- Email: [email protected]
Lipid peroxidation in aging brain and Alzheimer’s disease1, 2
Free Radical Biology & Medicine, Vol. 33, No. 5, pp. 620 – 626, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/02/$–see front matter
PII S0891-5849(02)00807-9
Serial Review: Causes and Consequences of Oxidative Stress in Alzheimer’s Disease Guest Editors: Mark A. Smith and George Perry LIPID PEROXIDATION IN AGING BRAIN AND ALZHEIMER’S DISEASE THOMAS J. MONTINE,* M. DIANA NEELY,* JOSEPH F. QUINN,† M. FLINT BEAL,‡ WILLIAM R. MARKESBERY,§ L. JACKSON ROBERTS II,* and JASON D. MORROW* *Departments of Pharmacology, Pathology, and Medicine and the Vanderbilt Brain Institute, Vanderbilt University Medical Center, Nashville, TN, USA; †Department of Neurology, Veterans Administration Medical Center, Portland, OR, USA; ‡ Department of Neurology and Neuroscience, New York Presbyterian Hospital/Cornell University Medical College, New York, NY, USA; and §Departments of Pathology and Neurology and the Sanders-Brown Center on Aging, University of Kentucky, Lexington, KY, USA (Received 6 February 2002; Accepted 21 February 2002)
Abstract—Lipid peroxidation is one of the major outcomes of free radical-mediated injury that directly damages membranes and generates a number of secondary products, both from fission and endocyclization of oxygenated fatty acids that possess neurotoxic activity. Numerous studies have demonstrated increased lipid peroxidation in brain of patients with Alzheimer’s disease (AD) compared with age-matched controls. These data include quantification of fission and endocyclized products such as 4-hydroxy-2-nonenal, acrolein, isoprostanes, and neuroprostanes. Immunohistochemical and biochemical studies have localized the majority of lipid peroxidation products to neurons. A few studies have consistently demonstrated increased cerebrospinal fluid (CSF) levels of isoprostanes in AD patients early in the course of their dementia, and one study has suggested that CSF isoprostanes may improve the laboratory diagnostic accuracy for AD. Similar analyses of control individuals over a wide range of ages indicate that brain lipid peroxidation is not a significant feature of usual aging. Quantification of isoprostanes in plasma and urine of AD patients has yielded inconsistent results. These results indicate that brain lipid peroxidation is a potential therapeutic target in probable AD patients, and that CSF isoprostanes may aid in the assessment of antioxidant experimental therapeutics and the laboratory diagnosis of AD. © 2002 Elsevier Science Inc. Keywords—Free radicals, Brain, Lipid peroxidation, Aging, Alzheimer’s disease, Isoprostanes, Neuroprostanes
INTRODUCTION
membranes and generation of secondary products. Membrane damage derives from the generation of fragmented fatty acyl chains, lipid-lipid cross-links, endocyclization to produce novel fatty acid esters, and lipid-protein cross-links [1]. In total, these processes combine to produce changes in the biophysical properties of membranes that can have profound effects on the activity of membrane-bound proteins. Secondary products from lipid peroxidation may be divided conceptually into those generated by fragmentation or rearrangement of oxygenated lipid; however, it should be kept in mind that lipid peroxidation is an indiscriminate process and formation of both types of products occurs simultaneously.
Biochemistry of lipid peroxidation Lipid peroxidation is one of the major outcomes of free radical-mediated injury to tissue. Peroxidation of fatty acyl groups, mostly in membrane phospholipids, has three well-described phases: initiation, propagation, and termination. An important aspect of these reactions is that lipid peroxidation will proceed until substrate is consumed or termination occurs. There are two broad outcomes to lipid peroxidation, viz., structural damage to This article is part of a series of reviews on “Causes and Consequences of Oxidative Stress in Alzheimer’s Disease.” The full list of papers may be found on the homepage of the journal. Address correspondence to: Dr. Thomas J. Montine, Alvord Professor of Neuropathology, Department of Pathology, University of Washington, Harborview Medical Center, Box 359791, 325 9th Ave., Seattle, WA 98104, USA; E-Mail: [email protected].
Fragmentation products of lipid peroxidation Fragmentation of lipid hydroperoxides liberates a number of diffusible products, some of which are potent 620
Brain lipid peroxidation in aging and AD
621
Scheme 1. Shown are the structures of some reactive aldehyde products of lipid peroxidation (HNE, 4-hydroxy-2-nonenal; HHE, 4-hydroxy-2hexenal).
electrophiles [2]. The most abundant diffusible products of lipid peroxidation are chemically reactive aldehydes such as malondialdehyde (MDA), acrolein, 4-hydroxy2-hexenal (HHE), and 4-hydroxy-2-nonenal (HNE) (Scheme 1). Several mechanisms have been proposed for the formation of these reactive aldehydes during lipid peroxidation [3,4]. Reactive aldehydes from lipid peroxidation are biologically active. Most studies have focused on their reactivity with a number of cellular nucleophiles, including proteins, nucleic acids, and some lipids [2]. Indeed, many of the cytotoxic effects of lipid peroxidation can be reproduced directly by electrophilic lipid peroxidation products such as HNE [1]. These include depletion of glutathione, dysfunction of structural proteins, reduction in enzyme activities, and induction of cell death. More recent work, much in ex vivo models of the central nervous system, has highlighted alterations in ion channel, transporter, and second messenger system activities following exposure to reactive aldehydes from lipid peroxidation [5–7]. How these many effects are interrelated is not known. We found that microtubule organization and neurite outgrowth were highly vulnerable to HNE exposure in a neuroblastoma cell line [8] (Fig. 1). The mechanisms by which HNE leads to microtubule disruption are not fully clarified, but may involve modification of tubulin [9]. Other lipid peroxidation products, HHE and acrolein, also cause microtubule disruption; however, this may not be the most vulnerable target for these reactive aldehydes [10 –12]. Numerous antioxidant mechanisms have evolved, including several metabolic routes to detoxify fragmentation products from lipid peroxidation. The diffusible reactive aldehydes generated from lipid peroxidation are excellent substrates for glutathione transferases (GSTs) and a number of oxidoreductases that act to detoxify these molecules [2]. Each of the major classes of cytosolic GSTs, ␣, , and , are present throughout rat brain. Two reports have described localization of cytosolic GST immunoreactivity in human brain; one reported GST- and GST- immunoreactivity in glia but not neurons, while the other described GST-␣ and GST-
Fig. 1. Shown are mouse neuroblastoma Neuro2A cells exposed to vehicle or HNE (25 M for 60 min) and microtubules visualized by immunofluorescence microscopy using an anti--tubulin antibody. (A) Control cells show a dense microtubule network that extends to the periphery of the cell. (B) Cells exposed to HNE show severe loss of microtubules. Arrows point to the few remaining microtubules.
localization to glia and vascular cells but not neurons. We have determined the regional, cellular, and class distribution in human brain of the four major oxidoreductases comprising the other major pathways for detoxifying HNE: aldehyde dehydrogenase (ALDH), aldose reductase, aldehyde reductase, and alcohol dehydrogenase (ADH) [13]. Of these four enzymes, only ALDH and aldose reductase are expressed in cerebral cortex, hippocampus, basal ganglia, and midbrain; all four enzymes are present in cerebellum. In cerebrum and hippocampus, aldose reductase was localized to pyramidal neurons and mitochondrial class 2 ALDH was localized to glia, suggesting that neurons and glia may employ different mechanisms to detoxify reactive aldehydes from lipid peroxidation. In addition to liberating diffusible aldehydes, fragmentation of lipid hydroperoxides leaves an abnormally shortened fatty acyl group esterified to lipid. In at least one instance, these chemically modified lipids gain re-
622
T. J. MONTINE et al.
Scheme 2. Shown are the pathways of production and ring structures for F-ring and D/E-ring IsoPs and NeuroPs.
ceptor-binding activity; peroxidation and fragmentation of polyunsaturated fatty acyl groups in phosphatidylcholines can generate platelet activating factor (PAF) analogues that stimulate the PAF receptor [14]. Endocyclized products of lipid peroxidation Rather than pursue fragmentation chemistry, a minority of lipid hydroperoxy radicals undergo internal cyclization to generate endoperoxide intermediates that then are converted to multiple isomers with ring structures analogous to prostaglandins (Scheme 2). When derived from arachidonic acid these compounds are called isoprostanes (IsoPs) and when derived from docosahexaenoic acid these compounds are termed neuroprostanes (NeuroPs) [15,16]. In contrast to arachidonic acid (AA) that is evenly distributed in brain tissue, docosahexaenoic acid (DHA) is highly enriched in neurons in brain [17]. For these reasons it was proposed that NeuroPs may be more sensitive and specific for oxidative damage to neurons in brain. We have characterized both the reduction of the endoperoxide intermediate to F-ring compounds or its isomerization to D- and E-compounds, and have proposed that the ratio of F-ring to D/E-ring compounds reflects the reducing environment in which IsoP or NeuroP formation occurred [18,19]. Unlike prostaglandins, IsoPs and NeuroPs are formed in situ from fatty acids esterified to lipid. Thus, IsoPs and
NeuroPs are thought to contribute to changes in membrane characteristics that accompany lipid peroxidation. Once formed, IsoPs and NeuroPs can undergo hydrolysis to liberate free IsoPs and NeuroPs that are detectable in body fluids [16,20 –22]. Although some products of the IsoP and NeuroP pathways are chemically reactive and can modify cellular proteins, the best studied products of these pathways, D-, E-, and F-ring compounds, do not react with cellular macromolecules, and are not extensively metabolized in situ. Due to their similarity to enzymatically derived prostanoids, several investigators have sought receptor-binding activity for free IsoPs. Indeed, one F2-IsoP isomer has been shown to possess renal and cerebral vasoconstrictor activity, likely through receptor-mediated mechanisms [22–24].
BRAIN LIPID PEROXIDATION
The following sections will review the evidence for brain lipid peroxidation in patients with Alzheimer’s disease and in age-matched control individuals without neurologic disease. Important variables to keep in mind are the methods used to quantify lipid peroxidation, the source of the sample to be analyzed, and whether tissue or fluid was obtained postmortem from definite AD patients or during life from probable AD patients.
Brain lipid peroxidation in aging and AD
Postmortem studies Brain. There is compelling evidence that the magnitude of lipid peroxidation in the brains of AD patients examined postmortem exceeds that in age-matched control individuals. Seminal experiments demonstrated significantly increased thiobarbiturate reactive substances (TBARS) in diseased regions of AD brain obtained postmortem compared to age-matched control individuals [25]. This same group of investigators has shown further that free HNE and acrolein are elevated in diseased regions of AD brain compared to controls [12,26]. Several groups have studied the immunolocalization of protein adducts with reactive aldehydes from lipid peroxidation products. Consistent with the quantitative studies described above, hippocampus and cerebral cortex from AD patients display protein modifications that are not detectable or barely detectable in the corresponding brain regions from age-matched control individuals [27–32]. Also, proteins modified by lipid peroxidation products are present in diseased regions of brain but not in regions uninvolved by AD. Within diseased regions of brain, most studies have observed immunoreactivity with neuronal cytoplasm and neurofibrillary tangles. One group observed that the tissue distribution of HNE-protein adducts varies with apolipoprotein E genotype (APOE) [30,32], but this was not observed by others in a small number of patients [28]. Another study has observed modified proteins in or adjacent to neuritic plaques [33], a distribution that is similar to aged genetically modified mice expressing a mutant human amyloid precursor protein gene [34]. Activity of some of the major metabolizing enzymes for HNE and related aldehydes is altered in brains of AD patients. Interestingly, decreased cytosolic GST activity has been reported in diseased regions of AD brain [35]. In contrast, class 2 ALDH activity is significantly increased in temporal cortex from patients with AD compared to age-matched controls [13]. Combined with the results described above, these findings suggest that increased glial mitochondrial ALDH activity may be a protective response to increased generation of reactive aldehydes from lipid peroxidation in AD brain. Altered metabolic activity of the detoxifying enzymes for reactive aldehydes in AD brain raises serious issues for interpreting the significance of increased levels of these aldehydes in AD brain. Assays for TBARS or specific secondary products, such as HNE, are accurate measures of lipid peroxidation in vitro when metabolism of the lipid peroxidation products does not occur. However, in more complicated systems, extensive metabolism of electrophilic lipid peroxidation products may compromise the accuracy of these assays [36,37]. Due to these limitations in quantifying reactive products of lipid
623
peroxidation, we measured F- and D/E-ring IsoPs and NeuroPs in temporal and parietal cortex, hippocampus, and cerebellum of 9 definite AD patients and 11 agematched controls, all with short postmortem intervals [38]. Total NeuroP (F- plus D/E-ring), but not total IsoP, levels are greater in AD patients than controls, but only in those brain regions involved by AD. The F- to D/Ering ratio for NeuroPs, but not IsoPs, was 40 to 70% lower in AD patients compared to controls; however, this occurred in all brain regions and not only those involved by AD. These data indicate greater free radical damage in DHA- than AA-containing compartments in diseased regions of AD brain, and diminished reducing capacity in DHA-containing compartments throughout AD brain. Using a very similar gas chromatography/mass spectrometric technique, others have demonstrated elevated F2IsoP levels in frontal and temporal lobes of AD patients compared to controls [39], a finding that was confirmed in the study cited above. Another group has reported that F4-NeuroPs (called F4-isoprostanes in their publication) are elevated in temporal and occipital, but not parietal, lobes of AD patients compared to controls, and that F4-NeuroPs are greater than F2-IsoPs in these regions [40]. However, interpretation of data from this study is limited by long postmortem intervals (average 47 h in AD patients) [40]. Cerebrospinal fluid. Cerebrospinal fluid (CSF) also has been investigated as a source of central nervous system tissue for the assessment of lipid peroxidation in AD brain. One study has measured free HNE in CSF obtained from the lateral ventricles postmortem and shown that its concentration is significantly elevated in AD patients compared to age-matched controls [41]. A few studies have determined the concentration of F2-IsoPs, and in one case F4-NeuroPs, in CSF obtained from the lateral ventricles postmortem and have shown significant elevations in AD patients compared to age-matched controls [20,42]. Importantly, CSF F2-IsoP concentrations in AD patients are significantly correlated with global indices of neurodegeneration, decreasing brain weight, degree of cerebral cortical atrophy, and increasing Braak stage, but not with APOE or the tissue density of neuritic plaques or neurofibrillary tangles [42]. It is noteworthy that all of the postmortem CSF studies in AD patients used material from individuals with very short (average 2 to 3 h) postmortem intervals. Intra vitam studies CSF. AD patients undergoing postmortem examination typically have advanced disease and an average duration of dementia of 8 to 12 years. Therefore, a serious limitation to interpretation of results from tissue obtained
624
T. J. MONTINE et al.
Fig. 2. Shown is the frequency distribution for F2-IsoP concentration in CSF obtained from the lumbar cistern of 36 controls and 41 probable AD patients. Similar to other CSF biomarkers of AD, there is considerable overlap between the two groups. However, values for probable AD patients are skewed significantly towards higher levels (MannWhitney test, p ⬍ .001).
postmortem is that the increased brain lipid peroxidation in AD patients might be a late stage consequence of disease. Obviously, a late stage consequence of AD would be a less attractive therapeutic target than a process contributing to disease progression at an earlier stage. The first study of probable AD patients early in the course of dementia showed that F2-IsoPs are significantly elevated in CSF obtained from the lumbar cistern compared to age-matched hospitalized patients without neurological disease [21]. The average duration of dementia in these probable AD patients at the time of CSF examination was less than 2 years, while the average duration of AD is between 8 and 12 years. This result was observed again in another group of probable AD patients and controls [43]. The combined results from these two studies are presented in Fig. 2. Importantly, a different laboratory examining a third group of probable AD patients and controls using a similar mass spectrometric method obtained similar results and confirmed the difference observed between AD patients and controls [44]. This last study suggested a correlation between higher CSF F2-IsoP levels and homozygosity for the 4 allele of APOE, an association not observed in other CSF and postmortem studies of AD patients [20,21,38,42]. In addition to providing mechanistic information about AD pathogenesis and a means to quantitatively assess response to antioxidant therapeutics, CSF F2-IsoP levels also may provide information that is useful in diagnosing AD. We tested the hypothesis that quantification of CSF F2-IsoPs, along with CSF A42 and tau levels, improves laboratory diagnostic accuracy for AD in patients with probable AD, dementias other than AD, and age-matched controls [43]. Individuals were classi-
Fig. 3. Shown is a scatter plot of F2-IsoP concentration in CSF obtained from the lumbar cistern of 42 control individuals of varying ages. Open symbols are for samples collected at the Massachusetts General Hospital and closed symbols are for samples collected at Oregon Health Sciences University. Neither group alone had a significant correlation between age and CSF F2-IsoP concentration. Correlation analysis for the combined groups yielded R2 ⫽ 0.00 and p ⬎ .7.
fied as AD or non-AD by a commercially available test using CSF A42 and tau levels (95% sensitivity, 50% specificity), by CSF F2-IsoP and A42 (90% sensitivity, 83% specificity), and by combined analysis using CSF F2-IsoP, A42, and tau levels (84% sensitivity, 89% specificity). These results indicate that CSF F2-IsoP quantification can enhance the accuracy of the laboratory diagnosis of AD; however, this conclusion is based on a single study and these findings need to be replicated. Each of these CSF-based studies has contained a group of age-matched control individuals. In none of these studies did we observe an association between CSF F2-IsoP concentration and age in controls; however, the range of ages examined was limited because individuals were selected to be age-matched for AD patients. Recently, we have augmented these data by determining CSF F2-IsoP levels in control individuals over a broad range of ages (Fig. 3). Our results show no relationship between age over the third to ninth decades of life and CSF F2-IsoP concentrations. These data do not support the model of progressively increasing oxidative damage to the human central nervous system with advancing age. However, these findings are consistent with a model suggested from studies in other species, viz., oxidative damage is exponentially related to advancing age and is not significantly elevated until the extremes of age [45]. Plasma and urine. Although obtaining CSF from the lumbar cistern is not associated with significant risks, even in the elderly, spinal taps are stressful to the patient and not easily obtained in most clinics. For these reasons, several investigators have pursued quantification of F2-
Brain lipid peroxidation in aging and AD
IsoPs in plasma or urine. Like most data for peripheral biomarkers of neurodegenerative disease, the results have been conflicting. In our study, plasma and urine were collected and analyzed over a 2 year period [46]. Over the ensuing 3 years, 23 of the 25 probable AD patients died and 17 underwent postmortem examination; 12 were diagnosed with definite AD, 4 with Dementia with Lewy Bodies, and 1 with Pick’s disease. There was no significant difference in plasma or urine F2-IsoP levels between definite AD patients and age-matched controls or definite AD patients and probable AD patients. Importantly, all individuals in this study were nonsmokers, a behavior that significantly influences plasma F2-IsoP levels [47]. At the time of sample collection, only 4 individuals were taking multivitamin supplements; their plasma and urine F2-IsoP levels did not differ from those not taking supplements. Another study, using an antibody-based detection method, also concluded that plasma F2-IsoP levels are not elevated in AD patients compared to controls [48]. In contrast, one group using an antibody-based method and another using a gas chromatography/mass spectrometric method both concluded that F2-IsoP levels were increased in AD patients’ plasma or urine compared to controls [44,49]. There are several possible reasons for this discrepancy; however, it is unlikely to be methodological because all of these methods have yielded similar results for CSF levels from AD patients and controls. More likely, the differences reside in criteria used to include or exclude patients and controls. For example, one study did not exclude smokers from the patient group [44]. In summary, results from these studies do not reproducibly support the hypothesis that patients with AD experience increased levels of systemic oxidative stress or that the increased levels of F2-IsoPs observed in CSF are detectable peripherally.
SUMMARY
Results from these studies clearly show that diseased regions of brain from patients with advanced AD have increased levels of lipid peroxidation products compared to controls. Importantly, patients with probable AD early in the course of their dementia have increased CSF F2-IsoP levels compared to age-matched controls. In combination with a large number of additional studies showing in vitro neurotoxic activity for reactive products from lipid peroxidation, these results indicate that lipid peroxidation likely significantly contributes to disease progression and is a potential therapeutic target in AD patients.
625
Acknowledgements — This work was supported by grants from the NIH (AG00774, AG16835, AG05114, GM15431, DK26657, and CA77839) as well as grants from the Alzheimer’s Association, the Abercrombie Foundation, and a Bourroughs Welcome Fund Clinical Scientist Award in Translational Research to J.D.M.
REFERENCES [1] Farber, J. L. Mechanisms of cell injury. In: Craighead, J. E., ed. Pathology of environmental and occupational disease. St. Louis, MO: Mosby-Year Book Inc.; 1995:287–302. [2] Esterbauer, H.; Schaur, R. J.; Zollner, H. Chemistry and biochemistry of 4-hydroxynonenal, malondialdehyde and related aldehydes. Free Radic. Biol. Med. 11:81–128; 1991. [3] Schneider, C.; Tallman, K. A.; Porter, N. A.; Brash, A. R. Two distinct pathways of formation of 4-hydroxynonenal: mechanisms of non-enzymatic transformation of the 9- and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J. Biol. Chem. 276:20831– 20838; 2001. [4] Porter, N. A.; Caldwell, S. E.; Mills, K. A. Mechanisms of free radical oxidation of unsaturated lipids. Lipids 30:277–290; 1995. [5] Mattson, M. P. Modification of ion homeostasis by lipid peroxidation: roles in neuronal degeneration and adaptive plasticity. Trends Neurosci. 20:53–57; 1998. [6] Keller, J. N.; Mattson, M. P. Roles of lipid peroxidation in modulation of cellular signaling pathways, cell dysfunction, and death in the nervous system. Rev. Neurosci. 9:105–116; 1998. [7] Smith, M.; Sayre, L.; Monnier, V.; Perry, G. Radical AGEing in Alzheimer’s disease. Trends Neurosci. 18:172–176; 1995. [8] Neely, M. D.; Sidell, K. R.; Graham, D. G.; Montine, T. J. The lipid peroxidation product 4-hydroxynonenal inhibits neurite outgrowth and disrupts neuronal microtubules. J. Neurochem. 72: 2323–2333; 1999. [9] Neely, M. D.; Zimmerman, L.; Picklo, M. J.; Ou, J. J.; Morales, C. R.; Montine, K. S.; Amarnath, V.; Montine, T. J. Congeners of N␣-acetyl-L-cysteine but not aminoguanidine act as neuroprotectants from the lipid peroxidation product 4-hydroxy-2-nonenal. Free Radic. Biol. Med. 29:1028 –1036; 2000. [10] Picklo, M. J.; Montine, T. J. Acrolein inhibits respiration in isolated brain mitochondria. Biochim. Biophys. Acta 1535:145– 152; 2001. [11] Lovell, M. A.; Xie, C.; Markesbery, W. R. Acrolein, a product of lipid peroxidation, inhibits glucose and glutatmate uptake in primary neuronal cultures. Free Radic. Biol. Med. 28:714 –720; 2000. [12] Lovell, M. A.; Xie, C.; Markesbery, W. R. Acrolein is increased in Alzheimer’s disease brain and is toxic to primary hippocampal cultures. Neurobiol. Aging 22:187–194; 2001. [13] Picklo, M. J.; Olson, S. J.; Markesbery, W. R.; Montine, T. J. Expression and activities of aldo-keto oxidoreductases in Alzheimer’s disease. J. Neuropathol. Exp. Neurol. 60:686 – 695; 2001. [14] McIntyre, T. M.; Zimmerman, G. A.; Prescott, S. M. Biologically active oxidized phospholipids. J. Biol. Chem. 274:25189 –25192; 1999. [15] Morrow, J.; Hill, K.; Burk, R.; Nammour, T.; Badr, K.; Roberts, L. A series of prostaglandin-like compounds produced in vivo in humans by a non-cyclooxygenase, free radical catalyzed mechanism. Proc. Natl. Acad. Sci. USA 87:9383–9387; 1990. [16] Roberts, L. J.; Montine, T. J.; Markesbery, W. R.; Tapper, A. R.; Hardy, P.; Chemtob, S.; Detbarn, W. D.; Morrow, J. D. Formation of isoprostane-like compounds (neuroprostanes) in vivo from docosahexaenoic acid. J. Biol. Chem. 273:13605–13612; 1998. [17] Salem, N.; Kim, H. Y.; Yergey, J. A. Docosahexaenoic acid: membrane function and metabolism. In: Simopoulos, A. P.; Kifer, R. R.; Martin, R. E., eds. Health effects of polyunsaturated acids in seafoods. New York: Academic Press; 1986:263–317. [18] Reich, E. E.; Zackert, W. E.; Brame, C. J.; Chen, Y.; Roberts, L. J. II; Hachey, D. L.; Montine, T. J.; Morrow, J. D. Formation of novel D-ring and E-ring isoprostane-like compounds (D4/E4-neuroprostanes) in vivo from docosahexaenoic acid. Biochemistry 39:2376 –2383; 2000.
626
T. J. MONTINE et al.
[19] Morrow, J. D.; Roberts, L. J.; Daniel, V. C.; Awad, J. A.; Mirochnitchenko, O.; Swift, L. L.; Burk, R. F. Comparison of formation of D2/E2-isoprostanes and F2-isoprostanes in vitro and in vivo— effects of oxygen tension and glutathione. Arch. Biochem. Biophys. 353:160 –171; 1998. [20] Montine, T. J.; Markesbery, W. R.; Morrow, J. D.; Roberts, L. J. Cerebrospinal fluid F2-isoprostanes are increased in Alzheimer’s disease. Ann. Neurol. 44:410 – 413; 1998. [21] Montine, T. J.; Beal, M. F.; Cudkowicz, M. E.; Brown, R. H.; O’Donnell, H.; Margolin, R. A.; McFarland, L.; Bachrach, A. F.; Zackert, W. E.; Roberts, L. J.; Morrow, J. D. Increased cerebrospinal fluid F2-isoprostane concentration in probable Alzheimer’s disease. Neurology 52:562–565; 1999. [22] Morrow, J. D.; Roberts, L. J. The isoprostanes: unique bioactive products of lipid peroxidation. Prog. Lipid. Res. 36:1–21; 1997. [23] Hou, X.; Gobeil, F.; Peri, K.; Speranza, G.; Marrache, A. M.; Lachapelle, P.; Roberts, L. J. II; Varma, D. R.; Chemtob, S. Augmented vasoconstriction and thromboxane formation by 15F2t-isoprostane (8-isoprostaglandin F2␣) in immature pig periventricular brain microvessels. Stroke 31:516 –525; 2000. [24] Hoffman, S. W.; Moore, S.; Ellis, E. F. Isoprostanes: free radicalgenerated prostaglandins with constrictor effects on cerebral arterioles. Stroke 28:844 – 849; 1997. [25] Lovell, M.; Ehmann, W.; Butler, S.; Markesbery, W. Elevated thiobarbituric acid-reactive substances and antioxidant enzyme activity in the brain in Alzheimer’s disease. Neurology 45:1594 – 1601; 1995. [26] Markesbery, W. R.; Lovell, M. A. Four-hydroxynonenal, a product of lipid peroxidation, is increased in the brain in Alzheimer’s disease. Neurobiol. Aging 19:33–36; 1998. [27] Calingasan, N. Y.; Uchida, K.; Gibson, G. E. Protein-bound acrolein: a novel marker of oxidative stress in Alzheimer’s disease. J. Neurochem. 72:751–756; 1999. [28] Sayre, L. M.; Zelasko, D. A.; Harris, P. L. R.; Perry, G.; Salomon, R. G.; Smith, M. A. 4-Hydroxynonenal-derived advanced lipid peroxidation end products are increased in Alzheimer disease. J. Neurochem. 68:2092–2097; 1997. [29] Smith, M. A.; Sayre, L. M.; Anderson, V. E.; Harris, P. L.; Beal, M. F.; Kowall, N.; Perry, G. Cytochemical demonstration of oxidative damage in Alzheimer’s disease by immunochemical enhancement of the carbonyl reaction with 2,4-dinitrophenylhydrazine. J. Histochem. Cytochem. 46:731–735; 1998. [30] Montine, K. S.; Olson, S. J.; Amarnath, V.; Whetsell, W. O.; Graham, D. G.; Montine, T. J. Immunochemical detection of 4-hydroxynonenal adducts in Alzheimer’s disease is associated with APOE4. Am. J. Pathol. 150:437– 443; 1997. [31] Montine, K. S.; Kim, P. J.; Olson, S. J.; Markesbery, W. R.; Montine, T. J. 4-Hydroxy-2-nonenal pyrrole adducts in human neurodegenerative disease. J. Neuropathol. Exp. Neurol. 56:866 – 871; 1997. [32] Montine, K.; Reich, E.; Olson, S. J.; Markesbery, W. R.; Montine, T. Distribution of reducible 4-hydroxynonenal adduct immunoreactivity in Alzheimer’s disease is associated with APOE genotype. J. Neuropathol. Exp. Neurol. 57:415– 425; 1998. [33] Ando, Y.; Brannstrom, T.; Uchida, K.; Nyhlin, N.; Nasman, B.; Suhr, O.; Yamashita, T.; Olsson, T.; Salhy, M.; Uchino, M.; Ando, M. Histochemical detection of 4-hydroxynonenal protein in Alzheimer amyloid. J. Neurol. Sci. 156:172–176; 1998. [34] Smith, M. A.; Keisuke, H.; Hsiao, K.; Pappolla, M. A.; Harris,
[35] [36] [37] [38]
[39]
[40]
[41] [42]
[43]
[44]
[45] [46]
[47]
[48]
[49]
P. L.; Siedlak, S. L.; Massimo, T.; Perry, G. Amyloid- deposition in Alzheimer transgenic mice is associated with oxidative stress. J. Neurochem. 70:2212–2215; 1998. Lovell, M. A.; Xie, C.; Markesbery, W. R. Decreased glutathione transferase activity in brain and ventricular fluid in Alzheimer’s disease. Neurology 51:1562–1566; 1998. Moore, K.; Roberts, L. J. Measurement of lipid peroxidation. Free Radic. Res. 28:659 – 671; 1998. Gutteridge, J. M. C.; Halliwell, B. The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. 15:129 –135; 1990. Reich, E.; Markesbery, W.; Roberts, L. II; Swift, L.; Morrow, J.; Montine, T. Brain regional quantification of F-ring and D/E-ring isoprostanes and neuroprostanes in Alzheimer’s disease. Am. J. Pathol. 158:293–297; 2001. Pratico, D.; Lee, V. M.; Trojanowski, J. Q.; Rokach, J.; Fitzgerald, G. A. Increased F2-isoprostanes in Alzheimer’s disease: evidence for enhanced lipid peroxidation in vivo. FASEB J. 12: 1777–1784; 1998. Nourooz-Zadeh, J.; Liu, E. H.; Yhlen, B.; Anggard, E. E.; Halliwell, B. F4-isoprostanes as specific marker of docosahexaenoic acid production in Alzheimer’s disease. J. Neurochem. 72:734 – 740; 1999. Lovell, M.; Ehmann, W.; Mattson, M.; Markesbery, W. Elevated 4-hydroxynonenal in ventricular fluid in Alzheimer’s disease. Neurobiol. Aging 18:457– 471; 1997. Montine, T. J.; Markesbery, W. R.; Zackert, W.; Sanchez, S. C.; Roberts, L. J.; Morrow, J. D. The magnitude of brain lipid peroxidation correlates with the extent of degeneration but not with density of neuritic plaques or neurofibrillary tangles, or with APOE genotype in Alzheimer’s disease patients. Am. J. Pathol. 155:863– 868; 1999. Montine, T.; Kaye, J.; Montine, K.; McFarland, L.; Morrow, J.; Quinn, J. CSF A42, tau, and F2-isoprostane concentrations in patients with Alzheimer’s disease, other dementias, and agematched controls. Arch. Pathol. Lab. Med. 125:510 –512; 2001. Pratico, D.; Clack, C. M.; Lee, V. M. Y.; Trojanowski, J. Q.; Rokach, J.; FitzGerald, G. Increased 8,12-iso-iPF2␣-IV in Alzheimer’s disease: correlation of a noninvasive index of lipid peroxidation with disease severity. Ann. Neurol. 48:809 – 812; 2000. Beckmam, K. B.; Ames, B. N. The free radical theory of aging matures. Physiol. Rev. 78:547–581; 1998. Montine, T. J.; Shinobu, L.; Montine, K. S.; Roberts, L. J. II; Beal, M. F.; Morrow, J. D. No difference in plasma or urine F2-isoprostanes among patients with Huntington’s disease or Alzheimer’s disease, and controls. Ann. Neurol. 48:950; 2000. Morrow, J. D.; Frei, B.; Longmire, A. W.; Gaziano, M.; Lynch, S. M.; Shyr, Y.; Strauss, W. E.; Oates, J. A.; Roberts, L. J. II Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers. N. Engl. J. Med. 332:1198 –1203; 1995. Feillet-Coudray, C.; Tourtauchaux, R.; Niculescu, M.; Rock, E.; Tauveron, I.; Alexandre-Gouabau, M. C.; Rayssiguier, Y.; Jalenques, I.; Mazur, A. Plasma levels of 8-epiPGF2␣, an in vivo marker of oxidative stress, are not affected by aging or Alzheimer’s disease. Free Radic. Biol. Med. 27:463– 469; 1999. Waddington, E.; Croft, K.; Clarnette, R.; Mori, T.; Martins, R. Plasma F2-isoprostane levels are increased in Alzheimer’s disease: evidence of increased oxidative stress in vivo. Alzheimer’s Rep. 2:277–282; 1999.
PII S0891-5849(02)00807-9
Serial Review: Causes and Consequences of Oxidative Stress in Alzheimer’s Disease Guest Editors: Mark A. Smith and George Perry LIPID PEROXIDATION IN AGING BRAIN AND ALZHEIMER’S DISEASE THOMAS J. MONTINE,* M. DIANA NEELY,* JOSEPH F. QUINN,† M. FLINT BEAL,‡ WILLIAM R. MARKESBERY,§ L. JACKSON ROBERTS II,* and JASON D. MORROW* *Departments of Pharmacology, Pathology, and Medicine and the Vanderbilt Brain Institute, Vanderbilt University Medical Center, Nashville, TN, USA; †Department of Neurology, Veterans Administration Medical Center, Portland, OR, USA; ‡ Department of Neurology and Neuroscience, New York Presbyterian Hospital/Cornell University Medical College, New York, NY, USA; and §Departments of Pathology and Neurology and the Sanders-Brown Center on Aging, University of Kentucky, Lexington, KY, USA (Received 6 February 2002; Accepted 21 February 2002)
Abstract—Lipid peroxidation is one of the major outcomes of free radical-mediated injury that directly damages membranes and generates a number of secondary products, both from fission and endocyclization of oxygenated fatty acids that possess neurotoxic activity. Numerous studies have demonstrated increased lipid peroxidation in brain of patients with Alzheimer’s disease (AD) compared with age-matched controls. These data include quantification of fission and endocyclized products such as 4-hydroxy-2-nonenal, acrolein, isoprostanes, and neuroprostanes. Immunohistochemical and biochemical studies have localized the majority of lipid peroxidation products to neurons. A few studies have consistently demonstrated increased cerebrospinal fluid (CSF) levels of isoprostanes in AD patients early in the course of their dementia, and one study has suggested that CSF isoprostanes may improve the laboratory diagnostic accuracy for AD. Similar analyses of control individuals over a wide range of ages indicate that brain lipid peroxidation is not a significant feature of usual aging. Quantification of isoprostanes in plasma and urine of AD patients has yielded inconsistent results. These results indicate that brain lipid peroxidation is a potential therapeutic target in probable AD patients, and that CSF isoprostanes may aid in the assessment of antioxidant experimental therapeutics and the laboratory diagnosis of AD. © 2002 Elsevier Science Inc. Keywords—Free radicals, Brain, Lipid peroxidation, Aging, Alzheimer’s disease, Isoprostanes, Neuroprostanes
INTRODUCTION
membranes and generation of secondary products. Membrane damage derives from the generation of fragmented fatty acyl chains, lipid-lipid cross-links, endocyclization to produce novel fatty acid esters, and lipid-protein cross-links [1]. In total, these processes combine to produce changes in the biophysical properties of membranes that can have profound effects on the activity of membrane-bound proteins. Secondary products from lipid peroxidation may be divided conceptually into those generated by fragmentation or rearrangement of oxygenated lipid; however, it should be kept in mind that lipid peroxidation is an indiscriminate process and formation of both types of products occurs simultaneously.
Biochemistry of lipid peroxidation Lipid peroxidation is one of the major outcomes of free radical-mediated injury to tissue. Peroxidation of fatty acyl groups, mostly in membrane phospholipids, has three well-described phases: initiation, propagation, and termination. An important aspect of these reactions is that lipid peroxidation will proceed until substrate is consumed or termination occurs. There are two broad outcomes to lipid peroxidation, viz., structural damage to This article is part of a series of reviews on “Causes and Consequences of Oxidative Stress in Alzheimer’s Disease.” The full list of papers may be found on the homepage of the journal. Address correspondence to: Dr. Thomas J. Montine, Alvord Professor of Neuropathology, Department of Pathology, University of Washington, Harborview Medical Center, Box 359791, 325 9th Ave., Seattle, WA 98104, USA; E-Mail: [email protected].
Fragmentation products of lipid peroxidation Fragmentation of lipid hydroperoxides liberates a number of diffusible products, some of which are potent 620
Brain lipid peroxidation in aging and AD
621
Scheme 1. Shown are the structures of some reactive aldehyde products of lipid peroxidation (HNE, 4-hydroxy-2-nonenal; HHE, 4-hydroxy-2hexenal).
electrophiles [2]. The most abundant diffusible products of lipid peroxidation are chemically reactive aldehydes such as malondialdehyde (MDA), acrolein, 4-hydroxy2-hexenal (HHE), and 4-hydroxy-2-nonenal (HNE) (Scheme 1). Several mechanisms have been proposed for the formation of these reactive aldehydes during lipid peroxidation [3,4]. Reactive aldehydes from lipid peroxidation are biologically active. Most studies have focused on their reactivity with a number of cellular nucleophiles, including proteins, nucleic acids, and some lipids [2]. Indeed, many of the cytotoxic effects of lipid peroxidation can be reproduced directly by electrophilic lipid peroxidation products such as HNE [1]. These include depletion of glutathione, dysfunction of structural proteins, reduction in enzyme activities, and induction of cell death. More recent work, much in ex vivo models of the central nervous system, has highlighted alterations in ion channel, transporter, and second messenger system activities following exposure to reactive aldehydes from lipid peroxidation [5–7]. How these many effects are interrelated is not known. We found that microtubule organization and neurite outgrowth were highly vulnerable to HNE exposure in a neuroblastoma cell line [8] (Fig. 1). The mechanisms by which HNE leads to microtubule disruption are not fully clarified, but may involve modification of tubulin [9]. Other lipid peroxidation products, HHE and acrolein, also cause microtubule disruption; however, this may not be the most vulnerable target for these reactive aldehydes [10 –12]. Numerous antioxidant mechanisms have evolved, including several metabolic routes to detoxify fragmentation products from lipid peroxidation. The diffusible reactive aldehydes generated from lipid peroxidation are excellent substrates for glutathione transferases (GSTs) and a number of oxidoreductases that act to detoxify these molecules [2]. Each of the major classes of cytosolic GSTs, ␣, , and , are present throughout rat brain. Two reports have described localization of cytosolic GST immunoreactivity in human brain; one reported GST- and GST- immunoreactivity in glia but not neurons, while the other described GST-␣ and GST-
Fig. 1. Shown are mouse neuroblastoma Neuro2A cells exposed to vehicle or HNE (25 M for 60 min) and microtubules visualized by immunofluorescence microscopy using an anti--tubulin antibody. (A) Control cells show a dense microtubule network that extends to the periphery of the cell. (B) Cells exposed to HNE show severe loss of microtubules. Arrows point to the few remaining microtubules.
localization to glia and vascular cells but not neurons. We have determined the regional, cellular, and class distribution in human brain of the four major oxidoreductases comprising the other major pathways for detoxifying HNE: aldehyde dehydrogenase (ALDH), aldose reductase, aldehyde reductase, and alcohol dehydrogenase (ADH) [13]. Of these four enzymes, only ALDH and aldose reductase are expressed in cerebral cortex, hippocampus, basal ganglia, and midbrain; all four enzymes are present in cerebellum. In cerebrum and hippocampus, aldose reductase was localized to pyramidal neurons and mitochondrial class 2 ALDH was localized to glia, suggesting that neurons and glia may employ different mechanisms to detoxify reactive aldehydes from lipid peroxidation. In addition to liberating diffusible aldehydes, fragmentation of lipid hydroperoxides leaves an abnormally shortened fatty acyl group esterified to lipid. In at least one instance, these chemically modified lipids gain re-
622
T. J. MONTINE et al.
Scheme 2. Shown are the pathways of production and ring structures for F-ring and D/E-ring IsoPs and NeuroPs.
ceptor-binding activity; peroxidation and fragmentation of polyunsaturated fatty acyl groups in phosphatidylcholines can generate platelet activating factor (PAF) analogues that stimulate the PAF receptor [14]. Endocyclized products of lipid peroxidation Rather than pursue fragmentation chemistry, a minority of lipid hydroperoxy radicals undergo internal cyclization to generate endoperoxide intermediates that then are converted to multiple isomers with ring structures analogous to prostaglandins (Scheme 2). When derived from arachidonic acid these compounds are called isoprostanes (IsoPs) and when derived from docosahexaenoic acid these compounds are termed neuroprostanes (NeuroPs) [15,16]. In contrast to arachidonic acid (AA) that is evenly distributed in brain tissue, docosahexaenoic acid (DHA) is highly enriched in neurons in brain [17]. For these reasons it was proposed that NeuroPs may be more sensitive and specific for oxidative damage to neurons in brain. We have characterized both the reduction of the endoperoxide intermediate to F-ring compounds or its isomerization to D- and E-compounds, and have proposed that the ratio of F-ring to D/E-ring compounds reflects the reducing environment in which IsoP or NeuroP formation occurred [18,19]. Unlike prostaglandins, IsoPs and NeuroPs are formed in situ from fatty acids esterified to lipid. Thus, IsoPs and
NeuroPs are thought to contribute to changes in membrane characteristics that accompany lipid peroxidation. Once formed, IsoPs and NeuroPs can undergo hydrolysis to liberate free IsoPs and NeuroPs that are detectable in body fluids [16,20 –22]. Although some products of the IsoP and NeuroP pathways are chemically reactive and can modify cellular proteins, the best studied products of these pathways, D-, E-, and F-ring compounds, do not react with cellular macromolecules, and are not extensively metabolized in situ. Due to their similarity to enzymatically derived prostanoids, several investigators have sought receptor-binding activity for free IsoPs. Indeed, one F2-IsoP isomer has been shown to possess renal and cerebral vasoconstrictor activity, likely through receptor-mediated mechanisms [22–24].
BRAIN LIPID PEROXIDATION
The following sections will review the evidence for brain lipid peroxidation in patients with Alzheimer’s disease and in age-matched control individuals without neurologic disease. Important variables to keep in mind are the methods used to quantify lipid peroxidation, the source of the sample to be analyzed, and whether tissue or fluid was obtained postmortem from definite AD patients or during life from probable AD patients.
Brain lipid peroxidation in aging and AD
Postmortem studies Brain. There is compelling evidence that the magnitude of lipid peroxidation in the brains of AD patients examined postmortem exceeds that in age-matched control individuals. Seminal experiments demonstrated significantly increased thiobarbiturate reactive substances (TBARS) in diseased regions of AD brain obtained postmortem compared to age-matched control individuals [25]. This same group of investigators has shown further that free HNE and acrolein are elevated in diseased regions of AD brain compared to controls [12,26]. Several groups have studied the immunolocalization of protein adducts with reactive aldehydes from lipid peroxidation products. Consistent with the quantitative studies described above, hippocampus and cerebral cortex from AD patients display protein modifications that are not detectable or barely detectable in the corresponding brain regions from age-matched control individuals [27–32]. Also, proteins modified by lipid peroxidation products are present in diseased regions of brain but not in regions uninvolved by AD. Within diseased regions of brain, most studies have observed immunoreactivity with neuronal cytoplasm and neurofibrillary tangles. One group observed that the tissue distribution of HNE-protein adducts varies with apolipoprotein E genotype (APOE) [30,32], but this was not observed by others in a small number of patients [28]. Another study has observed modified proteins in or adjacent to neuritic plaques [33], a distribution that is similar to aged genetically modified mice expressing a mutant human amyloid precursor protein gene [34]. Activity of some of the major metabolizing enzymes for HNE and related aldehydes is altered in brains of AD patients. Interestingly, decreased cytosolic GST activity has been reported in diseased regions of AD brain [35]. In contrast, class 2 ALDH activity is significantly increased in temporal cortex from patients with AD compared to age-matched controls [13]. Combined with the results described above, these findings suggest that increased glial mitochondrial ALDH activity may be a protective response to increased generation of reactive aldehydes from lipid peroxidation in AD brain. Altered metabolic activity of the detoxifying enzymes for reactive aldehydes in AD brain raises serious issues for interpreting the significance of increased levels of these aldehydes in AD brain. Assays for TBARS or specific secondary products, such as HNE, are accurate measures of lipid peroxidation in vitro when metabolism of the lipid peroxidation products does not occur. However, in more complicated systems, extensive metabolism of electrophilic lipid peroxidation products may compromise the accuracy of these assays [36,37]. Due to these limitations in quantifying reactive products of lipid
623
peroxidation, we measured F- and D/E-ring IsoPs and NeuroPs in temporal and parietal cortex, hippocampus, and cerebellum of 9 definite AD patients and 11 agematched controls, all with short postmortem intervals [38]. Total NeuroP (F- plus D/E-ring), but not total IsoP, levels are greater in AD patients than controls, but only in those brain regions involved by AD. The F- to D/Ering ratio for NeuroPs, but not IsoPs, was 40 to 70% lower in AD patients compared to controls; however, this occurred in all brain regions and not only those involved by AD. These data indicate greater free radical damage in DHA- than AA-containing compartments in diseased regions of AD brain, and diminished reducing capacity in DHA-containing compartments throughout AD brain. Using a very similar gas chromatography/mass spectrometric technique, others have demonstrated elevated F2IsoP levels in frontal and temporal lobes of AD patients compared to controls [39], a finding that was confirmed in the study cited above. Another group has reported that F4-NeuroPs (called F4-isoprostanes in their publication) are elevated in temporal and occipital, but not parietal, lobes of AD patients compared to controls, and that F4-NeuroPs are greater than F2-IsoPs in these regions [40]. However, interpretation of data from this study is limited by long postmortem intervals (average 47 h in AD patients) [40]. Cerebrospinal fluid. Cerebrospinal fluid (CSF) also has been investigated as a source of central nervous system tissue for the assessment of lipid peroxidation in AD brain. One study has measured free HNE in CSF obtained from the lateral ventricles postmortem and shown that its concentration is significantly elevated in AD patients compared to age-matched controls [41]. A few studies have determined the concentration of F2-IsoPs, and in one case F4-NeuroPs, in CSF obtained from the lateral ventricles postmortem and have shown significant elevations in AD patients compared to age-matched controls [20,42]. Importantly, CSF F2-IsoP concentrations in AD patients are significantly correlated with global indices of neurodegeneration, decreasing brain weight, degree of cerebral cortical atrophy, and increasing Braak stage, but not with APOE or the tissue density of neuritic plaques or neurofibrillary tangles [42]. It is noteworthy that all of the postmortem CSF studies in AD patients used material from individuals with very short (average 2 to 3 h) postmortem intervals. Intra vitam studies CSF. AD patients undergoing postmortem examination typically have advanced disease and an average duration of dementia of 8 to 12 years. Therefore, a serious limitation to interpretation of results from tissue obtained
624
T. J. MONTINE et al.
Fig. 2. Shown is the frequency distribution for F2-IsoP concentration in CSF obtained from the lumbar cistern of 36 controls and 41 probable AD patients. Similar to other CSF biomarkers of AD, there is considerable overlap between the two groups. However, values for probable AD patients are skewed significantly towards higher levels (MannWhitney test, p ⬍ .001).
postmortem is that the increased brain lipid peroxidation in AD patients might be a late stage consequence of disease. Obviously, a late stage consequence of AD would be a less attractive therapeutic target than a process contributing to disease progression at an earlier stage. The first study of probable AD patients early in the course of dementia showed that F2-IsoPs are significantly elevated in CSF obtained from the lumbar cistern compared to age-matched hospitalized patients without neurological disease [21]. The average duration of dementia in these probable AD patients at the time of CSF examination was less than 2 years, while the average duration of AD is between 8 and 12 years. This result was observed again in another group of probable AD patients and controls [43]. The combined results from these two studies are presented in Fig. 2. Importantly, a different laboratory examining a third group of probable AD patients and controls using a similar mass spectrometric method obtained similar results and confirmed the difference observed between AD patients and controls [44]. This last study suggested a correlation between higher CSF F2-IsoP levels and homozygosity for the 4 allele of APOE, an association not observed in other CSF and postmortem studies of AD patients [20,21,38,42]. In addition to providing mechanistic information about AD pathogenesis and a means to quantitatively assess response to antioxidant therapeutics, CSF F2-IsoP levels also may provide information that is useful in diagnosing AD. We tested the hypothesis that quantification of CSF F2-IsoPs, along with CSF A42 and tau levels, improves laboratory diagnostic accuracy for AD in patients with probable AD, dementias other than AD, and age-matched controls [43]. Individuals were classi-
Fig. 3. Shown is a scatter plot of F2-IsoP concentration in CSF obtained from the lumbar cistern of 42 control individuals of varying ages. Open symbols are for samples collected at the Massachusetts General Hospital and closed symbols are for samples collected at Oregon Health Sciences University. Neither group alone had a significant correlation between age and CSF F2-IsoP concentration. Correlation analysis for the combined groups yielded R2 ⫽ 0.00 and p ⬎ .7.
fied as AD or non-AD by a commercially available test using CSF A42 and tau levels (95% sensitivity, 50% specificity), by CSF F2-IsoP and A42 (90% sensitivity, 83% specificity), and by combined analysis using CSF F2-IsoP, A42, and tau levels (84% sensitivity, 89% specificity). These results indicate that CSF F2-IsoP quantification can enhance the accuracy of the laboratory diagnosis of AD; however, this conclusion is based on a single study and these findings need to be replicated. Each of these CSF-based studies has contained a group of age-matched control individuals. In none of these studies did we observe an association between CSF F2-IsoP concentration and age in controls; however, the range of ages examined was limited because individuals were selected to be age-matched for AD patients. Recently, we have augmented these data by determining CSF F2-IsoP levels in control individuals over a broad range of ages (Fig. 3). Our results show no relationship between age over the third to ninth decades of life and CSF F2-IsoP concentrations. These data do not support the model of progressively increasing oxidative damage to the human central nervous system with advancing age. However, these findings are consistent with a model suggested from studies in other species, viz., oxidative damage is exponentially related to advancing age and is not significantly elevated until the extremes of age [45]. Plasma and urine. Although obtaining CSF from the lumbar cistern is not associated with significant risks, even in the elderly, spinal taps are stressful to the patient and not easily obtained in most clinics. For these reasons, several investigators have pursued quantification of F2-
Brain lipid peroxidation in aging and AD
IsoPs in plasma or urine. Like most data for peripheral biomarkers of neurodegenerative disease, the results have been conflicting. In our study, plasma and urine were collected and analyzed over a 2 year period [46]. Over the ensuing 3 years, 23 of the 25 probable AD patients died and 17 underwent postmortem examination; 12 were diagnosed with definite AD, 4 with Dementia with Lewy Bodies, and 1 with Pick’s disease. There was no significant difference in plasma or urine F2-IsoP levels between definite AD patients and age-matched controls or definite AD patients and probable AD patients. Importantly, all individuals in this study were nonsmokers, a behavior that significantly influences plasma F2-IsoP levels [47]. At the time of sample collection, only 4 individuals were taking multivitamin supplements; their plasma and urine F2-IsoP levels did not differ from those not taking supplements. Another study, using an antibody-based detection method, also concluded that plasma F2-IsoP levels are not elevated in AD patients compared to controls [48]. In contrast, one group using an antibody-based method and another using a gas chromatography/mass spectrometric method both concluded that F2-IsoP levels were increased in AD patients’ plasma or urine compared to controls [44,49]. There are several possible reasons for this discrepancy; however, it is unlikely to be methodological because all of these methods have yielded similar results for CSF levels from AD patients and controls. More likely, the differences reside in criteria used to include or exclude patients and controls. For example, one study did not exclude smokers from the patient group [44]. In summary, results from these studies do not reproducibly support the hypothesis that patients with AD experience increased levels of systemic oxidative stress or that the increased levels of F2-IsoPs observed in CSF are detectable peripherally.
SUMMARY
Results from these studies clearly show that diseased regions of brain from patients with advanced AD have increased levels of lipid peroxidation products compared to controls. Importantly, patients with probable AD early in the course of their dementia have increased CSF F2-IsoP levels compared to age-matched controls. In combination with a large number of additional studies showing in vitro neurotoxic activity for reactive products from lipid peroxidation, these results indicate that lipid peroxidation likely significantly contributes to disease progression and is a potential therapeutic target in AD patients.
625
Acknowledgements — This work was supported by grants from the NIH (AG00774, AG16835, AG05114, GM15431, DK26657, and CA77839) as well as grants from the Alzheimer’s Association, the Abercrombie Foundation, and a Bourroughs Welcome Fund Clinical Scientist Award in Translational Research to J.D.M.
REFERENCES [1] Farber, J. L. Mechanisms of cell injury. In: Craighead, J. E., ed. Pathology of environmental and occupational disease. St. Louis, MO: Mosby-Year Book Inc.; 1995:287–302. [2] Esterbauer, H.; Schaur, R. J.; Zollner, H. Chemistry and biochemistry of 4-hydroxynonenal, malondialdehyde and related aldehydes. Free Radic. Biol. Med. 11:81–128; 1991. [3] Schneider, C.; Tallman, K. A.; Porter, N. A.; Brash, A. R. Two distinct pathways of formation of 4-hydroxynonenal: mechanisms of non-enzymatic transformation of the 9- and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J. Biol. Chem. 276:20831– 20838; 2001. [4] Porter, N. A.; Caldwell, S. E.; Mills, K. A. Mechanisms of free radical oxidation of unsaturated lipids. Lipids 30:277–290; 1995. [5] Mattson, M. P. Modification of ion homeostasis by lipid peroxidation: roles in neuronal degeneration and adaptive plasticity. Trends Neurosci. 20:53–57; 1998. [6] Keller, J. N.; Mattson, M. P. Roles of lipid peroxidation in modulation of cellular signaling pathways, cell dysfunction, and death in the nervous system. Rev. Neurosci. 9:105–116; 1998. [7] Smith, M.; Sayre, L.; Monnier, V.; Perry, G. Radical AGEing in Alzheimer’s disease. Trends Neurosci. 18:172–176; 1995. [8] Neely, M. D.; Sidell, K. R.; Graham, D. G.; Montine, T. J. The lipid peroxidation product 4-hydroxynonenal inhibits neurite outgrowth and disrupts neuronal microtubules. J. Neurochem. 72: 2323–2333; 1999. [9] Neely, M. D.; Zimmerman, L.; Picklo, M. J.; Ou, J. J.; Morales, C. R.; Montine, K. S.; Amarnath, V.; Montine, T. J. Congeners of N␣-acetyl-L-cysteine but not aminoguanidine act as neuroprotectants from the lipid peroxidation product 4-hydroxy-2-nonenal. Free Radic. Biol. Med. 29:1028 –1036; 2000. [10] Picklo, M. J.; Montine, T. J. Acrolein inhibits respiration in isolated brain mitochondria. Biochim. Biophys. Acta 1535:145– 152; 2001. [11] Lovell, M. A.; Xie, C.; Markesbery, W. R. Acrolein, a product of lipid peroxidation, inhibits glucose and glutatmate uptake in primary neuronal cultures. Free Radic. Biol. Med. 28:714 –720; 2000. [12] Lovell, M. A.; Xie, C.; Markesbery, W. R. Acrolein is increased in Alzheimer’s disease brain and is toxic to primary hippocampal cultures. Neurobiol. Aging 22:187–194; 2001. [13] Picklo, M. J.; Olson, S. J.; Markesbery, W. R.; Montine, T. J. Expression and activities of aldo-keto oxidoreductases in Alzheimer’s disease. J. Neuropathol. Exp. Neurol. 60:686 – 695; 2001. [14] McIntyre, T. M.; Zimmerman, G. A.; Prescott, S. M. Biologically active oxidized phospholipids. J. Biol. Chem. 274:25189 –25192; 1999. [15] Morrow, J.; Hill, K.; Burk, R.; Nammour, T.; Badr, K.; Roberts, L. A series of prostaglandin-like compounds produced in vivo in humans by a non-cyclooxygenase, free radical catalyzed mechanism. Proc. Natl. Acad. Sci. USA 87:9383–9387; 1990. [16] Roberts, L. J.; Montine, T. J.; Markesbery, W. R.; Tapper, A. R.; Hardy, P.; Chemtob, S.; Detbarn, W. D.; Morrow, J. D. Formation of isoprostane-like compounds (neuroprostanes) in vivo from docosahexaenoic acid. J. Biol. Chem. 273:13605–13612; 1998. [17] Salem, N.; Kim, H. Y.; Yergey, J. A. Docosahexaenoic acid: membrane function and metabolism. In: Simopoulos, A. P.; Kifer, R. R.; Martin, R. E., eds. Health effects of polyunsaturated acids in seafoods. New York: Academic Press; 1986:263–317. [18] Reich, E. E.; Zackert, W. E.; Brame, C. J.; Chen, Y.; Roberts, L. J. II; Hachey, D. L.; Montine, T. J.; Morrow, J. D. Formation of novel D-ring and E-ring isoprostane-like compounds (D4/E4-neuroprostanes) in vivo from docosahexaenoic acid. Biochemistry 39:2376 –2383; 2000.
626
T. J. MONTINE et al.
[19] Morrow, J. D.; Roberts, L. J.; Daniel, V. C.; Awad, J. A.; Mirochnitchenko, O.; Swift, L. L.; Burk, R. F. Comparison of formation of D2/E2-isoprostanes and F2-isoprostanes in vitro and in vivo— effects of oxygen tension and glutathione. Arch. Biochem. Biophys. 353:160 –171; 1998. [20] Montine, T. J.; Markesbery, W. R.; Morrow, J. D.; Roberts, L. J. Cerebrospinal fluid F2-isoprostanes are increased in Alzheimer’s disease. Ann. Neurol. 44:410 – 413; 1998. [21] Montine, T. J.; Beal, M. F.; Cudkowicz, M. E.; Brown, R. H.; O’Donnell, H.; Margolin, R. A.; McFarland, L.; Bachrach, A. F.; Zackert, W. E.; Roberts, L. J.; Morrow, J. D. Increased cerebrospinal fluid F2-isoprostane concentration in probable Alzheimer’s disease. Neurology 52:562–565; 1999. [22] Morrow, J. D.; Roberts, L. J. The isoprostanes: unique bioactive products of lipid peroxidation. Prog. Lipid. Res. 36:1–21; 1997. [23] Hou, X.; Gobeil, F.; Peri, K.; Speranza, G.; Marrache, A. M.; Lachapelle, P.; Roberts, L. J. II; Varma, D. R.; Chemtob, S. Augmented vasoconstriction and thromboxane formation by 15F2t-isoprostane (8-isoprostaglandin F2␣) in immature pig periventricular brain microvessels. Stroke 31:516 –525; 2000. [24] Hoffman, S. W.; Moore, S.; Ellis, E. F. Isoprostanes: free radicalgenerated prostaglandins with constrictor effects on cerebral arterioles. Stroke 28:844 – 849; 1997. [25] Lovell, M.; Ehmann, W.; Butler, S.; Markesbery, W. Elevated thiobarbituric acid-reactive substances and antioxidant enzyme activity in the brain in Alzheimer’s disease. Neurology 45:1594 – 1601; 1995. [26] Markesbery, W. R.; Lovell, M. A. Four-hydroxynonenal, a product of lipid peroxidation, is increased in the brain in Alzheimer’s disease. Neurobiol. Aging 19:33–36; 1998. [27] Calingasan, N. Y.; Uchida, K.; Gibson, G. E. Protein-bound acrolein: a novel marker of oxidative stress in Alzheimer’s disease. J. Neurochem. 72:751–756; 1999. [28] Sayre, L. M.; Zelasko, D. A.; Harris, P. L. R.; Perry, G.; Salomon, R. G.; Smith, M. A. 4-Hydroxynonenal-derived advanced lipid peroxidation end products are increased in Alzheimer disease. J. Neurochem. 68:2092–2097; 1997. [29] Smith, M. A.; Sayre, L. M.; Anderson, V. E.; Harris, P. L.; Beal, M. F.; Kowall, N.; Perry, G. Cytochemical demonstration of oxidative damage in Alzheimer’s disease by immunochemical enhancement of the carbonyl reaction with 2,4-dinitrophenylhydrazine. J. Histochem. Cytochem. 46:731–735; 1998. [30] Montine, K. S.; Olson, S. J.; Amarnath, V.; Whetsell, W. O.; Graham, D. G.; Montine, T. J. Immunochemical detection of 4-hydroxynonenal adducts in Alzheimer’s disease is associated with APOE4. Am. J. Pathol. 150:437– 443; 1997. [31] Montine, K. S.; Kim, P. J.; Olson, S. J.; Markesbery, W. R.; Montine, T. J. 4-Hydroxy-2-nonenal pyrrole adducts in human neurodegenerative disease. J. Neuropathol. Exp. Neurol. 56:866 – 871; 1997. [32] Montine, K.; Reich, E.; Olson, S. J.; Markesbery, W. R.; Montine, T. Distribution of reducible 4-hydroxynonenal adduct immunoreactivity in Alzheimer’s disease is associated with APOE genotype. J. Neuropathol. Exp. Neurol. 57:415– 425; 1998. [33] Ando, Y.; Brannstrom, T.; Uchida, K.; Nyhlin, N.; Nasman, B.; Suhr, O.; Yamashita, T.; Olsson, T.; Salhy, M.; Uchino, M.; Ando, M. Histochemical detection of 4-hydroxynonenal protein in Alzheimer amyloid. J. Neurol. Sci. 156:172–176; 1998. [34] Smith, M. A.; Keisuke, H.; Hsiao, K.; Pappolla, M. A.; Harris,
[35] [36] [37] [38]
[39]
[40]
[41] [42]
[43]
[44]
[45] [46]
[47]
[48]
[49]
P. L.; Siedlak, S. L.; Massimo, T.; Perry, G. Amyloid- deposition in Alzheimer transgenic mice is associated with oxidative stress. J. Neurochem. 70:2212–2215; 1998. Lovell, M. A.; Xie, C.; Markesbery, W. R. Decreased glutathione transferase activity in brain and ventricular fluid in Alzheimer’s disease. Neurology 51:1562–1566; 1998. Moore, K.; Roberts, L. J. Measurement of lipid peroxidation. Free Radic. Res. 28:659 – 671; 1998. Gutteridge, J. M. C.; Halliwell, B. The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. 15:129 –135; 1990. Reich, E.; Markesbery, W.; Roberts, L. II; Swift, L.; Morrow, J.; Montine, T. Brain regional quantification of F-ring and D/E-ring isoprostanes and neuroprostanes in Alzheimer’s disease. Am. J. Pathol. 158:293–297; 2001. Pratico, D.; Lee, V. M.; Trojanowski, J. Q.; Rokach, J.; Fitzgerald, G. A. Increased F2-isoprostanes in Alzheimer’s disease: evidence for enhanced lipid peroxidation in vivo. FASEB J. 12: 1777–1784; 1998. Nourooz-Zadeh, J.; Liu, E. H.; Yhlen, B.; Anggard, E. E.; Halliwell, B. F4-isoprostanes as specific marker of docosahexaenoic acid production in Alzheimer’s disease. J. Neurochem. 72:734 – 740; 1999. Lovell, M.; Ehmann, W.; Mattson, M.; Markesbery, W. Elevated 4-hydroxynonenal in ventricular fluid in Alzheimer’s disease. Neurobiol. Aging 18:457– 471; 1997. Montine, T. J.; Markesbery, W. R.; Zackert, W.; Sanchez, S. C.; Roberts, L. J.; Morrow, J. D. The magnitude of brain lipid peroxidation correlates with the extent of degeneration but not with density of neuritic plaques or neurofibrillary tangles, or with APOE genotype in Alzheimer’s disease patients. Am. J. Pathol. 155:863– 868; 1999. Montine, T.; Kaye, J.; Montine, K.; McFarland, L.; Morrow, J.; Quinn, J. CSF A42, tau, and F2-isoprostane concentrations in patients with Alzheimer’s disease, other dementias, and agematched controls. Arch. Pathol. Lab. Med. 125:510 –512; 2001. Pratico, D.; Clack, C. M.; Lee, V. M. Y.; Trojanowski, J. Q.; Rokach, J.; FitzGerald, G. Increased 8,12-iso-iPF2␣-IV in Alzheimer’s disease: correlation of a noninvasive index of lipid peroxidation with disease severity. Ann. Neurol. 48:809 – 812; 2000. Beckmam, K. B.; Ames, B. N. The free radical theory of aging matures. Physiol. Rev. 78:547–581; 1998. Montine, T. J.; Shinobu, L.; Montine, K. S.; Roberts, L. J. II; Beal, M. F.; Morrow, J. D. No difference in plasma or urine F2-isoprostanes among patients with Huntington’s disease or Alzheimer’s disease, and controls. Ann. Neurol. 48:950; 2000. Morrow, J. D.; Frei, B.; Longmire, A. W.; Gaziano, M.; Lynch, S. M.; Shyr, Y.; Strauss, W. E.; Oates, J. A.; Roberts, L. J. II Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers. N. Engl. J. Med. 332:1198 –1203; 1995. Feillet-Coudray, C.; Tourtauchaux, R.; Niculescu, M.; Rock, E.; Tauveron, I.; Alexandre-Gouabau, M. C.; Rayssiguier, Y.; Jalenques, I.; Mazur, A. Plasma levels of 8-epiPGF2␣, an in vivo marker of oxidative stress, are not affected by aging or Alzheimer’s disease. Free Radic. Biol. Med. 27:463– 469; 1999. Waddington, E.; Croft, K.; Clarnette, R.; Mori, T.; Martins, R. Plasma F2-isoprostane levels are increased in Alzheimer’s disease: evidence of increased oxidative stress in vivo. Alzheimer’s Rep. 2:277–282; 1999.