeled in PC, PI and PE in descending order but PS almost not at all. [3H]Serine labeled PS predominantly, but small amounts were found in PC, PE and PI. The trephine retina buttons can be utilized for multiple-precursor incubations and studies of differential metabolism in retinal regions, particularly when studying scarce tissue from mutant animals or humans with inherited retinal degenerations.

The acylation of rat rhodopsin in vitro and in vivo

Rat retinas were incubated with [3H] palmitate and [14C] leucine and subsequently detergent-extracted. Glycoproteins were isolated on Con A-Sepharose columns and separated by gel electrophoresis. Leucine labeled the newly synthesized opsin but palmitate was esterified to both mature rhodopsin and newly synthesized opsin which migrated more slowly because of its untrimmed oligosaccharide chains. Crude rod outer segments were found to contain most of the palmitate-labeled mature rhodopsin, while the retinal debris contained most of the doubly labeled newly synthesized opsin. Homogenization of double-labeled retinas followed by centrifugation in a linear sucrose gradient gave rise to several bands of particles including purified rod outer segments, a Golgi-enriched fraction and a pellet enriched in endoplasmic reticulum. Newly synthesized opsin was first found in the pellet at the earliest incubation times and subsequently appeared in the Golgi fraction and finally in the rod outer segments. Palmitate-labeled mature rhodopsin was found only in the rod outer segments. It appeared at the earliest time points and increased with time. Thus the acylation of opsin occurs in the endoplasmic reticulum, shortly after polypeptide synthesis, and in the rod outer segments, the latter possibly as an exchange reaction. Most of the newly synthesized opsin remained in the pellet and did not pass through the Golgi to the rod outer segments. Intravitreal injection of [3H] palmitate and [14C] leucine gave rise to doubly labeled opsin that appeared to remain untrimmed for at least 6 hr in vivo. After 17 hr, both labels were found only in mature rhodopsin, thus accumulation of new molecules in the endoplasmic reticulum may occur in vivo. In addition, leucine maximally labeled the opsin-rhodopsin pool early in the first day whereas palmitate did not maximally label rhodopsin until 2- or 3 days post injection. Moreover, while leucine label was lost at day 9 because of rod outer-segment renewal and shedding, the palmitate label in rhodopsin remained unchanged. Thus, palmitate labeling in vivo reflects the pattern seen in vitro with a prolonged equilibration of rod outer-segment rhodopsin with the fatty-acid pool, probably mediated by a fatty acyl exchange reaction.