Klebsiella pneumoniae induces an inflammatory response in an in vitro model of blood-retinal barrier.

Klebsiella pneumoniae has become an important pathogen in recent years. Although most cases of K. pneumoniae endogenous endophthalmitis occur via hematogenous spread, it is not yet clear which microbial and host factors are responsible for the ability of K. pneumoniae to cross the blood-retinal barrier (BRB). In the present study, we show that in an in vitro model of BRB based on coculturing primary bovine retinal endothelial cells (BREC) and primary bovine retinal pericytes (BRPC), K. pneumoniae infection determines changes of transendothelial electrical resistance (TEER) and permeability to sodium fluorescein. In the coculture model, bacteria are able to stimulate the enzyme activities of endothelial cytosolic and Ca(2+)-independent phospholipase A2s (cPLA2 and iPLA2). These results were confirmed by the incremental expression of cPLA2, iPLA2, cyclo-oxygenase-1 (COX1), and COX2 in BREC, as well as by cPLA2 phosphorylation. In supernatants of K. pneumoniae-stimulated cocultures, increases in prostaglandin E2 (PGE2), interleukin-6 (IL-6), IL-8, and vascular endothelial growth factor (VEGF) production were found. Incubation with K. pneumoniae in the presence of arachidonoyl trifluoromethyl ketone (AACOCF3) or bromoenol lactone (BEL) caused decreased PGE2 and VEGF release. Scanning electron microscopy and transmission electron microscopy images of BREC and BRPC showed adhesion of K. pneumoniae to the cells, but no invasion occurred. K. pneumoniae infection also produced reductions in pericyte numbers; transfection of BREC cocultured with BRPC and of human retinal endothelial cells (HREC) cocultured with human retinal pericytes (HRPC) with small interfering RNAs (siRNAs) targeted to cPLA2 and iPLA2 restored the pericyte numbers and the TEER and permeability values. Our results show the proinflammatory effect of K. pneumoniae on BREC, suggest a possible mechanism by which BREC and BRPC react to the K. pneumoniae infection, and may provide physicians and patients with new ways of fighting blinding diseases.

Loss of aromatase cytochrome P450 function as a risk factor for Parkinson's disease?

The final step in the physiological synthesis of 17beta estradiol (E(2)) is aromatization of precursor testosterone by a CYP19 gene product, cytochrome P450 estrogen aromatase in the C19 steroid metabolic pathway. Within the central nervous system (CNS) the presence, distribution, and activity of aromatase have been well characterized. Developmental stage and injury are known modulators of brain enzyme activity, where both neurons and glial cells reportedly have the capability to synthesize this key estrogenic enzyme. The gonadal steroid E(2) is a critical survival, neurotrophic and neuroprotective factor for dopaminergic neurons of the substantia nigra pars compacta (SNpc), the cells that degenerate in Parkinson's disease (PD). In previous studies we underlined a crucial role for the estrogenic status at the time of injury in dictating vulnerability to the parkinsonian neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). Our ongoing studies address the contribution of brain aromatase and extragonadal E(2) as vulnerability factors for PD pathology in female brain, by exposing aromatase knockout (ArKO, -/-) female mice which are unable to synthesize estrogens to MPTP. Our initial results indicate that aromatase deficiency from early embryonic life significantly impairs the functional integrity of SNpc tyrosine hydroxylase-positive neurons and dopamine transporter innervation of the caudate-putamen in adulthood. In addition, ArKO females exhibited a far greater vulnerability to MPTP-induced nigrostriatal damage as compared to their Wt type gonadally intact and gonadectomized counterparts. Characterization of this novel implication of P450 aromatase as determining factor for PD vulnerability may unravel new avenues for the understanding and development of novel therapeutic approaches for Parkinson's disease.

High glucose and advanced glycation end products induce phospholipid hydrolysis and phospholipid enzyme inhibition in bovine retinal pericytes.

In the present study, we investigated the possible role of oxidative stress and the modulation of phospholipid turnover in two related models of pericyte injury, i.e., treatment with high glucose or advanced glycation end products (AGEs). Growing microcapillary pericytes from bovine retinas in culture were incubated, for 3 weeks, with 20-50 mM glucose or 2-20 microM AGEs, and peroxidation parameters (malondialdehyde, conjugated diene, hydroperoxide, glutathione (GSH) levels and lactate dehydrogenase (LDH) release) were evaluated. Arachidonate (AA) and choline release from membrane phospholipids was determined in pericytes prelabeled with [1-(14)C]arachidonate and [Me-(3)H]choline, respectively, and stimulated with elevated glucose or AGEs for 30 min or 2 h. [1-(14)C]arachidonate and [Me-(3)H]choline incorporation into phospholipids, for 2 h and 3 h respectively, was also studied in conditioned and serum-starved cultures. Finally, lysates of treated and control cells were assayed for cytosolic phospholipase A(2) (cPLA(2)), acyl-CoA:1-acyl-sn-glycero-3-phosphocholine O-acyltransferase (AT), CTP:phosphocholine cytidylyltransferase (CT) and microsomal choline phosphotransferase (CPT) enzyme activities. We found that high glucose and AGEs caused neither significant production of reactive oxygen species nor cell toxicity or death, unlike other cell types. Both agents had no significant effect on the cellular ultrastructure, evaluated by light and electron microscopy, AA incorporation and release, cytosolic phospholipase A(2) (cPLA(2)) and AT activities. On the contrary, choline incorporation into phosphatidylcholine, CT and CPT activities were significantly reduced either by 50 mM glucose or 20 microM AGEs. Simultaneously, [Me-(3)H]choline release was significantly stimulated by both agents. We conclude that prolonged treatments with high glucose or AGEs are not able to induce oxidative injury in bovine retinal capillary pericytes. Nevertheless, they do induce phospholipid hydrolysis and phospholipid enzyme activity inhibition.

t-Butyl hydroperoxide and oxidized low density lipoprotein enhance phospholipid hydrolysis in lipopolysaccharide-stimulated retinal pericytes.

Free radicals induced by organic peroxides or oxidized low density lipoprotein (oxLDL) play a critical role in the development of atherosclerosis. In investigating this process, and the concomitant inflammatory response, the role of pericytes, cells supporting the endothelial ones in blood vessels, has received little attention. In this study we tested the hypothesis that tert-butyl hydroperoxide (t-BuOOH) and oxLDL, administered in sublethal doses to the culture medium of retinal pericytes, function as prooxidant signals to increase the stimulation of the peroxidation process induced by lipopolysaccharide (LPS). Confluent cell monolayers were exposed to t-BuOOH (25-400 microM), native LDL or oxLDL (3.4-340 nmol hydroperoxides/mg protein, 1-100 micro). LPS (1 microg/ml), t-BuOOH (200 microM), and oxLDL (100 microM), but not native LDL, incubated for 24 h with cells, markedly increased lipid peroxidation, cytosolic phospholipase A2 (cPLA2) activity and arachidonic acid (AA) release in a time- and dose-dependent manner. AACOCF(3), a potent cPLA2 inhibitor, and the antioxidant alpha-tocopherol strongly inhibited the prooxidant-stimulated AA release. Long-term exposure to maximal concentrations of t-BuOOH (400 microM) or oxLDL (100 microM) had a sharp cytotoxic effect on the cells, described by morphological and biochemical indices. The presence of t-BuOOH or oxLDL at the same time, synergistically increased phospholipid hydrolysis induced by LPS alone. 400 microM t-BuOOH or 100 microM oxLDL had no significant effect on the stimulation of an apoptosis process estimated by DNA laddering and light and electron microscopy. The results indicate that (i) pericytes may be the target of extensive oxidative damage; (ii) activation of cPLA2 mediates AA liberation; (iii) as long-term regulatory signals, organic peroxide and specific constituents of oxLDL increase the pericyte ability to degrade membrane phospholipids mediated by LPS which was used, in the present study, to simulate in vitro an inflammatory burst in the retinal capillaries.

Susceptibility of rat retina acyl-CoA:1-acyl-sn-glycero-3-phosphocholine O-acyltransferase and CTP:phosphocholine cytidylyltransferase activity to lipid peroxidation and hydroperoxide treatment.

Two enzyme activities involved in phospholipid metabolism in the rat retina were determined after in vivo and in vitro peroxidation according to several model systems. The in vivo models were based on: (i) intravenous administration of a sonicated emulsion of phospholipid and linoleate photooxidized mixture to normal rat for a period of one week; (ii) acute injection of Fe2+ solution (20 mM) or (iii) 0.5 mg of hydroperoxylinoleate into the vitreous body, and collection of retinal tissue 4 h or 4 days later, respectively. Oleoyl CoA:lysophosphatidylcholine acyltransferase activity was unchanged or exhibited significant inhibition. On the contrary, CTP:phosphocholine cytidylyltransferase activity was stimulated. By incubating in vitro the retina with: (i) Fe(2+)-ascorbate; (ii) photooxidized phospholipid mixture (0.1-5 mM) or individual phospholipid classes; (iii) hydroperoxylinoleate (0.25-2 mM), with or without Fe2+, a significant inactivation of acyltransferase (six-fold maximum loss of initial activity) and a slight stimulation of cytidylyltransferase were seen. Altogether, the results suggest that in situ oxygen radical generation by a variety of agents irreversibly perturbs enzymes and/or membrane structures in which the enzymes are inserted; these events may bea causal factor in retinal degeneration accompanying some ocular diseases.

1-Acyl-2-lysophosphatidylcholine transport across the blood-retina and blood-brain barrier.

The transport of lysophospholipids through the rat blood-retina and blood-brain barrier was determined by using radioactive 1-palmitoyl-2-lysophosphatidylcholine (Pam-lysoPtdCho) and by measuring the uptake of this labeled compound into the retina and various brain regions after short in situ carotid perfusion. The transport was not affected by probenecid (0.25 mM), but it was inhibited, in a dose-dependent manner, by circulating albumin which is able to bind tightly to lysophosphatidylcholine and lowered the availability of the latter for tissue extraction. Radiotracer transfer in the retina was higher than in brain regions. The permeability-surface area products (PS) changed with the inclusion of unlabeled Pam-lysoPtdCho, showing that transport across retinal and brain microvessels is mainly saturable. The data provided an estimate of transport constants (Vmax, Km and non-saturable constant Kd). However, we could not distinguish whether this saturable process represents the saturation of a transport carrier or simple passive diffusion followed by the saturation of enzymatic reactions. In brain tissue lipid extract, 20 s after carotid injection, radiolabel was associated by 45% to unmetabolized Pam-lysoPtdCho. Partial acylation to phosphatidylcholine, as well as hydrolysis and redistribution of the fatty acyl moiety into main phospholipid classes also occurred. The present results, compared to our previous data, indicate that PamlysoPtdCho is transported faster and/or in greater amounts than unesterified fatty acids.

Phosphatidylcholine synthesis-related enzyme activities of bovine brain microvessels exhibit susceptibility to peroxidation.

In microvessels isolated from bovine brain, microsomal enzyme activities involved in phosphatidylcholine biosynthesis and degradation were determined. The microvessels possessed acyl-CoA:1-acyl-sn-glycero-3-phosphocholine (AT) and glycerophosphocholine phosphodiesterase (GroPChoPDE) activity at a higher level compared with bovine and rat brain or rat liver microsomes whereas they expressed CTP:phosphocholine cytidylyltransferase (CT) and choline phosphotransferase (CPT) activity at a lower level. Each enzyme has been characterized in terms of response to inhibitors or activators revealing properties very similar to those in brain and liver microsomes. In the homogenate prepared from t-butylhydroperoxide-treated microvessels (10 min exposure to 10 microM up to 1 mM concentrations), AT and CPT activities exhibited a significant dose-dependent inhibition. In contrast, GroPChoPDE activity was unaffected. CT was inhibited only at 1 mM concentration. Short treatment of microvessels with Fe2+ (20 microM)-ascorbate (0.25 mM) or 100 microM linoleate hydroperoxide did not have any effect on the activity of the four enzymes. Strong inhibition of all enzymes was noted when the linoleate hydroperoxide system was fortified by Fe2+ ions (100 microM). AT inactivation was also found when oxidized low density lipoprotein was preincubated with microvessels. On the other hand, oxidized LDL left unchanged CPT and GroPChoPDE activities whereas it promoted a slight stimulation of cytidylyltransferase activity. Overall, the results suggest a link between oxygen radical generation and the perturbation of the microvessel membrane structure in which the four enzymes are incorporated, coupled to a direct sulfhydryl protein modification.

Levels of ethanolamine intermediates in the human and rat visual system structures: comparison with neural tissues of a lower vertebrate (Mustelus canis) and an invertebrate (Loligo pealei).

Levels of ethanolamine intermediates in the retina and optic nerves of autopsied human donors and in the rat visual system (retina, optic nerve, lateral geniculate body, superior colliculus) were measured. Amounts were also obtained from the retina, optic nerve, and optic tectum of a primitive elasmobranch, the smooth dogfish Mustelus canis, and from the related nervous structures (retina, optic lobe, fin nerve, and stellate ganglia) of a marine invertebrate, the squid Loligo pealei. In all regions of the human and rat nervous system, the pool size of CDP-ethanolamine (values ranging between 10-31 nmol/g wet wt) was much smaller than that of free ethanolamine (values ranging between 197-395 nmol/g wet wt), whereas glycerophosphorylethanolamine was present in relatively high content (values ranging between 125-280 nmol/g wet wt). In nervous system regions of the dogfish and squid, the distribution of values followed the same general trend as observed for humans and rats, even if all regions had less ethanolamine intermediates compared to the mammalian counterpart. In dogfish and squid retina, glycerophosphorylethanolamine showed the highest pool size among the ethanolamine derivatives analyzed (16 and 44 nmol/g wet wt, respectively). The present study confirms the basic similarity of ethanolamine intermediate pool size patterns in the nervous system structures (with the exception of the retina) of animal species which have widely different phylogenetic positions. The data support the proposal that the levels reached by ethanolamine and its derivatives in the nervous tissue is the result of an ancient evolutionary development of metabolic pathways for the maintenance of phosphatidylethanolamine membraneous content.

Characterization of phospholipase A2 and acyltransferase activities in squid (Loligo pealei) axoplasm: comparison with enzyme activities in other neural tissues, axolemma and axoplasmic subfractions.

Phospholipase A2 and acyltransferase were assayed and characterized in pure axoplasm and neural tissues of squid. Intracellular phospholipase A2 activity was highest in giant fiber lobe and axoplasm, followed by homogenates from retinal fibers, optic lobe and fin nerve. In most preparations, exogenous calcium (5 mM) caused a slight stimulation of activity. EGTA (2 mM) was somewhat inhibitory, indicating that low levels of endogenous calcium may be required for optimum activity. Phospholipase A2 was inhibited by 0.1 mM p-bromophenacylbromide, and was completely inactivated following heating. The level of acylCoA: lysophosphatidylcholine acyltransferase activity was higher in axoplasm and giant fiber lobe than in other neural tissues of the squid. Km (apparent) and Vmax (apparent) for oleoyl-CoA and lysophosphatidylcholine were quite similar for axoplasm and giant fiber lobe enzyme preparations. Acyltransferase activity was inactivated by heat treatment, and greatly inhibited by 0.2 mM p-chloromercuribenzoate, and to a lesser extent by 20 mM N-ethylmaleimide. Phospholipase A2 activity was present in fractions enriched in axolemmal membranes (separated from squid retinal fibers and garfish olfactory nerve) from both tissues, and it was also highly concentrated in vesicles derived from squid axoplasm. In all three preparations, phospholipase A2 activity was stimulated by Ca++ (5 mM) and inhibited by EGTA (2 mM). In addition, axoplasmic cytosol (114,000 g supernatant) retained a substantial portion of a Ca(++)-independent phospholipase A2, active in the presence of 2 mM EGTA. Acyltransferase activity was present at high content in both axolemma membrane rich fractions, and among subaxoplasmic fractions and axoplasmic vesicles.

Levels of choline intermediates in the visual system structures and in peripheral nerve of the rat: Comparison with neural tissues of a lower vertebrate (Mustelus canis) and an invertebrate (Loligo pealei).

The levels of choline intermediate endogenous pools in structures of the visual system (retina, optic nerve, lateral geniculate body, superior colliculus) and in sciatic nerve of adult (4-month-old) and young (30-day-old) rats were measured. The amounts were also obtained from retina, optic nerve, optic tectum and cranio-spinal nerves of a primitive elasmobranch, the smooth dogfish Mustelus canis, and from related nervous structures (retina, optic lobe, fin nerve, stellar nerve and stellate ganglia) of a marine invertebrate, the squid Loligo pealei. In all regions of rat nervous system, the pool size of CDP-choline was much smaller than that of free choline, whereas GroPCho was present in a relatively higher content. The pool sizes of choline intermediates in 30- and 120-day-old rats were nearly the same. In nervous system regions of the dogfish and squid, the values followed the same general trend as observed for rat. Squid nervous tissues had the lowest choline and GroPCho contents. The rat retina showed the lowest glycerophosphorylcholine phosphodiesterase activity. The chemical studies described here confirm the basic similarity in the pattern of choline intermediate pool sizes among animal species widely different in phylogenetic position. The data highly reinforce the idea that the precursor role of choline and catabolic pathways for the maintenance of the PtdCho membraneous pool seem highly conserved during evolution.

Lipid peroxidation inhibits oleoyl-CoA: 1-acyl-sn-glycero-3-phosphocholine O-acyltransferase in rat CNS axolemma-enriched fractions.

The effect of phospholipid peroxidation on the acylation of lysoPtdCho (lysophosphatidylcholine) by axolemma-enriched fraction prepared from rat brain stem was investigated. After two types of peroxidative treatments, the in vitro induction of malondialdehyde and conjugated dienes formation in axolemmal membranes correlated to a shift in the ratio of saturated/unsaturated fatty acids. By using an Fe2+ (20 microM)-ascorbate (0.25 mM) peroxidation system, the residual acyltransferase activity was 55% of the initial one. No change in Km value for either oleoyl-CoA or lysoPtdCho was found, whereas a loss of 24% in Vmax was observed. After 5 min preincubation with 150 mM t-BuOOH, 70% inactivation of the acylation reaction was observed. A near suppression of enzyme activity was reached with 400 mM. The apparent Km for oleoyl-CoA decreased sharply (from 6.6 microM in control preparations to 4.1 microM in t-BuOOH-treated membranes), indicating a 2-fold increase in the enzymatic affinity for this substrate. The apparent Km for lysoPtdCho increased markedly (from 1.56 microM in the control preparations to 5.88 microM in t-BuOOH-treated membranes) whereas a decrease of Vmax (from 1.65 to 0.80 nmol/min/mg protein) for the same substrate was observed. Significant enzyme inactivation (loss of 60% of initial activity) was seen when 10 mumol of photooxidized phospholipids were preincubated with axolemmal membranes. Significant dose-dependent enzyme inactivation was brought about by addition of 10-60 mumol of peroxidized PtdEtn/100 micrograms axolemmal protein. The percent enzyme inhibition by peroxidized PtdCho at equivalent amounts was lower than that by PtdEtn.(ABSTRACT TRUNCATED AT 250 WORDS)

Aging does not affect the susceptibility to lipid peroxidation and lysosomal enzyme release of rat visual system structures and sciatic nerve.

The aim of the present study was to clarify the issue of lipid peroxidation operating in visual system structures and sciatic nerve of the rat as a contributing factor to senescence. In 4-, 14- and 28-month-old male rats, the amount of endogenous malondialdehyde, conjugated dienes and extractable phospholipids were all taken as indices of lipid peroxidation. In addition, the total free and released enzyme activities of four lysosomal hydrolases were evaluated. The susceptibility of all these parameters to in vitro iron-induced peroxidation was also taken as an age-related indicator of the endogenous peroxidative potential of the nervous tissues examined. Our data show that the content of malondialdehyde and phospholipids did not change in an age-related fashion. Furthermore, the susceptibility of rat visual system structures to lipid peroxidation, together with the release of lysosomal enzymes were unchanged as a function of aging. The results do not lend support to the hypothesis that an increase in overall lipid peroxidation is peculiar to the aging phenomenon of the central nervous system areas which delimit the rat visual pathway.

Lipid peroxidation inhibits acyl-CoA:-1-acyl-sn-glycero-3-phosphocholine O-acyltransferase but not CTP: phosphocholine cytidylyltransferase activity in rat brain membranes.

In brain tissue in vivo peroxidized according to three model systems, we determined two microsomal enzyme activities involved in phospholipid biosynthesis. The first, short-term model, was based on the i.v. administration to normal rats, twice a day, for a period of 1 week, of a sonicated emulsion of a peroxidized mixture of phospholipids and linoleate (4:1, w/w; 500 mg/day; hydroperoxides: 200-250 nmol/mg lipid). The half-life time of the injected toxic lipid species in the blood circulation was about 1 h. At the end of the week's treatment, brain and liver malondialdehyde, conjugated diene and lipid hydroperoxide levels were significantly higher in treated rats than in the controls. The second model consisted of the acute injection of aqueous Fe2+ solution (50 mM) into lateral ventricles, and the collection of brain tissue 2 h later. The third model was based on two consecutive injections of hydroperoxylinoleate (1 mg each) into lateral ventricles over a period of 18 h, and the collection of brain tissue 2 h after the second administration. In brain microsomal membranes prepared from peroxide- or iron-treated rats, lysophosphatidylcholine acyltransferase activity exhibited a significant inhibition. On the contrary, in microsomal preparations derived from the short-term model, CTP:phosphocholine cytidylyltransferase activity was slightly stimulated. Intraventricular injection of linoleate or linoleic acid hydroperoxide left this enzyme activity unchanged. The effect of in vitro membrane peroxidation on both microsomal enzyme activities was investigated. By using an Fe2+ (20 microM)-ascorbate (0.25 mM) peroxidation system, the residual acyltransferase and cytidylyltransferase activities were 80 and 72% of initial activity respectively. Significant dose-dependent inactivation of acyltransferase (maximum loss of 45% of initial activity) was seen when 0.1-10 mumol of photooxidized phospholipids were preincubated with 100 micrograms of microsomal membranes. Unoxidized or photooxidized phospholipids (1 mM) promoted a slight stimulation of cytidylyltransferase activity. Altogether, the results suggest a link between oxygen radical generation and the perturbation of the membrane structure in which the enzymes are located.

Arachidonate transport through the blood-retina and blood-brain barrier of the rat after reperfusion of varying duration following complete cerebral ischemia.

The permeability-surface area product (PS) of [1-14C]arachidonate at the blood-retina and blood-brain barrier was determined by short carotid perfusion in young Wistar rats 1 or 6 h after recovery period following complete cerebral ischemia induced by temporary cardiac arrest. For the retina and structures of visual system, hypothalamus and olfactory bulb there was no significant difference over sham-operated rats among mean PSs. For cortex, hippocampus and striatum, significant increases were found at both time intervals of recovery after cardiac arrest. The ischemia-reperfusion model was characterized by a significant increase in tissue conjugated diene in the hippocampus and microsomal lysophosphatidylcholine acyltransferase activity in the cortex. Consistent with these findings, we also show ultrastructural evidence mainly represented by partial opening of interendothelial junctions and mild signs of tissue edema in surrounding neuropil, suggesting barrier leakiness predominantly in the cortex, hippocampus and striatum but almost absent in the retina microvessels. Our results indicate that ischemia-reperfusion does affect influex through blood-brain barrier into regional structures of rat central nervous system of arachidonate, a metabolic substrate and lipid mediator rapidly incorporated into microcapillary and brain lipids. The data also suggested that: (i) reactive oxyradicals were moderately generated during the early phase of ischemic-reperfusion process in the rat; (ii) after reperfusion, in vitro susceptibility of different brain regions to iron-induced peroxidation was highest in the hippocampus and lowest in the cortex and striatum; (iii) membrane phospholipid repair mechanisms were activated at the same time.

Amyloid beta but not bradykinin induces phosphatidylcholine hydrolysis in immortalized rat brain endothelial cells.

We describe the inhibitory effect of A beta (25-35) fragment of amyloid-beta peptide and bradykinin (BK) on phosphatidylcholine (PtdCho) metabolism in immortalized rat brain GP8.39 endothelial cells (EC). Cultures were incubated either with A beta for 24-48 h, or with BK for 30 min-4 h. The peroxidation indices (malondialdehyde, conjugated dienes) and lactate dehydrogenase (LDH) release significantly increased after A beta peptide (10-50 microM) treatment. The BK (10 microM) stimulation of cells brought about an increase in conjugated dienes and LDH release only after 4 h. Following 24 h treatment with 50 microM A beta peptide, the [Me-3H]choline incorporation into PtdCho strongly decreased while the [3H]choline release increased, indicating PtdCho hydrolysis. The effect was most likely due to peptide prooxidant effect. After 4 h preincubation with BK, the [Me-3H]choline incorporation into PtdCho strongly decreased, but no significant [3H]choline release was found. Following long-term treatment, the action of 50 microM A beta on [3H]choline release was not enhanced by 10 microM BK. Cell exposure to alpha-tocopherol (1 mM) prior to the addition of both agents did not abolish stimulated PtdCho breakdown. The data suggest that: (a) A beta peptide and BK may modulate phospholipid turnover in microvessel cells; (b) they could not synergistically interact in vascular EC damage during processes involving amyloid accumulation and inflammatory response.

Palmitate transport through the blood-retina and blood-brain barrier of rat visual system during aging.

The permeability-surface area product (PA) of [1-14C]palmitate at the blood-retina (BRB) and blood-brain barrier (BBB) was determined after short carotid perfusion in male Sprague-Dawley rats at 4, 14 and 28 months of age. For the retina, optic nerve and tract, lateral geniculate body, visual and parietal cortex, there was no significant difference among mean PAs in any age group. For superior colliculus, frontal cortex, striatum, hippocampus and olfactory bulb, a slight but significant increase of PA values was observed between young (4-month-old) and senescent (28-month-old) rats. Our results indicate that aging does not affect influx into retina and other structures of rat visual system of the palmitate, a metabolic substrate for which carrier-mediated transport across the BRB and BBB has not been demonstrated.

Differential transport of docosahexaenoate and palmitate through the blood-retina and blood-brain barrier of the rat.

The permeability-surface area product (PS) of [1-14C]docosahexaenoate and [1-14C]palmitate at the blood-retina (BRB) and blood-brain barrier (BBB) was determined after in situ brain perfusion in Sprague-Dawley rats. The intracarotid injection procedure involved continuous infusion of albumin-bound fatty acids for up to 20 s. In the retina, and visual, parietal and frontal cortex, there was a significant decrease in mean PSs for docosahexaenoate compared to the palmitate group. For optic nerve and tract, superior colliculus, lateral geniculate, striatum, hippocampus and olfactory bulb, the comparison of PS values between the two fatty acid-injected groups of animals did not reveal any difference. It is suggested that the lower docosahexaenoate transport, compared to 16:0, across microvascular endothelium of the retina and other cortical regions might help explain the highest availability and selective retention of the essential 22:6(n-3) fatty acid in these nervous system structures.

Lipid hydroperoxides induce changes in palmitate uptake across the rat blood-retina and blood-brain barrier.

The blood-retina and blood-brain transport of fatty acids was studied in control and lipid hydroperoxide-treated rats by measuring the permeability-surface area product (PS) to [1-14C]palmitate. An in situ carotid perfusion method was used. PS values were evaluated: (1) just after intracarotid injection of hydroperoxides; or (2) after a short-term systemic treatment for 1 week with sonicated emulsion of phospholipids-linoleate peroxidized mixture. Compared with saline-treated rats, PS remarkably decreased in the retina and most brain regions studied after acute, arterial injection of hydroperoxide preparations. On the contrary, the transport index significantly increased in the retina and almost all the brain areas after 7 days i.v. treatment with hydroperoxide emulsion. It is suggested that hydroperoxides acutely administered before transport radiotracer brought about a reinforcement of microvasculature junctional area or hampered substrate diffusion across endothelial membrane. On the other hand, upon short-term i.v. administration, hydroperoxides presumably triggered a lipid structure derangement of endothelial cell membranes and zonulae occludens due to their local accumulation and/or high capability of generating oxygen-free radicals.

Arachidonate transport through the blood-retina and blood-brain barrier of the rat during aging.

The permeability-surface area product (PS) of [I-14C]arachidonate at the blood-retina (BRB) and blood-brain barrier (BBB) was determined after short carotid perfusion in Wistar rats at 4, 12 and 28 months of age. For the visual system structures, parietal and frontal cortex, striatum, hypothalamus, hippocampus and olfactory bulb there was no significant difference among mean PSs in any age group. Our results indicate that: (1) arachidonate is able to cross at relevant rate BRB and BBB; (2) in all brain regions except retina, optic tract and hippocampus, blood barriers have a transport capacity for arachidonate significantly higher than that for docosahexaenoate and palmitate as well; (3) aging does not affect influx into retina and other structures of rat central nervous system of the arachidonate, a metabolic substrate rapidly incorporated into microcapillary and brain lipids, and for which simple diffusion transport across the BRB and BBB may be postulated.

Amyloid beta(1-42) and its beta(25-35) fragment induce in vitro phosphatidylcholine hydrolysis in bovine retina capillary pericytes.

We describe the inhibitory effect of full-length Abeta(1-42) and Abeta(25-35) fragment of amyloid-beta peptide on phosphatidylcholine (PtdCho) metabolism in bovine retina capillary pericytes. Cell cultures were incubated with Abetas for 24 h. Peroxidation indices (malondialdehyde and lactate dehydrogenase release) significantly increased after 20-50 microM Abeta(1-42) or Abeta(25-35) treatment. In addition, [Me-3H]choline incorporation into PtdCho strongly decreased while either 3H-choline or 14C-arachidonic acid release from prelabeled cells increased, indicating PtdCho hydrolysis. The effect was very likely due to prooxidant action of both Abeta peptides. Reversed-sequence Abeta(35-25) peptide did not depress 3H-choline incorporation nor stimulate PtdCho breakdown. With addition of Abetas at low concentrations (2-20 microM) to pericytes, marked ultrastructural changes, well connected to metabolic alterations, emerged including shrinkage of cell bodies, retraction of processes, disruption of the intracellular actin network. Cells treated with higher concentrations (50-200 microM) displayed characteristics of necrotic cell death. The data suggest that: (a) Abeta(1-42) and Abeta(25-35) peptides may modulate phospholipid turnover in microvessel pericytes; (b) together with endothelial cells, pericytes could be the target of vascular damage during processes involving amyloid accumulation.