Reduction in Water and Metabolite Apparent Diffusion Coefficients During Energy Failure Involves Cation-Dependent Mechanisms. A Proton Nuclear Magnetic Resonance Study of Rat Cortical Brain Slices

Abstract

Proton (¹H) nuclear magnetic resonance (NMR) diffusion spectroscopy was used to assess apparent diffusion coefficients (ADCs) in rat brain slices. Aglycemic hypoxia caused reductions in the ADC of N-acetylaspartate (NAA) (0.15 to 0.09 × 10⁻³ mm²/s) and “slow” diffusion coefficient (D₂) of tissue water (0.51 to 0.37 × 10⁻³ mm²/s), together with a 32 ± 11% increase in tissue water volume, attributable to tissue swelling. The ADC and D₂ reductions were diminished, however, by removing external Ca²⁺, and under 10 mmol/L Mg²⁺, normoxic diffusion coefficients persisted until 40 minutes of hypoxia. The data suggest that the shift of water into the intracellular space alone cannot satisfactorily explain the reduced cerebral diffusion upon energy failure and that external Mg²⁺ and Ca²⁺ play crucial modulatory roles.

Keywords

Diffusion Water Brain Hypoxia Calcium Magnesium Macromolecules-Nuclear magnetic resonance spectroscopy

The apparent diffusion coefficient (ADC) of tissue water declines within a few minutes of vascular occlusion and energy failure, leading to greatly enhanced detection of afflicted tissue by diffusion-weighted magnetic resonance imaging (Moseley et al., 1990). The method has consequently become particularly effective at identifying acute stroke victims (Baird and Warach, 1998). Paradoxically, the mechanisms affecting in vivo ADC are numerous, and their individual contributions to the overall diffusion change have proven to be difficult to quantify. The key hypothesis involves a shift of ions and water from the extracellular space into an inherently more restricted intracellular space in energetically compromised tissue, substantiated by the rapid perturbations known to occur in cerebral ion and water homeostasis (Hossmann, 1994) and the reduced extracellular volume fraction (Hansen, 1985).

Proton (¹H) nuclear magnetic resonance (NMR) diffusion spectroscopy has been recently used to study the effects of ischemia on the ADCs of cerebral metabolites such as choline-containing compounds, creatine and phosphocreatine, and N-acetylaspartate (NAA) in ischemic stroke models in vivo (Wick et al., 1995; van der Toorn et al., 1996). The results show that the intracellular metabolites experience ADC reductions akin to water. The nature of this reduction has not been experimentally explained, however. In this study, we have examined the ADCs of an intracellular neuronal metabolite, NAA, and cytoplasmic proteins, as they probe an intracellular microenvironment, together with determining the ADC and apparent fractional distribution of water in the cortical brain slice preparation. We have also evaluated the dependency of these effects on external Ca²⁺ and Mg²⁺ to gain more insight into the underlying mechanisms of cerebral ADC reduction.

MATERIALS AND METHODS

Animal experiments were approved by the Ethical Committee of the National Laboratory Animal Center, University of Kuopio. Transverse cortical brain slices (thickness 0.35 mm) from adult male Wistar rats (200 to 250 g) were prepared using a McIlwain tissue chopper and superfused in a 15-mm NMR tube in Krebs-Henseleit buffer (KHB) containing 124 mmol/L NaCl, 5 mmol/L KCl, 1.2 mmol/L MgSO₄, 1.2 mmol/L CaCl₂, 26 mmol/L NaHCO₃, 1.2 mmol/L KH₂PO₄, and 10 mmol/L D-glucose (324.8 mOsmol/L) equilibrated with O₂/CO₂ (95%/5%) at 37°C. During periods of aglycemic hypoxia, the slices were superfused in KHB with glucose omitted in otherwise normal KHB, in KHB with calcium omitted (0 mmol/L Ca²⁺), or in KHB containing 10 mmol/L Mg²⁺(Brooks and Kauppinen, 1993). Hypoxias lasting 10, 20, or 40 minutes were introduced by bubbling the medium with N₂ (95%)/CO₂ (5%). Slices were exposed to KHB containing 15 mmol/L KCl (high K⁺) for 20 minutes to assess ADC and volume changes. These conditions are known to depolarize cells without compromising their energy state (Lukkarinen et al., 1997). Slices were also superfused in KHB containing 248 mmol/L NaCl (572.8 mOsmol/L).

Proton (¹H) NMR spectra were recorded using a Varian 16-mm NMR probe without the original aluminum cover (Varian Associates, Palo Alto, CA, U.S.A.) to reduce eddy currents imposed by the presence of metal in a 400-MHz vertical magnet interfaced to an SMIS microimaging and spectroscopy console (Surrey Medical Imaging Systems, Surrey, U.K.). Gradient system (Magnex Scientific, Abdingdon, U.K.; maximal amplitude 30 G/cm) performance was assessed by diffusion measurements on phantoms containing a 0.9% (wt/vol) NaCl solution of NAA (20 mmol/L) at 22°C, as described earlier (Hakumäki et al., 1998). Diffusion-weighted metabolite spectra were acquired using a binomial spin echo sequence, with excitation set at 2.0 ppm, and a pair of diffusion gradients along the Z-axis using four serial b values between 0 and 6,600 s/mm² [gradient duration (δ) = 5.5 ms, gradient spacing (Δ) = 60 ms, dummy scans = 2, repetition time = 2.5 s, echo time = 120 ms]. Water spectra were obtained also in spectroscopy mode, with maximal excitation set on water, and with 16 serial b values between 0 and 4,600 s/mm² (δ = 6.5 ms, Δ = 30 ms, dummy scans = 2, repetition time = 1.5 s, echo time = 60 ms). Choice of spectroscopy instead of imaging mode is substantiated by several factors such as improved temporal resolution without compromising signal-to-noise ratio (Busza et al., 1992; Davis et al., 1994) and the possibility of obtaining high b values (Niendorf et al., 1996).

Macromolecular diffusion was quantified using a stimulated echo sequence (1331 -t-90°-t-90°-t-acquire) with four serial b values between 0 and 17,200 s/mm² (δ = 6 ms, Δ = 80 ms, dummy scans = 2, repetition time = 3.5 s, echo time = 20 ms) and maximal excitation set at 1.3 ppm. Buffer flow and rotor motion were stopped, and the slice preparations retained their stand-still position in the sensitive volume of the coil for the duration of the ADC measurements (1 minute for water, 5 minutes for metabolites and macromolecules). The temperature was maintained at 37.0 ± 0.5°C using a variable-temperature unit. Metabolite and macromolecule spectra were deconvolved using a moderate sine bell function over the free induction decay and by exponential multiplication corresponding to line broadening of 6 Hz. Spectral resonance areas were obtained as integrals using MR console software (Specana; SMIS). All values shown are mean ± SD.

RESULTS

The ADCs of NAA and water in 0.9% (wt/vol) NaCl solution at 22°C gave values of 0.52 ± 0.03 and 2.1 ± 0.02 × 10⁻³ mm²/s (n = 3), respectively. Biexponential decay was observed for water ADC in the brain slice preparation (Fig. 1A) analogously to that reported for in vitro cell studies containing a large extracellular fraction in the NMR cuvette (van Zijl et al., 1994). The data were best fitted according to a superposition of two exponential decays (R² > 0.98):

\frac{S}{S_{0}} = f_{1} e^{{- b D}_{1}} + f_{2} e^{{- b D}_{2}}

(1)

where S and S₀ are the resonance signal intensities with and without diffusion weighting, 1 and 2 denote the water population fractions (f) and diffusion coefficients (D), and b denotes the diffusion-weighting factor.

FIG. 1.

(A) Biexponential water signal decays as a function of b, observed during normoxia and aglycemic hypoxia. Dotted line denotes fit for f₁ water component and dashed lines for the respective f₂ water components in normoxic (filled squares) and hypoxic (open squares) conditions. (B) Typical diffusion-weighted ¹H nuclear magnetic resonance spectra from hypoxic cortical tissue with corresponding b values (s/mm²). Maximal binomial spectral excitations are centered at 2.0 ppm on the N-acetylaspartate resonance. Acquisition parameters are as described in Materials and Methods. Glx, cerebral glutamate + glutamine.

In normoxic brain slices, D₁ and D₂ were 3.40 ± 0.20 and 0.51 ± 0.05 × 10⁻³ mm²/s, for the fast f₁ and slow f₂ decay fractions, respectively, with the f₂ component accounting for 17 ± 3% of the signal. The actual physical volume fraction of tissue amounted to 17 ± 2% of NMR tube volume (n = 8). The diffusion-weighted NAA spectra (Fig. 1B) decayed monoexponentially (R² > 0.95), similarly to the macromolecular methyl resonance at 0.9 ppm (not shown).

Elevated K⁺ leads to membrane depolarization and cell swelling mimicking of cerebral energy failure. However, when superfused with 15 mmol/L KCl for 20 minutes under these conditions, the phosphocreatine/ATP ratios are only slightly reduced (Lukkarinen et al., 1997) compared with the collapse observed in aglycemic hypoxia (Brooks and Kauppinen, 1993). Exposure to elevated KCl led to a 44% increase in f₂, accompanied by D₂ at 0.50 ± 0.03 × 10⁻³ mm²/s, and the ADC of NAA remained at 0.16 ± 0.02 × 10⁻³ mm²/s (Fig. 2A).

FIG. 2.

Proton nuclear magnetic resonance diffusion spectroscopy data from cortical slice preparation during depolarization with high (15 mmol/L) KCl and subsequent recovery (A) and high NaCl with subsequent hypoxia in Krebs-Henseleit buffer with 248 mmol/L NaCl (B). The apparent diffusion coefficient values and tissue water fractions were obtained as described in Materials and Methods. Data are mean ± SD for four independent experiments. *P < 0.05, **P < 0.01 compared with control, Student's paired t-test.

Aglycemic hypoxia in the presence of physiological KHB (Fig. 3A) led to a significant (29%) increase in f₂ by 20 minutes. This was concurrent with a reduction in D₂ to 0.43 ± 0.04 × 10⁻³ mm²/s, but the D₁ did not change. By 40 minutes of hypoxia, f₂ was elevated to 25 ± 2% and D₂ reduced to 0.37 ± 0.04 × 10⁻³ mm²/s. N-Acetylaspartate is expressed exclusively in adult neurons, with a predominantly cytoplasmic location (Moffett et al., 1991), thus probing the intracellular milieu. The average ADC for NAA in normoxia was 0.15 ± 0.01 × 10⁻³ mm²/s, being significantly lower by 20 and 40 minutes of hypoxia (0.10 ± 0.02 and 0.09 ± 0.03 × 10⁻³ mm²/s, respectively), consistent with reduced diffusivity during energy failure in the cytoplasm. The ADC of the mobile protein resonance at 0.9 ppm (Kauppinen et al., 1992) was also reduced from 0.03 ± 0.01 × 10⁻³ mm²/s in normoxia to 0.02 ± 0.01 × 10⁻³ mm²/s after 40 minutes of hypoxia (P = 0.07, n = 5, Student's paired t test).

FIG. 3.

Proton nuclear magnetic resonance diffusion spectroscopy data from cortical slice preparation during aglycemic hypoxia in plain Krebs-Henseleit buffer (A), with no external calcium (B), and with high external magnesium (C) present. Apparent diffusion coefficient values and tissue water fractions were obtained as described in Materials and Methods. REC, 60-minute recovery in oxygenated Krebs-Henseleit buffer after 20 minutes of hypoxia. Data are mean ± SD for four to six independent experiments. *P < 0.05, **P < 0.01 compared with control, Student's paired t-test.

Aglycemic hypoxia leads to complete energy failure with similar kinetics in KHB with 0 Ca²⁺ or 10 mmol/L Mg²⁺, yet the metabolic recovery of cerebral tissue after reoxygenation is greatly improved relative to plain KHB (Brooks and Kauppinen, 1993). When Ca²⁺ was omitted from the buffer during hypoxia, a similar increase in f₂ was observed, but reductions in diffusion coefficients of both water and NAA were attenuated (Fig. 3B). However, when hypoxic slices were superfused with plain KHB containing 10 mmol/L external Mg²⁺, the ADC of NAA and water D₂ remained at control levels until 40 minutes of hypoxia, despite the large increase in f₂, as observed in hypoxia in plain KHB (Fig. 3C). There were no differences in ADCs after recovery from 20 minutes of aglycemic hypoxia, except for NAA after hypoxia induced in 10 mmol/L Mg²⁺ KHB.

As an increase in f₂ was detected in hypoxic tissue independent of external Ca²⁺ and Mg²⁺ status, we tested the effects of high NaCl, known to induce cell shrinkage in an ex vivo brain preparation due to an osmotic effect (Krizaj et al., 1996). In normoxia, KHB supplemented with additional 124 mmol/L NaCl resulted in a severely reduced f₂, with no significant effect on D₂ of water and ADC of NAA. Hypoxia under the same conditions caused a large increase in f₂, with subsequent reductions in the diffusivity of water and NAA (Fig. 2B).

The apparent intracellular tortuosity values (λ_app) were calculated for normoxic and hypoxic cortical brain tissue with the approach of van der Toorn and co-workers (1996) using ADC data for NAA in vivo and in vitro. For macromolecules, a reference in vitro value of D ∼0.12 × 10⁻³ mm²/s, as observed for polypeptides such as bovine pancreatic trypsin inhibitor or ubiquitin, was used (Iiyina et al., 1997). Interestingly, the λ_app values using the water fraction f₂ were very similar. In hypoxia with external Ca²⁺ omitted, similar reductions were observed (Table 1).

TABLE 1.

Apparent intracellular tortuosity values (λ_app) from cortical slice preparate diffusion data

	Normoxia	Hypoxia (20 min)	Hypoxia (40 min)	Recovery (60 min)
Plain KHB
Water (f₂)	2.08 ± 0.13	2.21 ± 0.08†	2.38 ± 0.12†	2.10 ± 0.05
NAA	1.88 ± 0.12	2.28 ± 0.19*	2.33 ± 0.30†	2.06 ± 0.15
MM	2.09 ± 0.20	2.36 ± 0.20	—	—
KHB (Ca²⁺, 0 mmol/L)
Water (f₂)	2.11 ± 0.03	2.19 ± 0.03*	2.18 ± 0.06†	2.17 ± 0.10
NAA	1.88 ± 0.14	2.05 ± 0.16†	2.30 ± 0.19†	2.00 ± 0.23
KHB (Mg²⁺, 10 mmol/L)
Water (f₂)	2.12 ± 0.05	2.13 ± 0.09	2.22 ± 0.09†	2.13 ± 0.08
NAA	1.90 ± 0.13	1.96 ± 0.15	2.18 ± 0.25†	2.00 ± 0.22

P < 0.01.

†

P < 0.05, Student's paired t-test relative to normoxia. Recovery is after corresponding 20 minute hypoxia. KHB, Krebs–Henseleit buffer; MM, macromolecules.

DISCUSSION

The present data show that the ADC of NAA responds to external cation manipulations during energy failure. The ADC reduction can be attenuated by omitting external Ca²⁺ and, more extensively, delayed by the presence of 10 mmol/L external Mg²⁺. The latter effect is particularly interesting because it demonstrates that diffusion changes can be dissociated from volume changes. The fact that high external NaCl cannot prevent intracellular ADC change, unlike high Mg²⁺, strongly argues for a contributing cytoplasmic process behind the altered diffusivity upon hypoxic energy failure. Mg²⁺ is a natural, nonspecific Ca²⁺ blocker, affecting several cation channels (Iseri and French, 1984; Mayer et al., 1984), and its intracellular concentration follows the extracellular one (Garcia-Martin et al., 1995). Energy failure per se, however, does not cause an increase in the intracellular concentration of Mg²⁺(Heinonen and Åkerman, 1987).

The reduction in water diffusion upon ischemic energy failure has been interpreted as a shift of more freely diffusing extracellular water into the more restricted, possibly more viscous intracellular space after loss of membrane potential, reducing the extracellular-to-intracellular volume ratio (Hansen, 1985) and resulting in increased extracellular restriction (Szafer et al., 1995). This theory has faced a challenge from recent ischemia studies demonstrating concomitant reductions also in the ADCs of intracellular metabolites (Wick et al., 1995; van der Toorn et al., 1996; Duong et al., 1998; Dijkuizen et al., 1999) and ions (Neil et al., 1996)in vivo. In fact, a recent NMR study using phosphorus metabolites as compartment-specific probes demonstrated similar reductions in the ADCs of both intra- and extracellular spaces during energy failure (Duong et al., 1998). This observation together with our data imply that the dominant restriction to overall diffusion must take place intracellularly. The metabolite data also imply that choline-containing compounds, creatine plus phosphocreatine, NAA, and inositol (Wick et al., 1995) are nearly uniformly affected.

There are several factors in the experimental design that need to be discussed when interpreting the present water diffusion coefficients. We use b values up to 4,600 s/mm² in quantifying water diffusion coefficients, whereas the majority of diffusion magnetic resonance imaging studies are carried out with b values up to 1,500 s/mm². The NMR data acquisition conditions inevitably weight the data for given water populations within the biological sample (van Zijl et al., 1994; Niendorf et al., 1996). In the brain slice preparation, it is necessary to use high b values to suppress the signal from the buffer, and the fast water component attains zero signal intensity at b of ∼2,000 s/mm² (Fig. 1). It should be kept in mind that in vivo, ∼80% of brain water is intracellular, and thus intracellular water is likely to play a dominant role in diffusion images under commonly used magnetic resonance imaging acquisition conditions with typical b values in the order of 1,500 s/mm². The assignment of the two water fractions in the brain slice preparation to physical equivalents is not difficult, but data interpretation at the level of specific subcellular compartments may not be as straightforward as for NAA, however. We assign f₁ to the buffer because its ADC matches that of free water at 37°C and the NMR signal from f₁ is completely suppressed with similar b values as the extracellular water in an in vitro cell preparation (van Zijl et al., 1994). The f₂ and D₂ represent “tissue water” and, indeed, a D₂ of 0.51 × 10⁻³ mm²/s determined in normoxic slices is in a good agreement with the average water ADC determined in an immobilized hippocampal brain slice using magnetic resonance imaging at similar b values (Buckley et al., 1999). The f₂ gains contributions from various water tissue compartments, and furthermore, the exchange between and within different tissue compartments at Δ = 30 ms is likely to bias assignment of f₂(Pfeuffer et al., 1998). Interestingly, diffusion data of NAA and water f₂ are in apparent agreement under all conditions, which appears to argue for a minor contribution for the expected exchange weighting both between extra- and intracellular space and also intracellularly to the values of f₂ and D₂. We therefore use f₂ as an apparent indicator of tissue water, bearing in mind the uncertainty associated with exchange processes between intra- and extracellular spaces within tissue (Pfeuffer et al., 1998). It should be pointed out that the present diffusion data should not be affected by expected T₁ and T₂ relaxation changes, as dummy scans were used to achieve magnetic equilibrium acquisition and the data were fitted into exact equations.

An important technical issue is that the diffusion measurements were carried out with slices in a stand-still situation, which will inevitably lead to anoxia over time. However, the dissolved oxygen in KHB bubbled with O₂/CO₂ is sufficient to maintain tissue energetics for several minutes after flow is halted, whereas the residual oxygen in a hypoxic preparation is exhausted in 2 minutes (Kauppinen and Williams, 1990). The ADC of NAA determined in our preparation is essentially identical to that from previous in vivo reports (Wick et al., 1995; van der Toorn et al., 1996), showing that the normoxic cortical slices are under energetically physiological conditions for the duration of the measurement. Furthermore, the proportional ADC reductions of NAA and water upon induction of aglycemic hypoxia agree with in vivo ischemia studies (Davis et al., 1994; Wick et al., 1995; van der Toorn et al., 1996).

Brain slices are a well-documented model for cortical brain in metabolic studies by NMR spectroscopy, maintaining similar phosphocreatine/ATP ratio as brain in vivo and low lactate concentration (Bachelard and Badar-Goffer, 1993; Kauppinen and Williams, 1990). Brain slices actively metabolize [¹³C]glucose (Lukkarinen et al., 1997), similar to the brain in vivo(Fitzpatrick et al., 1990). Accepting the drawback of nonphysiological extracellular-to-intracellular water volume ratio as discussed above and the possibility that cell swelling might be greater ex vivo than in vivo, the brain slice preparation offers a wider attainable range of experimental conditions than the brain in vivo. The ADC data of NAA from the slice preparation both under controlled and compromised energetic conditions are consistent with those obtained in in vivo conditions (Wick et al., 1995; van der Toorn et al., 1996; Dijkhuizen et al., 1999), thus validating the use of the brain slice preparation in studying the mechanistic aspects of molecular diffusion.

The present data show that the early diffusion drop in brain cortex in vitro upon hypoxia is expressed only if (1) ATP is exhausted, (2) external Ca²⁺ is present, and (3) external Mg²⁺ is present at a physiological level. Membrane depolarization without energy failure and acute cell swelling in the absence of ATP with high external Mg²⁺ apparently are not sufficiently strong perturbations of tissue homeostasis to influence ADC within the first 20 minutes of hypoxia, when intracellular sodium concentrations are uninfluenced (Brooks et al., 1993). The observation with 15 mmol/L KCl is in line with a previous hippocampal brain slice study showing no diffusion change in the two cerebral diffusion coefficients upon ouabain poisoning but an increase in the fraction of slowly diffusing water, possibly reflecting cell swelling (Buckley et al., 1999). The cation dependency of ADC changes in energy failure suggests biochemically driven mechanisms behind this phenomenon. These can be brought forth by increased intracellular tortuosity and viscosity due to structural reorganization such as disassembly of cytoplasmic proteins or of the cytoskeleton (Pettigrew et al., 1996) or reduced microcirculation (Duong et al., 1998). Interestingly, ADC reductions of similar magnitude are observed in cortical spreading depression induced by applying concentrated KCl topically in vivo(Gardner-Medwin et al., 1994). Although our model with 15 mmol/L external KCl mimics the situation, it must be stressed that spreading depression induces a severe 40 to 45% maximal reduction in cortical ATP and a 75% reduction in phosphocreatine (Gault et al., 1994), which might facilitate an adverse, albeit temporary, structural aberration within the cell interior.

Footnotes

Abbreviations used

Acknowledgements

The authors thank Dr. Kai Kaila for comments on the manuscript.

References

Bachelard

Badar-Goffer

(1993) NMR spectroscopy in neurochemistry. J Neurochem 61:412–429

Baird

Warach

(1998) Magnetic response imaging of acute stroke. J Cereb Blood Flow Metab 18:583–609

Brooks

Kauppinen

(1993) Calcium-mediated damage following hypoxia in cerebral cortex ex vivo studied by NMR spectroscopy. Evidence for direct involvement of voltage-gated Ca ²⁺-channels. Neurochem Int 23:441–450

Brooks

Pirttilä

T-RM

Kauppinen

(1993) Delayed increase in intracellular Na⁺ in cerebral cortical slices during severe hypoxia as measured by douple quantum filtered ²³Na⁺ NMR. Neuroreport 4:139–142

Buckley

Bui

Phillips

Zelles

Inglis

Plant

Blackband

(1999) The effect of ouabain on water diffusion in the rat hippocampal slice measured by high resolution NMR imaging. Magn Res Med 41:137–142

Busza

Allen

King

van Bruggen

Williams

Gadian

(1992) Diffusion-weighted imaging studies of cerebral ischemia in gerbils. Potential evidence to energy failure. Stroke 23:1601–1612

Davis

Ulatowski

Eleff

Izuta

Mori

Shunagu

van Aijl

(1004) Rapid monitoring of changes in water diffusion coefficients during reversible ischemia in cat and rat brain. Magn Res Med 31:454–460

Dijkhuizen

de Graaf

Tulleken

Nicolay

(1999) Changes in the diffusion of water and intracellular metabolites after excitotoxic injury and global ischemia in neonatal rat brain. J Cereb Blood Flow Metab 19:341–349

Duong

Ackerman

JJH

Ying

Neil

(1998) Evaluation of extra- and intracellular apparent diffusion in normal and globally ischemic rat brain via ¹⁹F NMR. Magn Res Med 40:1–13

10.

Fitzpatrick

Hetherington

Behar

Shulman

(1990) The flux from the glucose to glutamate in the rat brain in vivo as determined by ¹H-observed, ¹³C-edited NMR spectroscopy. J Cereb Blood Flow Metab 10:170–179

11.

Garcia-Martin

Martin-Romero

Guiterrez-Merino

(1995) Intra-synaptosomal free Mg²⁺ concentration measured with the fluorescent indicator mag-fura-2: modulation by Na⁺ gradient and by extrasynaptosomal ATP. J Neurochem 65:2757–2764

12.

Gardner-Medwin

van Bruggen

Williams

Ahier

(1994) Magnetic resonance imaging of propagating waves of spreading depression in anaesthetised rat. J Cereb Blood Flow Metab 14:7–11

13.

Gault

Lin

C-W

LaManna

Lust

(1994) Changes in energy metabolites, cGMP and intracellular pH during cortical spreading depression. Brain Res 641:176–180

14.

Hakumäki

Poptani

Puumalainen

A-M

Loimas

Paljarvi

Yla-Herttuala

Kauppinen

(1998) Quantitative ¹H NMR diffusion spectroscopy of BT4C rat glioma during thymidine kinase-mediated gene therapy in vivo: identification of apoptotic response. Cancer Res 58:3791–3799

15.

Hansen

(1985) Effect of anoxia on ion distribution in the brain. Physiol Rev 65:101–148

16.

Heinonen

Åkerman

(1987) Intracellular free magnesium in synaptosomes measured with entrapped eriochrome blue. Biochim Biophys Acta 898:331–337

17.

Hossmann

K-A

(1994) Glutamate-mediated injury in focal cerebral ischemia—the excitotoxic hypothesis revised. Brain Pathol 4:23–36

18.

Iiyina

Roongta

Pan

Woodward

Mayo

(1997) A pulsedfield gradient NMR study of bovine pancreatic trypsin inhibitor self-association. Biochemistry 36:3383–3388

19.

Iseri

French

(9184) Magnesium: nature's physiologic calcium blocker. Am Heart J 108:188–193

20.

Kauppinen

Williams

(1990) Cerebral energy metabolism and intracellular pH during severe hypoxia and recovery. A study using ¹H, ³¹P and ¹H{¹³C} nuclear magnetic resonance spectroscopy in the guinea pig cerebral cortex in vitro. J Neurosci Res 26:359–369

21.

Kauppinen

Nissinen

Kärkkäinen

A-M

Pirtila

Palvimo

Kokko

Williams

(1992) Detection of thymosin β4 in situ in a guinea pig cerebral cortex preparation using ¹H NMR spectroscopy. J Biol Chem 267:9905–9910

22.

Krizaj

Rice

Wardle

Nicholson

(1996) Water compart-mentalization and extracellular tortuosity after osmotic changes in cerebellum of Trachemys scripta. J Physiol (Lond) 492:887–896

23.

Lukkarinen

Oja

JME

Turunen

Kauppinen

(1997) Quantitative determination of glutamate turnover by ¹H-observed, ¹³C-edited nuclear magnetic resonance spectroscopy in the cerebral cortex ex vivo: interrelationships with oxygen consumption. Neurochem Int 31:95–104

24.

Mayer

Westbrook

Guthrie

(1984) Voltage-dependent block by Mg²⁺ of NMDA responses in spinal cord neurones. Nature 309:261–263

25.

Moffett

Mamboodiri

MAA

Cangro

Neale

(1991) Immunohistochemical localization of N-acetylaspartate in rat brain. Neuroreport 2:131–134

26.

Moseley

Cohen

Mintorovitch

Chileuitt

(1990) Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Res Med 14:330–346

27.

Neil

Duong

Ackerman

JJH

(1996) Evaluation of intracellular diffusion in normal and globally ischemic rat brain via ¹³³Cs NMR. Magn Res Med 35:329–335

28.

Niendorf

Dijkhuizen

Norris

van Lookeren Campagne

Nicolay

(1996) Biexponential diffusion attenuation in various states of brain tissue: implications for diffusion-weighted imaging. Magn Res Med 36:847–857

29.

Pettigrew

Holtz

Craddock

Minger

Hall

Geddes

(1996) Microtubular proteolysis in focal cerebral ischemia. J Cereb Blood Flow Metab 16:1189–1202

30.

Pfeuffer

Dreher

Sykova

Leibfritz

(1998) Water signal attenuation in diffusion-weighted ¹H NMR experiments during cerebral ischemia: influence of intracellular restrictions, extracellular tortuosity, and exchange. Magn Res Imag 16:1023–1032

31.

Szafer

Zhong

Gore

(1995) Theoretical model for water diffusion in tissues. Magn Res Med 33:697–712

32.

van der Toorn

Dijkhuizen

Tulleken

Nicolay

(1996) Diffusion of metabolites in normal and ischemic rat brain measured by localized ¹H MRS. Magn Res Med 36:914–922

33.

van Ziji

PCM

Davis

Moonen

CTW

(1994) Diffusion spectroscopy in living systems. In: NMR in Physiology and Biomedicine ( Gillies

, ed), San Diego: Academic Press, pp. 185–198

34.

Wick

Nagatomo

Prielmeier

Frahm

(1005) Alteration of intracellular metabolite diffusion in rat brain in vivo during ischemia and reperfusion. Stroke 26:1930–1933