Stephen J. Kish

To examine the possible causal contribution of normal or accelerated aging to the neurodegenerative process of Parkinson's disease, we measured the influence of aging on subregional striatal dopamine and homovanillic acid levels in postmortem brain of 23 neurologically and psychiatrically normal human subjects 14-92 years old. We observed a significant decline in striatal dopamine levels and increase in the homovanillic acid/dopamine molar ratios with increasing age. The dopamine loss, on average, was of the same magnitude in the caudate nucleus and the putamen (-60% in the 84-year-old group as compared with the 22-year-old group), with the caudal component of both nuclei being more affected than the rostral subdivisions. The level of subregional dopamine metabolism, as measured by the homovanillic acid/dopamine ratio, in our young individuals (mean age, 22 years) was found to be inversely correlated to the degree of subregional dopamine loss suffered by the individuals in the older age groups. We conclude the following: (a) Striatal subdivisions with physiologically higher dopamine metabolism are not at a greater risk of suffering dopamine neuronal damage with advancing age, as would seem to be implied by the oxidative stress hypothesis; thus, formation of dopamine-derived oxy radicals in the human striatum appears unlikely to be a primary factor responsible for the age-related striatal dopamine loss. (b) The regional and subregional pattern of striatal dopamine loss in normal aging differs substantially from the pattern typically observed in idiopathic Parkinson's disease; therefore, the cause of idiopathic Parkinson's disease cannot be primarily an age-dependent neurodegenerative process.

Controversy exists as to the clinical importance, cause, and disease specificity of the cytochrome oxidase (CO) activity reduction observed in some patients with Alzheimer's disease (AD). Although it is assumed that the enzyme is present in normal amount in AD, no direct measurements of specific CO protein subunits have been conducted. We measured protein levels of CO subunits encoded by mitochondrial (COX I, COX II) and nuclear (COX IV, COX VIc) DNA in autopsied brain of patients with AD whom we previously reported had decreased cerebral cortical CO activity. To assess disease specificity, groups of patients with spinocerebellar ataxia type I and Friedreich's ataxia were also included. As compared with the controls, mean protein concentrations of all four CO subunits were significantly decreased (-19 to -47%) in temporal and parietal cortices in the AD group but were not significantly reduced (-12 to -17%) in occipital cortex. The magnitude of the reduction in protein levels of the CO subunits encoded by mitochondrial DNA (-42 to -47%) generally exceeded that encoded by nuclear DNA (-19 to -43%). In the spinocerebellar ataxia disorders, COX I and COX II levels were significantly decreased in cerebellar cortex (-22 to -32%) but were normal or close to normal in cerebral cortex, an area relatively unaffected by neurodegeneration. We conclude that protein levels of mitochondrial- and nuclear-encoded CO subunits are moderately reduced in degenerating but not in relatively spared brain areas in AD and that the decrease is not specific to this disorder. The simplest explanation for our findings is that CO is decreased in human brain disorders as a secondary event in brain areas having reduced neuronal activity or neuronal/synaptic elements consequent to the primary neurodegenerative process.

Lysophospholipids are generated during the turnover and breakdown of membrane phospholipids. We have identified and partially characterized three enzymes involved in the metabolism of lysophospholipids in human brain, namely, lysophospholipase, lysophospholipid:acyl-CoA acyltransferase (acyltransferase), and lysophospholipid:lysophospholipid transacylase (transacylase). Each enzyme displayed comparable levels of activity in biopsied and autopsied human brain, although in all cases the activity was somewhat lower in human than that in rat brain. All three enzymes were localized predominantly in the particulate fraction, with lysophospholipase possessing the greatest activity followed by acyltransferase and transacylase. Lysophosphatidylcholine possessed a Km in the micromolar range for lysophospholipase and transacylase, and in the millimolar range for acyltransferase, whereas arachidonyl-CoA displayed a Km in the micromolar range for acyltransferase. The three enzymes differed in their pH optima, with lysophospholipase being most active at pH 8.0, transacylase at pH 7.5, and acyltransferase at pH 6.0. Both bromophenacyl bromide and N-ethylmaleimide inhibited lysophospholipase activity and, to a lesser extent, that of acyltransferase and transacylase. None of the enzyme activities were affected by the presence of dithiothreitol or EDTA, although particulate lysophospholipase was activated approximately two-fold by the addition of 5 mM MgCl2 or CaCl2 but not KCl. Transacylating activity was stimulated by CoA, the EC50 of activation being 6.8 microM. Acyltransferase displayed an approximately threefold preference for arachidonyl-CoA over palmitoyl-CoA, whereas the acylation rate of different lysophospholipids was in the order lysophosphatidylinositol > 1-palmitoyl lysophosphatidylcholine > 1-oleoyl lysophosphatidylcholine >> lysophosphatidylserine > lysophosphatidylethanolamine. This, and the preference of human brain phospholipase A2 for phosphatidylinositol, suggests that this phospholipid may possess a higher turnover rate than the other phospholipid classes examined. Human brain homogenates also possessed the ability to transfer fatty acid from lysophosphatidylcholine to lysophosphatidylethanolamine. In addition, we also present evidence that diacylglycerophospholipids can act as acyl donors for the transacylation of lysophospholipids. We have therefore demonstrated the presence of, and partially characterized, three enzymes that are involved in the metabolism of lysophospholipids in human brain. Our results suggest that lysophospholipase may be the major route by which lysophospholipids are removed from the cell membrane in human brain. However, all three enzymes likely play an important role in the remodeling of membrane composition and thereby contribute to the overall functioning of membrane-associated processes.

Previous studies of patients with spinocerebellar atrophy type 1 (SCA-1) and Friedreich's ataxia (FA) have suggested the occurrence of membrane disturbances in both disorders. We measured concentrations of phosphatidylcholine (PC), diacyl and plasmalogen phosphatidylethanolamine (PE), and phosphatidylserine (PS), along with their fatty acid profiles, in the brains of eight patients with Friedreich's ataxia (FA) and nine patients with dominantly inherited spinocerebellar atrophy type 1 (SCA-1). Compared with the controls, levels of all phospholipid types (PE, PS, and PC) were reduced in the cerebellar but not occipital cortex of SCA-1 patients. In contrast, in the FA group, levels of PS and PE, but not PC, were reduced in both cerebellar and occipital cortices. The fatty acid composition of individual brain phospholipids was altered in both FA and SCA-1 patients, most markedly in the plasmalogen PE and PS classes of cerebellar phospholipids. Given the neuropathologic characteristics of each disorder, it is likely that altered fatty acid composition and phospholipid levels in SCA-1 cerebellar cortex occur as a consequence of pronounced cerebellar degeneration. In contrast, reduced phospholipid levels in FA cerebellar and occipital cortex, areas characterized by, at most, minimal neuronal loss in FA, may represent a widespread alteration in cellular phospholipid metabolism occurring in response to the specific gene defect in the disorder.