Subarachnoid hemorrhage mediates human neocortical network, membrane potential, and action potential bursting via glutamate receptors

Abstract

Subarachnoid hemorrhage (SAH) is a life-threatening neurological emergency with high mortality and morbidity rates. Despite its severity, the acute and direct neuronal effects of SAH remain poorly understood, particularly in human brain tissue. To address this gap, we applied cerebrospinal fluid from individuals with SAH (SAH-CSF) to human neocortical slices and examined neuronal activity and intrinsic properties. We found that SAH-CSF significantly increased population-wide neuronal activity, depolarized membrane potential, and either elevated firing rates or induced depolarization block of action potentials in these human neocortical slices. Notably, these neuronal changes were reversible upon washout of the SAH-CSF. Furthermore, kynurenic acid (KYNA), a glutamate receptor antagonist, effectively prevented SAH-CSF-induced neuronal changes. Together, these findings revealed key neuronal consequences of SAH-CSF in human brain tissue and suggest that glutamate receptor antagonists may offer therapeutic potential for SAH.

Fig. 1. SAH-CSF increases population-level neuronal activity in human neocortical slices.

a Schematic representation of human neocortex tissue resected during neurosurgical procedures for brain tumor removal (left), alongside a representative image of the resected neocortical slices (right). b Immunofluorescence staining of brain tumors and adjacent neocortical tissue using antibodies against astrocytes (GFAP antibody, red) and neurons (NeuN antibody, green). Nuclei were counterstained with DAPI. Scale bar = 50 μm. c Schematic illustrating two-photon imaging of human neocortical slices loaded with Cal-520. d Upper panels: Left, representative image of Cal-520-labeled cells in human neocortical slices under artificial cerebrospinal fluid (aCSF) conditions, captured via a two-photon microscope. Scale bar, 100 µm. Right, representative field of view highlighting regions of interest (ROIs) after cell extraction. Lower panel: Representative raw fluorescence traces from individual neurons under aCSF conditions. Scale bar: 60 sec, ΔF = 5. e Heatmap displaying neuronal activity of individual neurons from resected neocortical slices during sequential addition of different solutions. f Population-level neuronal activity plotted over time from human neocortical slices during continuous solution application. g Average neuronal activity within defined recording windows from resected neocortical slices. Data are presented as means ± s.d. or as box-plots showing the median and interquartile range (IQR), with the whiskers denoting the minimum and maximum values. Statistical analysis was performed using Two-way ANOVA followed by Scheffe post hoc comparison. *P  <  0.05, **P  <  0.01. N = 12 cells from one slice of a single subject. HC hydrocephalus, SAH subarachnoid hemorrhage, 3% SAH: 3 ul SAH-CSF + 97 ul aCSF; 5% HC: 5 ul HC-CSF + 95 ul aCSF.

Fig. 2. SAH-CSF increased population-wide neuronal activity in the neocortex of anesthetized mice.

a Schematic of experimental setup. Images for b injection setup and c external (left) and internal (right) views of the cranial window. d Average fluorescence ratios in recording window (∆F_{HC or SAH}/∆F_aCSF). Data are presented as box-plots showing the median and IQR, with the whiskers denoting the minimum and maximum values. Statistical analysis was performed using Wilcoxon matched-pairs signed rank test, ***P  <  0.001. N = 29 cells from three mouse brains (one male and two females). HC hydrocephalus, SAH subarachnoid hemorrhage. 0.5% SAH: 5 ul SAH-CSF + 995 ul aCSF; 0.5% HC: 5 ul HC-CSF + 995 ul aCSF.

Fig. 3. Neuronal and synaptic properties were reliably measured in excitatory and inhibitory neurons of human neocortical slices.

a, d, g Schematic illustration of whole-cell patch-clamp recordings performed on human neocortical slices. Representative images of excitatory neurons (b) and inhibitory neurons (c) from human neocortical slices. Excitatory and inhibitory neurons were distinguished via post hoc immunofluorescence staining using an anti-GAD67 antibody. Scale bar = 50 μm. Representative electrophysiological traces and corresponding quantification from excitatory (e) and inhibitory (f) neurons recorded in current-clamp mode. Scale bar = 50 μm. N = 9 excitatory neurons from nine slices of seven subjects; N = 25 inhibitory neurons from 25 slices of 13 subjects. Scale bar: 200 ms, 20 mV. Representative traces and quantitative data from excitatory (h) and inhibitory (i) neurons recorded in voltage-clamp mode. N = 9 excitatory neurons from nine slices of seven subjects; N = 9 inhibitory neurons from nine slices of nine subjects. Scale bar = 200 ms, 20 pA. Data are presented as means ± s.d. or as box-plots showing the median and IQR, with the whiskers indicating the minimum and maximum values. AP action potential, Rin input resistance.

Fig. 4. SAH-CSF depolarized membrane potentials and altered firing rates of action potentials in human neocortical slices.

a Schematic diagram outlining the experimental design. Two representative trace types (i, ii) were shown for four different conditions: aCSF (artificial cerebrospinal fluid), HC (hydrocephalus), SAH (subarachnoid hemorrhage), and washout with aCSF. Scale bar: 50 ms, 20 mV. b Schematic illustration of whole-cell patch-clamp recordings performed on human neocortical slices. c Membrane potential recordings from neurons under the four conditions. d Absolute values representing the difference in membrane potential between paired conditions. N = 9 cells (n = 5 excitatory neurons from five slices of five subjects and 4 inhibitory neurons from four slices of four subjects). Statistical analysis was performed using one-way repeated measures ANOVA followed by Dunn’s or Holm-Sidak’s post hoc comparison. **P  <  0.01, ***P < 0.001. Data are presented as means. Gray lines connect data from the same cells under different conditions. Vm: membrane potential.

Fig. 5. Glutamate receptor antagonists fully prevent SAH-CSF-induced phenotypes in human neocortical slices.

a Concentrations of potassium, sodium, chloride, and calcium in HC and SAH conditions. N = 4 for individuals with HC. N = 4 for individuals with SAH. Statistical analysis was performed using Mann–Whitney test, two-tailed. b Membrane potentials calculated using the Goldman–Hodgkin–Katz (GHK) equation for aCSF, HC, and SAH conditions. c Schematic illustration of experimental design. Representative traces (i, ii) were shown for four conditions: aCSF, HC, SAH, and washout with aCSF, with all conditions supplemented with glutamate receptor antagonists (kynurenic acid (KYNA), 5 mM). Scale bar: 50 ms, 20 mV. d Schematic of experimental setup of cell patching on human neocortical slices. e Membrane potential recordings from neurons under the four conditions. f Absolute values representing the difference in membrane potential between two assigned conditions. N = 6 cells (n = 1 excitatory neuron from one slice of a single subject and 5 inhibitory neurons from five slices of two subjects). Data are presented as means ± s.d. Data are presented as means or as box-plots showing the median and IQR, with the whiskers denoting the minimum and maximum values. Gray lines connect data from the same cells under different conditions. Vm membrane potential.