Hippocampus mitochondrial dysfunction with impaired electron transfer and increased oxidative damage was observed upon rat aging. Hippocampal mitochondria of aged (12 mo) and senescent (20 mo) rats showed, compared with young (4 mo) rats, marked decreases in the rate of state 3 respiration with NAD-dependent substrates (32–51%) and in the activities of mitochondrial complexes I (57–73%) and IV (33–54%). The activity of mitochondrial nitric oxide synthase was also decreased, 53–66%, with age. These losses in enzymatic activity were more marked in the hippocampus than in brain cortex or in whole brain. The histochemical assay of mitochondrial complex IV in the hippocampus showed decreased staining upon aging. Oxidative damage, determined as the mitochondrial content of thiobarbituric-acid reactive substances (TBARS) and protein carbonyls, increased in aged and senescent hippocampus (66–74% in TBARS and 48–96% in carbonyls). A significant statistical correlation was observed between mitochondrial oxidative damage and enzymatic activity. Mitochondrial dysfunction with shortage of energy supply is considered a likely cause of dysfunction in aged hippocampus.
- mitochondrial nitric oxide synthase
- reduced nicotinamide adenide dinucleotide dehydrogenase
- cytochrome oxidase
- oxidative damage
mammalian aging is characterized by a gradual and continuous loss, starting at full adulthood, of the quality of physiological functions and responses. The losses are more marked in the functions that depend on the integrated response of the central nervous system (19) than in the functions of the renal or cardiovascular systems. Mitochondria were brought to attention in mammalian aging biology because of the central role of mitochondria in producing biochemical energy (ATP) to meet cellular requirements in aerobic cells and to the decline of basal metabolic rate and of physical performance that are characteristic of aging (32).
The free radical theory of aging, based on the pioneer works of Gerschman et al. (18) and Harman (20), considers that aging is caused by the continuous inactivation of biologically essential macromolecules and subcellular structures due to chemical modifications produced by reactions mediated by oxygen free radicals. When the free radical theory of aging is focused in mitochondria, it emerges as the mitochondrial theory of aging (6, 21, 32, 47). Mitochondria are considered likely pacemakers of tissue aging because of their continuous production of superoxide radical (O2•−) and of nitric oxide (NO) and to the mitochondrial sensitivity to free radical-mediated oxidative damage (32).
Aged mammalian brain shows a decreased capacity to produce ATP by oxidative phosphorylation, and it is considered that this decreased capacity for energy production becomes limiting under physiological conditions in aged individuals. The current knowledge is that the impairment of mitochondrial function is due to decreased rates of electron transfer, with diminished activities of complexes I and IV, whereas inner membrane H+ impermeability and F1-ATP synthase activity are not significantly affected by aging (12, 32, 35).
Ames et al. (2) postulated that mitochondrial oxidants are the main source of the oxidative damage that accumulates with age and that these oxidation products are major contributors to cellular, tissue, and organism aging. Mitochondria isolated from the brain of aged animals show increased contents of the oxidation products of phospholipids, proteins, and DNA, decreased membrane potential, and increased size and fragility (6, 32). Mitochondrial dysfunction has been observed in whole brain mitochondria of aged rodents (6, 12, 32–37, 48), and it is likely that the phenomenon is more marked in specific brain areas. Hippocampus is a median temporal lobe formation that is atrophied during aging and in some neurological diseases, processes that are associated with memory and cognitive impairments (16). For instance, patients with early to moderate Alzheimer's disease have an ∼25% smaller hippocampal volume than healthy age-matched subjects (13, 24, 26) and show a postmortem marked ultrastructural damage in hippocampal mitochondria and neurons (4).
However, there are no available data on hippocampal mitochondrial function in aging. This lack of information is because of the unavailability of an effective isolation procedure for rat hippocampal mitochondria, a situation that was overcame by a recent report (28). In the present study, we describe the gradual impairment of mitochondrial function in rat hippocampus during aging compared with less marked mitochondrial dysfunctions observed in brain cortex and whole brain. The activity of mitochondrial nitric oxide synthase (mtNOS) was determined and considered in relation to the role of NO in mitochondrial biogenesis (38–39).
MATERIALS AND METHODS
Adult male Wistar rats, young (4 mo old, 262 ± 10 g), aged (12 mo old, 573 ± 16 g), and senescent (20 mo old, 731 ± 20 g), were used (n = 32/each age: 8 rats for whole brain and 24 rats for brain cortex and hippocampus in pools of 4 rats). The animals were grown at the Department of Experimental Animals of the University of Cadiz, housed in groups of three rats, and kept at 22 ± 2°C with 12:12-h light-dark cycles and with full access to water and food. Experiments were carried out in accordance with the Guiding Principles for Research Involving Animals and Human Beings of the American Physiological Society, the Guidelines of the European Union Council (86/609/CEE), and the Spanish regulations (BOE 67/8509-12, 1988) for laboratory animals and were approved by the Scientific Committee of the University of Cádiz.
Rats were killed by decapitation. The brain was quickly removed from the calvarium by opening the lateral sides of the skull with scissors as described by Palkovits and Brownstein (41). The rapid (2-min) dissection of rat hippocampus involved: 1) a lineal incision from the posterotemporal pole of one hemisphere to the anterior pole of the frontal lobe at a 30° angle (step 1) with the incision depth adjusted to cut only the cortex and subjacent corpus callosum; 2) incision in the opposite hemisphere and separation of the central cortex area (step 2); and 3) removal of the two hippocampal formations (step 3) as indicated in Fig. 1. After separation of the hippocampus, dorsomedial and dorsolateral cortex areas were separated by cutting out the outside gray areas. The procedure yielded 384 ± 12, 596 ± 15, and 629 ± 13 mg of brain cortex/rat of 4, 12, and 20 mo of rat age. The excised areas were washed and immersed in 230 mM mannitol, 70 mM sucrose, 1.0 mM EDTA, and 10 mM Tris·HCl, pH 7.4, at 0°C.
Isolation of mitochondria.
Pools of four rats were used for hippocampus and cortex and one rat for whole brain. Mitochondria were isolated from tissues homogenized in 230 mM mannitol, 70 mM sucrose, 1.0 mM EDTA, and 10 mM Tris·HCl, pH 7.40, at a ratio of 9 ml of homogenization medium/g of tissue in a Potter homogenizer with a Teflon pestle. The homogenate was centrifuged at 700 g for 10 min and the supernatant at 8,000 g for 10 min to pellet mitochondria that were washed in the same conditions (10, 28, 35) to obtain mitochondrial preparations with 0.24 ± 0.01, 0.22 ± 0.01, and 0.20 ± 0.1 nmol cytochrome aa3/mg protein for young, old, and senescent rats (P < 0.05 old vs. young rats), in agreement with the classical procedure for isolation of rat brain mitochondria (8, 11). The procedure used in this study provides hippocampal mitochondria with higher respiratory rates and with a higher contamination by other neural structures and enzymes than the classic and slower procedures using Ficoll gradients (11, 12). Isolated hippocampal mitochondria had a lactate dehydrogenase activity of 20–30 nmol·min−1·mg protein−1 and an acetylcholinesterase activity of 50–70 nmol·min−1·mg protein−1. Mitochondrial suspensions (∼20 mg protein/ml) were used immediately after isolation for respiration measurements or frozen in liquid N2 and kept at −80°C for other determinations. Mitochondrial fragments were obtained from mitochondria that were twice frozen and thawed and homogenized each time by passage through a tuberculin needle and were used for determination of mitochondrial enzymatic activities and of markers of oxidative damage. The protein content of mitochondrial preparations was determined using the Folin reagent and BSA as standard.
Mitochondrial oxygen uptake and mtNOS functional activity.
Mitochondrial O2 uptake was measured with a Clark electrode in a 1.5-ml chamber at 30°C in an air-saturated reaction medium consisting of 230 mM mannitol, 70 mM sucrose, 20 mM Tris·HCl, pH 7.40, 1.0 mM EDTA, 5.0 mM phosphate, 4.0 mM MgCl2, and 0.5–0.7 mg mitochondrial protein/ml at pH 7.40 (10). Respiratory rates were determined with 10 mM succinate or 5.0 mM malate-5.0 mM glutamate as substrates. State 4 respiration was determined without ADP addition, and state 3 respiration was established by addition of 0.50 mM ADP. Respiration is expressed in nanogram-atoms oxygen per minute per milligram protein. The respiratory control ratio, an indication of the quality of mitochondrial coupling in the preparation, was determined as state 3 rate/state 4 rate (10).
mtNOS functional activity was assayed by the determination of the difference between the rates of state 3 respiration in the presence of arginine (substrate) and in the presence of a NOS inhibitor. The difference in the measurements indirectly determines mtNOS activity (35, 46).
Mitochondrial electron transfer activities.
The enzyme activities of complexes I-III, II-III, and IV were determined spectrophotometrically at 30°C with the mitochondrial fragments suspended in 100 mM phosphate buffer, pH 7.3. For NADH-cytochrome c reductase (complexes I and III) and succinate-cytochrome c reductase (complexes II and III), mitochondrial membranes were added with 0.20 mM NADH or 5.0 mM succinate as substrates, 0.10 mM cytochrome c3+, and 1.0 mM KCN, and the enzymatic activity was determined at 550 nm (ε = 19 mM−1/cm) and expressed as nanomole cytochrome c reduced per minute per milligram protein. Cytochrome oxidase (complex IV) was determined in the same buffer added with 0.10 mM cytochrome c2+, prepared by reduction with NaBH4 and HCl. The rate of cytochrome c oxidation was calculated as the first-order reaction constant per milligram protein and expressed as nanomole cytochrome c oxidized at 10 μM cytochrome c per minute per milligram protein, which gives electron transfer rates of the order of physiological respiration (33–37).
Spectrophotometric determination of mtNOS activity.
Mitochondrial NO production was determined by the oxyhemoglobin (HbO2) oxidation assay at 30°C (9). The reaction medium consisted of 0.10 mM NADPH, 0.20 mM arginine, 1.0 mM CaCl2, 4.0 μM Cu,Zn-SOD, 0.10 μM catalase, and 25 μM HbO2 heme in 50 mM phosphate, pH 5.8, and 0.5–0.7 mg mitochondrial protein/ml. A diode array spectrophotometer (model 8453; Agilent, Palo Alto, CA) was used to follow the absorbance change at 577 nm with a reference wavelength at the isosbestic point of 591 nm (ε577–591 = 11.2 mM−1/cm). Production of NO was calculated from the absorbance change that was inhibited by 2 mM NG-methyl-l-arginine, usually 92–96%, and expressed as nanomole NO per minute per milligram protein.
Biochemical markers of oxidative damage.
Protein carbonyls and thiobarbituric acid reactive substances (TBARS) were determined in mitochondrial membranes by the assays of Fraga et al. (17) and of Oliver et al. (40), modified as described (34–37) and expressed as picomoles per milligram of mitochondrial protein.
Histological determination of complex I and complex IV.
Rats were anesthetized with chloral hydrate (35 mg/100 g body wt) and transcardially perfused with 4% formaldehyde in 0.120 M phosphate buffer, 20 mM CaCl2 (pH 7.40). Brains were rapidly removed, immersed in the same fixative during 15 min, and frozen. Serial 20-μm cryostat sections were placed in cold 0.120 M phosphate buffer (pH 7.40). Groups of one young, one aged, and one senescent rat were processed simultaneously. Complex I was assayed histochemically by the method of NADH-tetrazolium reductase activity (5). Hippocampal sections were incubated in 1.40 mM nitroblue tetrazolium, 5.0 mM MgCl2, and 2.5 mM NADH in 60 mM phosphate buffer, pH 7.40 during 30 min at 37°C. Control sections omitting NADH were used. The sections were dehydrated by exposure to graded ethanol rinses, cleared with Rio-Hortega mixture (carboxilol-creosote), and mounted in gelatinized slides. Complex IV was assayed histochemically by a modification of the classical diaminobenzidine method of Seligman et al. (45) intensified with heavy metals (44). Hippocampal sections were incubated in 1.40 mM diaminobenzidine, 40 μM cytochrome c, 2 μg/ml catalase, 1% CoCl2, and 1% NiSO4 in 60 mM phosphate buffer, pH 7.40, during 15 min at 37°C. Control sections were preincubated 5 min with 10 mM KCN and incubated in reaction medium supplemented with 10 mM KCN. Samples were dehydrated, cleared, and mounted as described above. High-resolution digital microphotographs were obtained in a Nikon Optiphot-2 microscope with a DXM1200F digital camera and processed with Image-J software (National Institutes of Health, Bethesda, MD) (1).
The differences between groups were analyzed by Student-Newman-Keul's as post hoc test after significant one-way ANOVA. A P value of <0.05 was considered statistically significant. Statistical analyses were carried out using the statistical package SPSS 11.5 for Windows.
Brain weights were 1.84 ± 0.05, 2.03 ± 0.06, and 2.37 ± 0.07 g at 4, 12, and 20 mo of rat age, and right and left hippocampi (together) weighed 156 ± 9, 165 ± 8, and 182 ± 9 mg at 4, 12, and 20 mo of rat age. The hippocampus/whole brain weight ratios were 8.47, 8.12, and 7.67% at 4, 12, and 20 mo of rat age, indicating that the rat hippocampus ages with a moderate atrophy, reaching ∼10% at 20 mo of age, which is comparable to the normal aging of human brain.
The respiration of rat hippocampal mitochondria with the NAD-dependent substrates malate-glutamate in the active state 3 (i.e., the respiratory activity that sustains ATP formation) decreased significantly with aging (Table 1). The observed state 3 respiratory rates and the derived respiratory control ratios are slightly higher than the ones first reported for rat hippocampal mitochondria (28) and are in the range of the values for whole brain (35). Compared with young rats (4 mo old), the state 3 respiration with malate-glutamate as substrate was decreased in aged (12 mo old) and senescent (20 mo old) rats to 92–80% in whole brain, to 79–70% in brain cortex, and to 68–49% in the hippocampus (the first and second values refer to 12- and 20-mo-old rats) (Table 1), indicating that mitochondrial dysfunction was more marked in the hippocampus than in the cortex or in the whole brain. The decrease in respiratory rate in whole brain mitochondria of senescent rats (20%) agrees with a similar report for whole brain mitochondria of senescent mice (35). Using succinate as substrate, the inhibition of state 3 respiration of hippocampal mitochondria was only 29% in senescent rats (Table 1). The higher respiratory impairment with NAD-dependent substrates reflects the higher sensitivity of complex I (NADH-ubiquinone reductase) and the decreased respiration with succinate as substrate reflects complex IV inhibition (see below). It is possible that the determined mitochondrial impairment is an underestimation because of the loss of dysfunctional mitochondria during isolation due to their increased size and fragility.
The partial inactivation of hippocampal complex I upon aging was confirmed by determination of the enzymatic activities of the respiratory complexes. The activities of complex I (determined as complex I + III activity) and of complex IV decreased significantly with age, whereas complex II activity (determined as complex II + III activity) was not significantly affected by aging (Table 2). Aged and senescent rats showed decreases of complex I activity of 17–35% in whole brain, 16–30% in the cortex, and 57–73% in the hippocampus. Similarly, the decreases in complex IV (cytochrome oxidase) activities were 19–36% in whole brain, 23–36% in the cortex, and 33–54% in the hippocampus. The loss in complex I activity was greater when assayed as complex I + III activity than when it was estimated from the state 3 rate of respiration with malate-glutamate, which indicates that normal complex I activity moderately exceeds the requirements for state 3 respiration.
The activity of mtNOS, the NOS isoenzyme located at the inner mitochondrial membrane involved in the regulation of O2 uptake (3), also decreased significantly with age: aged and senescent rats showed decreases of biochemical mtNOS activity of 29–64% in whole brain, 30–48% in brain cortex, and 53–66% in the hippocampus. This marked inhibition of mtNOS activity of mtNOS upon aging was confirmed by the determination of mtNOS functional activity in the inhibition of mitochondrial respiration (46). The functional activity of mtNOS in hippocampal mitochondria or, in other words, the effect of the NO produced by mtNOS in the inhibition of respiration with malate-glutamate as substrate and expressed as nanogram-atom oxygen per minute per milligram protein, decreased in aged and senescent rats by 40 and 73% in whole brain, by 37 and 47% in brain cortex, and by 52 and 74%, respectively (Table 3). When succinate was used as substrate, the functional activity of mtNOS was slightly less affected if the specific activity (ng O·min−1·mg protein−1) is considered and to a similar extent if the percentage of activity (%) is considered. The differences in the respiration with both substrates are the respiration rates, ∼60% higher with succinate than with malate-glutamate, and the involvement of complex I in malate-glutamate oxidation.
A histochemical study involving complexes I and IV with image analysis was performed in the hippocampus of young, aged, and senescent rats (Fig. 2). The architectural patterns of neuronal layers and regional densities were maintained upon aging. Both methods showed a lower relative staining of the synaptic layers of Ammon's horn in old rats compared with young animals, although the difference was clearer in complex IV histochemistry (Fig. 2, insets 1, 2, and 3). At higher magnification, young rats (Fig. 2, inset 4) clearly show more complex IV staining than senescent rats (Fig. 2, inset 5) in the CA3 and CA1 stratum oriens and CA1 stratum radiatum. Young rats exhibit higher staining in the cortical layers, with intensely stained synaptic mitochondria of pyramidal cells compared with old and senescent rats. This was made clearer when the gray stain was taken to a color-coded scale (Fig. 2, insets 6 and 7) that shows more blue and green (higher relative staining) and less red (lower relative staining) in young rats than in aged rats. At even higher magnification, large mitochondria were observed in senescent rats, mainly in neuronal perykaria, main dendritic pyramidal shafts, and the stratum lacunosum moleculare of Ammon's horn (Fig. 2, inset 8). Surface plots of the staining densities (Fig. 2, insets 9 and 10) reveal minute synaptic mitochondria, ∼0.2 μm in diameter, in young animals and enlarged mitochondria, >1 μm in diameter, in senescent rats. The histochemical assay for complex I activity (Fig. 2, insets 10, 11, and 12) revealed a pattern that correlated well with complex IV but achieved less organelle resolution that precluded further single-organelle analysis.
The level of mitochondrial lipid and protein oxidation products was determined as the contents of TBARS and protein carbonyls; these contents were increased significantly in hippocampus, brain cortex, and whole brain of aged and senescent rats. The increases of both markers of oxidative damage in aged and senescent rats, compared with young animals, were, respectively, 37–62% in whole brain, 43–52% in brain cortex, and 66–74% in hippocampus (Table 4).
Electron transfer in rat hippocampus mitochondria was markedly decreased, in the range of 51–73% and depending on the measured activity, in senescent rats. This age-dependent respiratory inactivation was higher in hippocampus than in brain cortex and in whole brain and was observed as a diminished state 3 respiration with NAD-dependent substrates and as decreased enzymatic activities of complexes I and IV, in agreement with previous reports in mice and rat whole brain (32). Decreased mitochondrial state 3 respiration with NAD-dependent substrates in aging has been reported by three research groups: early by Vitorica et al. (48), and more recently by Navarro et al. (35) and by Cocco et al. (12).
In young animals, the rate of electron transfer by complex I (284 nmol e−·min−1·mg protein−1; Table 2) exceeds active mitochondrial O2 uptake (196 nmol e−·min−1·mg protein−1; Table 1) and is not rate limiting. In aged and senescent hippocampus, the rate of electron transfer by complex I (122 and 77 nmol e−·min−1·mg protein−1 for 12- and 20-mo-old rats, respectively; Table 2) becomes rate limiting for active state 3 mitochondrial respiration (134 and 96 nmol e−·min−1·mg protein−1; Table 1). It follows that impaired electron transfer in old animals impedes hippocampal physiological responses that require an enhanced energy supply. A decreased ATP production and an increased oxidative damage are the two main concepts of the mitochondrial hypothesis of aging (6, 32, 47).
In aging rodents and humans, hippocampus appears a target of oxidative damage secondary to sustained glucocorticoid responses in a cumulative and chronic manner. When the hippocampus is unable to maintain an adequate control on the hypothalamic-pituitary-adrenal (HPA) axis a secondary oxidative damage follows driven by the allostatic load of the adrenal response (29). Moreover, chronic stress increases the hippocampus vulnerability to neurotoxic challenges (15). The age-related dysfunction of glucocorticoid receptors in the HPA contributes to hippocampal aging (30), and, in turn, the deficit in ATP production in aged hippocampus explains the profound changes in learning and memory and the altered regulation of the stress response in aged mammals (29). The reduced antioxidant capacity in hypothalamus and adrenals upon aging agrees with the concept that mitochondrial oxidative dysfunction is a factor in the dysregulation of the HPA axis upon aging (42).
The histochemical assay of complex IV in hippocampal tissue indicates a detrimental effect of aging in this enzyme (Fig. 2) in agreement with the biochemical determinations of complex IV activity in isolated hippocampal mitochondria and with the histochemical determinations of cytochrome oxidase in rat hippocampal dentate gyrus (7), human substantia nigra (23), and monkey brain (31). However, it is worth noting that the elegant histochemical assays are far less sensitive than the biochemical determinations for assessing the impairment of enzyme activities, i.e., complex I activity is found markedly decreased by biochemical assays and not significantly modified by histochemistry.
Impaired electron transfer at complex I has been associated with an increased production of O2•− radicals (22, 43). The free radicals O2•− and NO, primarily produced at the mitochondrial membranes, initiate a series of second-order and nonenzymatic reactions leading to the formation of other free radicals (HO·, R·, ROO·, and NO2) and related nonradical but reactive species (H2O2, ONOO−, ROOH, and 1O2) that react with the biochemical components of the biological membranes and lead to oxidative damage, usually determined by the products of lipid peroxidation and of protein carbonylation.
The aging of rat hippocampus, cortex, and whole brain was associated with cumulative oxidative damage, as indicated by the mitochondrial content of TBARS and protein carbonyls. The effect was again more marked in aged and senescent hippocampus than in brain cortex and in whole brain. Also, immunohistochemical analysis had indicated a progressive oxidative damage in rat hippocampal nucleic acids (8-hydroxyguanosine and 8-hydroxy-2′-deoxyguanosine contents) in aging (6, 27). Moreover, electron microscopy of rat hippocampus showed that dietary supplementation with acetyl-l-carnitine and/or R-α-lipoic acid, two substances that prevent tissue oxidative damage, reversed the age-associated mitochondrial ultrastructural damage (27).
An inverse relationship was observed when comparing the content of oxidation products and the enzymatic activities in brain mitochondria (Fig. 3). This significant correlation indicates that rat hippocampus, cortex, and whole brain age with simultaneous oxidative damage and mitochondrial enzyme dysfunction.
A role of NO in mitochondrial biogenesis is emerging as a concept that has straightforward applications in cell turnover and proliferation and in the tissue atrophy associated with aging. The regulation of the cell cycle with duplication of the mitochondrial mass by mitochondrial biogenesis and binary division involves a complex regulatory system with >1,000 genes and 20% of cellular proteins involved (38). The hypothesis that applies to aging is that mitochondrial NO, produced by mtNOS and regulated by the mitochondrial metabolic state, diffuses to the cytosol and activates guanylate cyclase and cGMP production that in turn activates a series of transcription factors, such as peroxisome proliferator-activated receptor G coactivator 1α, nuclear respiratory factors (NRF-1 and NRF-2), and mitochondrial transcription factor A (25, 38–39). The marked decrease in mtNOS activity observed in hippocampal mitochondria in aged animals is interpreted as a decrease in the NO signaling from mitochondria to the cytosol that it slows down mitochondrial biogenesis and affects mitochondrial homeostasis and function.
A decrease in nNOS-RNA level has been described in the hippocampus of aged mice. nNOS-RNA is expressed by posttranslational and highly regulated steps as cytosolic, synaptosomal, and mtNOS (14).
The observation of increased oxidation products in brain mitochondria upon aging has puzzled researchers considering that mitochondrial turnover, usually ∼30 days, is much faster than cellular turnover, which is considered negligible in postmitotic neurons. The current hypothesis is that mitochondria accumulate oxidation products generated by free radical-mediated oxidations, as by-products of respiration as a linear function of time and respiration. A decreased mitochondrial biogenesis leads to extended life times of the existing mitochondria with cumulative increased contents of oxidation products. An about double content of oxidation products (Table 4) is interpreted as about two times slower mitochondrial turnover time.
Perspectives and Significance
The mitochondrial dysfunction observed in hippocampus provides the hypothesis of a subcellular mechanism for the behavioral, cognitive, and neurophysiological changes that accompany aging. The identification of the molecular mechanisms of hippocampus and brain aging should give the opportunity to develop nutritional and pharmacological treatments to retard the decline of brain function in aging and in related neurodegenerative diseases.
This work was supported by Grant FIS PI050636 from the Ministerio de Sanidad y Consumo de España, Fondo de Investigación Sanitaria, Instituto de Salud Carlos III and by Plan Andaluz de Investigación 2005 (CTS-194) of Spain.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society