Diacylglycerol kinase epsilon regulates seizure susceptibility and long-term potentiation through arachidonoyl- inositol lipid signaling.

Arachidonoyldiacylglycerol (20:4-DAG) is a second messenger derived from phosphatidylinositol 4,5-bisphosphate and generated by stimulation of glutamate metabotropic receptors linked to G proteins and activation of phospholipase C. 20:4-DAG signaling is terminated by its phosphorylation to phosphatidic acid, catalyzed by diacylglycerol kinase (DGK). We have cloned the murine DGKepsilon gene that showed, when expressed in COS-7 cells, selectivity for 20:4-DAG. The significance of DGKepsilon in synaptic function was investigated in mice with targeted disruption of the DGKepsilon. DGKepsilon(-/-) mice showed a higher resistance to electroconvulsive shock with shorter tonic seizures and faster recovery than DGKepsilon(+/+) mice. The phosphatidylinositol 4,5-bisphosphate-signaling pathway in cerebral cortex was greatly affected, leading to lower accumulation of 20:4-DAG and free 20:4. Also, long-term potentiation was attenuated in perforant path-dentate granular cell synapses. We propose that DGKepsilon contributes to modulate neuronal signaling pathways linked to synaptic activity, neuronal plasticity, and epileptogenesis.

Alterations in lipid and calcium metabolism associated with seizure activity in the postischemic brain.

Transient ischemia is known to lead to a long-lasting depression of cerebral metabolic rate and blood flow and to an attenuated metabolic and circulatory response to physiological stimuli. However, the corresponding responses to induced seizures are retained, demonstrating preserved metabolic and circulatory capacity. The objective of the present study was to explore how a preceding period of ischemia (15 min) alters the release of free fatty acids (FFAs) and diacylglycerides (DAGs), the formation of cyclic nucleotides, and the influx/efflux of Ca(2+), following intense neuronal stimulation. For that purpose, seizure activity was induced with bicuculline for 30 s or 5 min at 6 h after the ischemia. Extracellular Ca(2+) concentration (Ca(2+)(e)) was recorded, and the tissue was frozen in situ for measurements of levels of FFAs, DAGs, and cyclic nucleotides. Six hours after ischemia, the FFA concentrations were normalized, but there was a lowering of the content of 20:4 in the DAG fraction. Cyclic AMP levels returned to normal values, but cyclic GMP content was reduced. Seizures induced in postischemic animals showed similar changes in Ca(2+)(e), as well as in levels of FFAs, DAGs, and cyclic nucleotides, as did seizures induced in nonischemic control animals, with the exception of an attenuated rise in 20:4 content in the DAG fraction. We conclude that, at least in the neocortex, seizure-induced phospholipid hydrolysis and cyclic cAMP/cyclic GMP formation are not altered by a preceding period of ischemia, nor is there a change in the influx/efflux of Ca(2+) during seizure discharge or in associated spreading depression.

Cortical impact injury in rats promotes a rapid and sustained increase in polyunsaturated free fatty acids and diacylglycerols.

Neurotrauma activates the release of membrane phospholipid-derived second messengers, such as free arachidonic acid (20:4n-6, AA) and diacylglycerols (DAGs). In the present study, we analyze the effect of cortical impact injury of low-grade severity applied to the rat frontal right sensory-motor cortex (FRC) on the accumulation of free fatty acids (FFAs) and DAGs in eight brain areas 30 min and 24 hours after the insult. At these times, accumulation of FFAs and DAGs occurred mainly in the damaged FRC. The cerebellum was the only other brain area that displayed a significant accumulation of DAGs by day one post-injury. By 30 min, accumulation of free AA in the FRC displayed the greatest relative increase (300% over sham value), followed by free docosahexaenoic acid (22:6n-3, DHA, 150%), while both 20:4-DAGs and 22:6-DAGs were increased 100% over sham values. At day one, free 22:6 and 22:6-DAGs showed the greatest increase (590% and 230%, respectively). These results suggest that TBI elicits the hydrolysis of phospholipids enriched in excitable membranes, targeting early on 20:4-phospholipids (by 30 min post- trauma) and followed 24 hours later by preferential hydrolysis of DHA-phospholipids. These lipid metabolic changes may contribute to the initiation and maturation of neuronal and fiber track degeneration observed following cortical impact injury.

Selective retinal pigment epithelial cell lipid metabolism and remodeling conserves photoreceptor docosahexaenoic acid following phagocytosis.

Retinal pigment epithelial cells (RPE) actively retrieve and recycle docosahexaenoic acid (DHA, 22:6n-3) from phagosomal phospholipids back to photoreceptor cells. Here we studied the fate of DHA in primary culture rat RPE cells after feeding with a suspension of rod outer segments (ROS) for 4 hr. Phospholipids (PLs), triacylglycerols (TAG), and free fatty acids were isolated from cells and media by thin layer chromatography (TLC), and their acyl groups quantified by gas liquid chromatography (GLC). In RPE cells, DHA-PLs increased 3. 5-fold by 4 hr, decreasing thereafter to 1.6-fold above basal by 24 hr. In contrast, 18:1-PLs were decreased by 13%-18% below RPE basal values by 8-24 hr, respectively. DHA-TAG showed the highest increase (21-fold) by 8 hr. Free DHA displayed a small increase in the cells with a preferential release and accumulation into the media by 24 hr. These results show that in rat RPE cells, photoreceptor cell DHA is transiently incorporated into TAG prior to its release and uptake into 18:1-PLs. These metabolic pathways and remodeling may be critical in the conservation of this essential, photoreceptor cell fatty acid.

Delayed phospholipid degradation in rat brain after traumatic brain injury.

Lipid second messengers such as arachidonic acid and its metabolites and diacylglycerols (DAGs) are affected in brain injury. Therefore, changes in the pool size and the fatty acid composition of free fatty acids (FFAs) and DAGs were analyzed in different rat brain areas 4 and 35 days after traumatic injury. Cortical impact injury of low-grade severity was applied in the right frontal somatosensory cortex. Four days after injury, FFAs and DAGs were increased by three- and twofold, respectively, in the injured cortex and to a lesser extent in the contralateral cortex compared with sham-operated animals. Docosahexaenoic acid followed by stearic acid, and arachidonic acid, displayed the greatest changes in both FFAs and DAGs. By day 35, free stearic, oleic, and arachidonic acids remained elevated in the damaged cortex (1.5-fold each). DAGs showed the greatest change, reaching values 2.7-fold higher than sham in all frontal and occipital cortical areas, including brainstem. Oleoyl- and arachidonoyl-DAGs (four- and threefold increase, respectively) followed by docosahexaenoyl-DAGs (twofold) contributed to the DAG accumulation. These results reveal that traumatic brain injury triggers a sustained and time-dependent activation of phospholipase-mediated signaling pathways leading to membrane phospholipid degradation and targeting, early on, docosahexaenoyl phospholipid-enriched excitable membranes.

Post-Golgi vesicles cotransport docosahexaenoyl-phospholipids and rhodopsin during frog photoreceptor membrane biogenesis.

Post-Golgi vesicles budding from the trans-Golgi network (TGN) are involved in the vectorial transport and delivery of rhodopsin to photoreceptor rod outer segments (ROS). We report here that newly synthesized docosahexaenoyl (DHA) phospholipids are sequestered and cotransported by rhodopsin-bearing post-Golgi vesicles to ROS. Frog retinas were pulse-labeled with [35S]methionine/cysteine and [3H]DHA prior to ROS isolation and subcellular fractionation. After a 1-h pulse, relatively uniform [3H]DHA-lipid labeling (DPM/microg protein) was observed in all fractions enriched in post-Golgi vesicles, TGN, Golgi, and endoplasmic reticulum (ER) membranes. During the subsequent 2-h chase translocation of free [3H]DHA from ROS to the photoreceptor inner segment contributed to an additional overall increase in labeling of lipids. The specific activity (dpm/nmol DHA) in ER-enriched fraction was similar or higher than in other subcellular fractions after both the pulse and the chase, indicating that the bulk of [3H]DHA-lipids was synthesized in the ER. After the chase a 2-fold increase in labeling of lipids in the ER and Golgi and a 2.6-fold in lighter TGN-enriched fractions was observed. The highest labeling was in the post-Golgi vesicle fraction (4-fold increase), with [3H]DHA-phosphatidylcholine and [3H]DHA-phosphatidylethanolamine showing the greatest increase. At the same time, newly synthesized [35S]rhodopsin shifted from the ER and Golgi toward TGN and post-Golgi fractions. Therefore, sequestration and association of [35S]rhodopsin and [3H]DHA-lipids in a TGN membrane domain occurs prior to their exit and subsequent vectorial cotransport on post-Golgi vesicles to ROS. Labeling of ROS lipids was very low, with phosphatidylinositol and diacylglycerols displaying the highest labeling. This indicates that other mechanisms by-passing Golgi, i.e. facilitated by lipid carrier proteins, may also contribute to molecular replacement of disc membrane DHA-phospholipids, particularly phosphatidylinositol.

Mediators of injury in neurotrauma: intracellular signal transduction and gene expression.

Membrane lipid-derived second messengers are generated by phospholipase A2 (PLA2) during synaptic activity. Overstimulation of this enzyme during neurotrauma results in the accumulation of bioactive metabolites such as arachidonic acid, oxygenated derivatives of arachidonic acid, and platelet-activating factor (PAF). Several of these bioactive lipids participate in cell damage, cell death, or repair-regenerative neural plasticity. Neurotransmitters may activate PLA2 directly when linked to receptors coupled to G proteins and/or indirectly as calcium influx or mobilization from intracellular stores is stimulated. The release of arachidonic acid and its subsequent metabolism to prostaglandins are early responses linked to neuronal signal transduction. Free arachidonic acid may interact with membrane proteins, i.e., receptors, ion channels, and enzymes, modifying their activity. It can also be acted upon by prostaglandin synthase isoenzymes (the constitutive prostaglandin synthase PGS-1 or the inducible PGS-2) and by lipoxygenases, with the resulting formation of different prostaglandins and leukotrienes. Glutamatergic synaptic activity and activation of postsynaptic NMDA receptors are examples of neuronal activity, linked to memory and learning processes, which activate PLA2 with the consequent release of arachidonic acid and platelet-activating factor (PAF), another lipid mediator. Both mediators may exert presynaptic and postsynaptic effects contributing to long-lasting changes in glutamate synaptic efficacy or long-term potentiation (LTP), PAF, a potential retrograde messenger in LTP, stimulates glutamate release. The PAF antagonist BN 52021 competes for receptors in presynaptic membranes and blocks this effect. PAF may also be involved in plasticity responses because PAF leads to the expression of early response genes and subsequent gene cascades. The PAF antagonist BN 50730, selective for PAF intracellular binding, blocks PAF-mediated induction of gene expression. A consequence of neural injury induced by ischemia, trauma, or seizures is an increased release of neurotransmitters, that in turn generates an overproduction of second messengers. Glutamate, a key player in excitotoxic neuronal damage, triggers increased permeation of calcium mediated by NMDA receptors and activation of PLA2 in postsynaptic neurons. NMDA receptor antagonists reduce the accumulation of free fatty acids and elicit neuroprotection in ischemic damage. Increased production of free arachidonic acid and PAF converges to exacerbate glutamate-mediated neurotransmission. These neurotoxic actions may be brought about by arachidonic acid-induced potentiation of NMDA receptor activity and decreased glutamate reuptake. On the other hand, PAF stimulates the further release of glutamate at presynaptic endings. The neuroprotective effects of the PAF antagonist BN 52021 in ischemia-reperfusion are due, at least in part, to an inhibition of presynaptic glutamate release. PAF also induces expression of the inducible prostaglandin synthase gene, and PAF antagonists selective for the intracellular sites inhibit this effect. The PAF antagonist also inhibits the enhanced abundance, due to vasogenic cerebral edema and ischemia-reperfusion damage, of inducible prostaglandin synthase mRNA in vivo. Therefore, PAF, an injury-generated mediator, may favor the formation of other cell injury and inflammation mediators by turning on the expression of the gene that encodes prostaglandin synthase.

Membrane lipid degradation is related to interictal cortical activity in a series of seizures.

Brain levels of free fatty acids (FFA) and diacylglycerols (DAG) rise rapidly with the onset of seizures, reflecting activation of phospholipases A2 (PLA2) and C (PLC), respectively. However, the ictal/interictal accumulation of FFA attenuates as recurrent seizures continue. To assess the role of neuronal activity in stimulating PLA2 and C, we compared FFA and DAG in rat cerebral cortex during recurrent ictal periods as a function of associated levels of interictal activity. Pentobarbital-anesthetized rats were paralyzed, ventilated with 30% O2 and subjected to periodic pentylenetetrazol seizures at intervals of 5 min. Animals were killed with focused-microwave irradiation during either the 3rd or 15th seizure. The rise in cortical FFA levels during early seizures for 20:4, 22:6, and 18:0 was 3.6-, 2.5-, and 2.2-fold greater, respectively, when adjacent interictal activity was intense as compared to weak activity. During late seizures, this difference dropped to 2.2-fold for 20:4, the only FFA that showed a significantly higher value between robust versus weak interictal activity. In contrast, accumulation of DAG during early and late seizures was observed only when adjacent interictal activity was high. These results indicate that the cortical accumulation of FFA and DAG during ictal periods of similar intensity and duration depends upon the electrocortical activity during adjacent interictal periods.

Review: pharmacological manipulation of docosahexaenoic-phospholipid biosynthesis in photoreceptor cells: implications in retinal degeneration.

Docosahexaenoic acid (22:6n-3, DHA) is derived in vertebrate animals from n-3 fatty acids present in the diet (i.e., alpha-linolenic acid, 18:3n-3 and/or other n-3-long chain polyunsaturated fatty acids) and is found in very high concentrations in phospholipids from membranes of the central nervous system. Disk membranes of photoreceptor outer segments and synaptic terminals display a preferential enrichment in DHA-phospholipids that appears to be necessary for normal excitable membrane functions. Because of the relevance of adequate DHA-phospholipid synthesis and sorting toward new assembled disk membranes and synaptic terminals, as well as the pathophysiological implications of abnormal DHA metabolism (including its synthesis, delivery to the retina, and incorporation into lipids by de novo and turnover pathways), we reviewed recent studies of: a) the preferential uptake and retention of DHA by photoreceptors and its metabolism as it is activated to DHA-CoA and incorporated preferentially into phospholipids; b) pharmacological manipulations using amphiphilic cationic drugs (i.e., propranolol) to show an active esterification of DHA into lipids via de novo synthesis; and c) perturbations in DHA metabolism in retinas from dogs with progressive rod-cone degeneration (prcd).

Docosahexaenoic acid is taken up by the inner segment of frog photoreceptors leading to an active synthesis of docosahexaenoyl-inositol lipids: similarities in metabolism in vivo and in vitro.

Retinal uptake and metabolism of docosahexaenoic acid (DHA) was studied in vivo in frogs 1, 2, and 6 hours after dorsal lymph sac injections of [3H]-DHA (50 microCi/g). Light microscope autoradiography and biochemical techniques were used to compare the profiles of cellular uptake and lipid labeling with those obtained from 6 hour [3H]-DHA retinal incubations (final DHA concentration, 0.11 and 25 microM). Light microscope autoradiography demonstrated that rod photoreceptor ellipsoids and synaptic terminals preferentially labeled both in vivo and in vitro conditions. Also, the cytoplasm and oil droplets of retinal pigment epithelial cells became very heavily labeled after 6 hours of in vivo labeling. Phosphatidic acid showed the highest labeling in one hour, while other phospholipids accumulated label throughout the 6 hours. At that time point, most label was recovered in phosphatidyl-ethanolamine (37%), phosphatidylcholine (27%), and phosphatidylinositol (16%), the latter displaying 1.6-fold higher labeling than phosphatidylserine. The profile of labeled lipids was similar to that obtained in vitro when the concentration of DHA was in the nanomolar range. Our results suggest that de novo lipid synthesis is a major route for esterification of [3H]-DHA into retinal lipids, giving rise to an early and rapid labeling of DHA-phosphatidylinositol, both in vivo and in vitro, when DHA is present at low concentrations. Furthermore, the profile of labeled retinal cells under in vivo conditions closely resembles in vitro DHA labeling.

Coupling among energy failure, loss of ion homeostasis, and phospholipase A2 and C activation during ischemia.

The objective of the present experiments was to correlate changes in cellular energy metabolism, dissipative ion fluxes, and lipolysis during the first 90 s of ischemia and, hence, to establish whether phospholipase A2 or phospholipase C is responsible for the early accumulation of phospholipid hydrolysis products. Ischemia was induced for 15-90 s in rats, extracellular K+ (K+e) was recorded, and neocortex was frozen in situ for measurements of labile tissue metabolites, free fatty acids, and diacylglycerides. Ischemia of 15- and 30-s duration gave rise to a decrease in phosphocreatine concentration and a decline in the ATP/free ADP ratio. Although these changes were accompanied by an activation of K+ conductances, there were no changes in free fatty acids until after 60 s, when free arachidonic acid accumulated. An increase in other free fatty acids and in total diacylglceride content did not occur until after anoxic depolarization. The results demonstrate that the early functional changes, such as activation of K+ conductances, are unrelated to changes in lipids or lipid mediators. They furthermore suggest that the initial lipolysis occurs via both phospholipase A2 and phospholipase C, which are activated when membrane depolarization leads to influx of calcium into cells.

Free fatty acid and diacylglycerol accumulation in the rat brain during recurrent seizures is related to cortical oxygenation.

Cerebral blood flow and oxygenation increase during the early seizures of a series, but the increase in cerebral blood flow attenuates during late seizures, sometimes resulting in decreased cortical oxygenation. Cortical free fatty acids (FFA) and diacylglycerols also increase during early seizures and the increase attenuates during late seizures. We analyzed the correlation between lipid accumulation and cortical O2 during periodic pentylenetetrazol-induced seizures. During early seizures, both FFA and diacylglycerols increased in the cerebral cortex, particularly arachidonate (20:4) and stearate (18:0). Changes in lipids were different during late seizures, depending on cortical O2 levels. An increase in cortical O2 during late seizures was associated with lower FFA levels compared with early seizures, and FFA levels recovered to basal levels during interictal periods. A decline in cortical O2 was associated with a further increase in FFA, which remained elevated during interictal periods. Our results indicate that periseizure lipid accumulation is related to cortical oxygenation.

Decreased electroconvulsive shock-induced diacylglycerols and free fatty acid accumulation in the rat brain by Ginkgo biloba extract (EGb 761): selective effect in hippocampus as compared with cerebral cortex.

The effect of Ginkgo biloba extract (EGb 761) treatment (100 mg/kg/day, per os, for 14 days) on electroconvulsive shock (ECS)-induced accumulation of free fatty acids (FFA) and diacylglycerols (DAG) was analyzed in rat cerebral cortex and hippocampus. EGb 761 reduced the FFA pool size by 33% and increased the DAG pool by 36% in the hippocampus. These endogenous lipids were unaffected in cerebral cortex. During the tonic seizure (10 s after ECS) the fast accumulation of FFA, mainly 20:4, was similar in sham- and EGb 761-treated rats, in both the cerebral cortex and hippocampus. However, further accumulation of free 18:0 and 20:4, observed in the hippocampus of sham-treated rats during clonic seizures (30 s to 2 min after ECS), did not occur in EGb 761-treated animals. The rise in DAG content triggered in the cortex and hippocampus by ECS was delayed by EGb 761 treatment from 10 s to 1 min, when values similar to those in sham animals were attained. Moreover, in the hippocampus the size of the total DAG pool was decreased by 19% during the tonic seizure. At later times, DAG content showed a faster decrease in EGb 761-treated rats. By 2 min levels of all DAG acyl groups decreased to values significantly lower than in sham animals in both cortex and hippocampus. This study shows that EGb 761 treatment affects, with high selectivity, lipid metabolism and lipid-derived second messenger release and removal in the hippocampus, while affecting to a lesser extent the cerebral cortex.

Pathways for the uptake and conservation of docosahexaenoic acid in photoreceptors and synapses: biochemical and autoradiographic studies.

Docosahexaenoic acid (22:6n-3) esterified into phospholipids represents by far the most prevalent fatty acid of rod photoreceptor disc membranes and synaptic terminals. During synaptogenesis and photoreceptor biogenesis, plasma lipoproteins, secreted mainly by the liver, are the main source of plasma 22:6n-3 for the central nervous system. This systemic route (the long loop) also operates in mature animals for morphogenesis and maintenance of excitable membranes (e.g., during constant renewal of photoreceptor disc membranes). When radiolabeled 18:3n-3, the dietary precursor of 22:6n-3, is systemically supplied to 3-day-old mouse pups, it is elongated and desaturated in the liver, leading to the synthesis of 22:6n-3-lipoproteins that shuttle the fatty acid through the bloodstream to retina and brain. When radiolabeled 22:6n-3 was used, a more efficient labeling of brain and retinal lipids was achieved. The retinal pigment epithelium is involved, not only in the uptake of 22:6n-3 from circulating lipoproteins in the choriocapillaris but also in the recycling of 22:6n-3 from degraded phagosomal phospholipids back to the inner segments of photoreceptors (the short loop), following each phagocytic event. An interplay among efficient 22:6n-6 delivery from the liver, selective uptake by retinal pigment epithelium photoreceptor cells, and avid retinal retention may contribute to the enrichment of excitable membranes of the retina with 22:6n-3-phospholipids.

EGb 761 inhibits stress-induced polydipsia in rats.

The effect of daily treatments with Ginkgo biloba extract (EGb 761, IPSEN, France) on body weight and water intake of rats was followed for 15 days. During this period, two groups of rats, under slight ether anesthesia, were intubated and fed either EGb 761 (100 mg/kg b.wt. in 5% ethanol) or, for sham controls, 5% ethanol alone (6.6 ml/kg b.wt.). The increase in body weight was similar for the control and experimental groups. However, during the same period of time, the water intake, ml water/g b.wt./24 h, increased 37% in the controls. In EGb 761-treated rats, water intake remained unchanged. This suggests that EGb 761 treatment inhibits the development of polydipsia due to the stress of daily handling and intubation.

Light stimulates in vivo inositol lipid turnover in frog retinal pigment epithelial cells at the onset of shedding and phagocytosis of photoreceptor membranes.

We have developed an experimental model to study in vivo inositol lipid metabolism in frog retinal pigment epithelial (RPE) cells, including the effect of light on phospholipase C-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate. RPE cells were rapidly isolated after either brief light or dark periods. Light and electron microscopy showed complete detachment of the retina from the RPE cells, and that the RPE cell suspensions were devoid of photoreceptor cell outer segments. Frog tissues were labeled in vivo for 20 hr by intravitreal injection of [3H]inositol (4 microCi, 4 microliters per eye) within a 24-hr constant illumination period. Following 1 hr of darkness (priming period), frogs were intravitreally injected with LiCl (0.5 M, 4 microliters per eye) 15 min before the onset of either 30-min light stimulation or an additional 30 min of darkness (controls). In order to preserve endogenous inositol phosphate pools present after dark and light exposure, the RPE cells were harvested in the shortest time possible, at low temperatures (18-20 degrees C), and in the presence of 10 mM LiCl. Total [3H]inositol-labeled water-soluble products (inositol plus inositol phosphates) were increased by 86% after 30 min of light. Inositol trisphosphate (IP3) showed the highest accumulation (a 5.5-fold increase), followed by inositol bisphosphate (1.9-fold increase) and inositol monophosphate (1.4-fold increase). Free [3H]inositol also accumulated (2.8-fold increase), reflecting only a partial inhibition of phosphomonoesterase by LiCl. These changes were paralleled by a 12% decrease in 3H-labeled phosphatidylinositol with no significant difference in the labeling of polyphosphoinositides.(ABSTRACT TRUNCATED AT 250 WORDS)

Retinal pigment epithelial cells play a central role in the conservation of docosahexaenoic acid by photoreceptor cells after shedding and phagocytosis.

The involvement of retinal pigment epithelial (RPE) cells in the recycling of docosahexaenoic acid (DHA), from phagocytized disc membranes back to the retina, was studied in frogs subsequent to injection of [3H]DHA via the dorsal lymph sac. Rod outer segments (ROS) gradually accumulated [3H]DHA as a dense, heavily labeled region that arrived at the distal tips by 28 days post-injection. Autoradiographic analysis at the time of maximal shedding and phagocytosis (1-2 hr after the onset of light) showed diffusely (before 28 days) and heavily (after 28 days) labeled phagosomes in RPE cells. Biochemical analysis of the [3H]DHA-containing lipids of discs that contribute to the labeling of RPE cells after phagocytosis was also performed. Between 27 and 34 days, when 12% of retinal [3H]DHA-lipids present in disc membranes are phagocytized by RPE cells, total retinal labeling remained unchanged. Taken together, these data suggest that the [3H]DHA of the densely labeled region of the ROS was recycled back to the photoreceptor cells only after it had reached the RPE cells following 28 days post-injection. We conclude that, following daily phagocytosis of ROS tips, RPE cells play a central role in the conservation and redelivery of ROS-derived DHA back to photoreceptor cells through the interphotoreceptor matrix.