Spreading depolarization-induced adenosine accumulation reflects metabolic status in vitro and in vivo

Abstract

Spreading depolarization (SD), a pathologic feature of migraine, stroke and traumatic brain injury, is a propagating depolarization of neurons and glia causing profound metabolic demand. Adenosine, the low-energy metabolite of ATP, has been shown to be elevated after SD in brain slices and under conditions likely to trigger SD in vivo. The relationship between metabolic status and adenosine accumulation after SD was tested here, in brain slices and in vivo. In brain slices, metabolic impairment (assessed by nicotinamide adenine dinucleotide (phosphate) autofluorescence and O2 availability) was associated with prolonged extracellular direct current (DC) shifts indicating delayed repolarization, and increased adenosine accumulation. In vivo, adenosine accumulation was observed after SD even in otherwise healthy mice. As in brain slices, in vivo adenosine accumulation correlated with DC shift duration and increased when DC shifts were prolonged by metabolic impairment (i.e., hypoglycemia or middle cerebral artery occlusion). A striking pattern of adenosine dynamics was observed during focal ischemic stroke, with nearly all the observed adenosine signals in the periinfarct region occurring in association with SDs. These findings suggest that adenosine accumulation could serve as a biomarker of SD incidence and severity, in a range of clinical conditions.

Figure 1

Brain slice thickness model for progressive metabolic impairment. Impaired metabolic status was confirmed by nicotinamide adenine dinucleotide (phosphate) (NAD(P)H) autofluorescence (360 nm excitation) and O₂ tension in brain slices acutely prepared at 250, 350, or 450 μm. (A) Representative brain slice showing relative position of O₂ probe and region of interest (ROI) for NAD(P)H assessment. Also visible are the KCl micropipette (right) and direct current (DC) recording electrode (left), used in subsequent experiments. (B and C) Increasing slice thickness progressively decreased pO₂, measured at the slice surface (n=5, 5, 4 slices from four animals) and increased autofluorescence attributable to NAD(P)H (n=10,11,11 slices from eight animals). Both parameters were measured before stimulation with SD. AFU, arbitrary fluorescence units. (D) NAD(P)H and O₂ were negatively correlated with each other (Spearman r=−0.60, P<0.05, n=14 slices from four animals). Bonferroni-corrected multiple comparisons: **P<0.01, ****P<0.0001. Error bars represent s.d.

Figure 2

Metabolic impairment prolonged direct current (DC) shifts and increased spreading depolarization (SD)-evoked adenosine accumulation. (A) Representative adenosine probe signals from hippocampal CA1 regions of 250, 350, and 450 μm slices after SD onset (arrow). (B) Population data (n=5 slices each, from four animals) from experiments shown in A, characterizing increases in adenosine concentration with progressive metabolic impairment. (C) Peak adenosine accumulation was positively correlated with basal metabolic status, as assessed by NAD(P)H autofluorescence (Pearson R²=0.3650, P<0.05, n=15 slices from four animals). (D and E) Representative and population data summarizing increases in DC shift duration with metabolic impairment (n=16, 19, 13 slices from eight animals). (F) Direct current duration was positively correlated with basal metabolic status as assessed by NAD(P)H autofluorescence (Spearman r=0.62, P<0.001, n=29 slices from eight animals). Bonferroni-corrected multiple comparisons: *P<0.05, ****P<0.0001. (G) Adenosine accumulation (estimated peak concentration) increased as DC shifts became longer, in brain slices subjected to SD. Pearson R²=0.7875, P=0.0008, n=14 slices from four animals. Error bars represent s.d.

Figure 3

Mitochondrial signals of oxidative metabolism correlated with adenosine accumulation. (A) Representative traces showing simultaneous nicotinamide adenine dinucleotide (phosphate) (NAD(P)H) and O₂ measurements during spreading depolarization (SD), expressed as change from baseline (dotted lines). The change in partial pressure of O₂ (pO₂) and the magnitude of the oxidative phase of NAD(P)H signals (initial negative deflections, top traces in 3A), were both markedly diminished as brain slice thickness increased. (B and C) Population data showing magnitudes of the initial oxidative phase of NAD(P)H transients and pO₂ changes from baseline as a function of slice thickness (n=11, 11, 8 from eight animals; n=5, 5, 4 from four animals). Absolute pO₂ minima were not significantly different between slices (analysis of variance P=0.16); thus, it is possible that oxygen consumption was limited by a floor effect. (D) Plot showing that peak adenosine accumulation after SD decreased as the amplitude of SD-induced NAD(P)H oxidative transients increased (Spearman r=−0.74, P<0.01, n=14 SDs, 1 SD per slice, from four animals). Data are from dual adenosine/NAD(P)H measurements made from the same preparations. Bonferroni-corrected multiple comparisons: **P<0.01, ***P<0.001, ****P<0.0001. Error bars represent s.d.

Figure 4

Graded metabolic impairment increased spreading depolarization (SD)-associated adenosine accumulation in brain slices of fixed thickness. (A) Representative recording from a 250 μm slice showing adenosine accumulation after SD in control conditions (95% O₂), then 21% O₂, and then during oxygen–glucose deprivation (OGD). Calibration steps show probe responses to 10 μmol/L and 20 μmol/L exogenous adenosine standards from the same recording, on an expanded time base. (B) Summary data from five preparations as shown in 4A (n=5 slices from three animals). Values are normalized to OGD-elicited adenosine accumulations in the same preparations. Bonferroni-corrected multiple comparisons: *P<0.05. Error bars represent s.d.

Figure 5

Detectable adenosine accumulation after spreading depolarization (SD) in vivo and enhancement under conditions of metabolic impairment. (A) Laser Speckle Contrast Imaging (LSCI) blood flow map depicting KCl application site (circle), adenosine probe (A) and direct correct (DC) recording electrode (Rec). A.U., arbitrary units. (B) Spreading depolarization-associated adenosine transients (measured as area under the curve (AUC)) increased with DC duration in untreated animals ( slope 1.47 mmol/L′s increase in adenosine AUC per 1 second longer DC duration, P<0.001, repeated measures analysis of variance, n=14 SDs from eight animals). (C) After injection of insulin (1.5 mU/g), animals became progressively hypoglycemic and DC shift durations increased (n=11, 12, 5 SDs from five animals). Bonferroni-corrected multiple comparisons: **P<0.01. (D and E) Simultaneous DC, electrocorticography (ECoG), and adenosine measurements from a single representative animal. (D) Control recordings before insulin exposure. Adenosine signals transiently increased after SD. Spreading depolarization was verified by DC shift, and suppression of ECoG activity. (E) Spreading depolarization induced in the same animal after establishment of hypoglycemia. Hypoglycemic SD was characterized by a prolonged DC shift and larger, longer lasting adenosine signal. Longer lasting suppression of ECoG was also observed under these conditions (see also Figure 7). Error bars represent s.d.

Figure 6

Adenosine accumulation in periinfarct tissue was linked to spreading depolarization (SD) events. (A) Laser Speckle Contrast Imaging (LSCI) blood flow map showing perfusion deficit in an animal with distal middle cerebral artery occlusion (dMCAO), before any periinfarct SDs. (B) Histogram of spontaneous SD incidence with respect to occlusion time (1-hour bins). Twenty-eight SDs were detected by propagating LSCI blood flow transients and/or DC shifts. Data shown are from the eight dMCAO animals with at least one SD. (C) Expanded DC shifts from the animal shown in D at time points indicated a and b. (D) Adenosine signals (bottom trace) increased after DC shifts (top trace) in each of five spontaneous SDs in an animal with dMCAO. At short inter-SD intervals, adenosine signals summated. Subtle electrocorticographic (ECoG) suppression could still be appreciated with each SD, even though ECoG power was diminished at baseline. Traces are from the same animal shown in A. SDs marked a and b are shown on an expanded time base in C above.

Figure 7

Electrocorticographic (ECoG) depression correlated with adenosine accumulation. Spreading depolarizations (SDs) from animals with insulin exposure (white squares) and without insulin (black circles), as well as SDs recorded during focal ischemia (individual SDs, black triangles; cluster, white triangle). Data shown are from non-isoelectric SDs only. Electrocorticographic suppression recorded during SDs correlated with adenosine accumulation (area under the curve (AUC)) (R²=0.5449, P<0.001). Each data set alone also showed this relationship (untreated: R²=0.9381, P<0.0001, n=10 SDs from six animals; insulin: R²=0.2440, P<0.05, n=18 SDs from three animals; evoked and spontaneous SDs in middle cerebral artery occlusion animals: (r=0.7167, P<0.05, n=9 SDs from six animals).