Inhaled Carbon Dioxide Improves Neurological Outcomes by Downregulating Hippocampal Autophagy and Apoptosis in an Asphyxia-Induced Cardiac Arrest and Resuscitation Rat Model

Abstract

Background Protracted cerebral hypoperfusion following cardiac arrest (CA) may cause poor neurological recovery. We hypothesized that inhaled carbon dioxide (CO₂) could augment cerebral blood flow (CBF) and improve post-CA neurological outcomes. Methods and Results After 6-minute asphyxia-induced CA and resuscitation, Wistar rats were randomly allocated to 4 groups (n=25/group) and administered with different inhaled CO₂ concentrations, including control (0% CO₂), 4% CO₂, 8% CO₂, and 12% CO₂. Invasive monitoring was maintained for 120 minutes, and neurological outcomes were evaluated with neurological function score at 24 hours post-CA. After the 120-minute experiment, CBF was 242.3% (median; interquartile range, 221.1%-267.4%) of baseline in the 12% CO₂ group while CBF fell to 45.8% (interquartile range, 41.2%-58.1%) of baseline in the control group (P<0.001). CBF increased along with increasing inhaled CO₂ concentrations with significant linear trends (P<0.001). At 24 hours post-CA, compared with the control group (neurological function score, 9 [interquartile range, 8-9]), neurological recovery was significantly better in the 12% CO₂ group (neurological function score, 10 [interquartile range, 9.8-10]) (P<0.001) while no survival difference was observed. Brain tissue malondialdehyde (P=0.02) and serum neuron-specific enolase (P=0.002) and S100β levels (P=0.002) were significantly lower in the 12% CO₂ group. TUNEL (terminal deoxynucleotidyl transferase-mediated biotin-deoxyuridine triphosphate nick-end labeling)-positive cell densities in hippocampal CA1 (P<0.001) and CA3 (P<0.001) regions were also significantly reduced in the 12% CO₂ group. Western blotting showed that beclin-1 (P=0.02), p62 (P=0.02), and LAMP2 (lysosome-associated membrane protein 2) (P=0.01) expression levels, and the LC3B-II:LC3B-I ratio (P=0.02) were significantly lower in the 12% CO₂ group. Conclusions Administering inhaled CO₂ augmented post-CA CBF, mitigated oxidative brain injuries, ameliorated neuronal injury, and downregulated apoptosis and autophagy, thereby improving neurological outcomes.

Keywords: apoptosis; autophagy; carbon dioxide; cardiac arrest; cerebral blood flow.

Figure 1. Study procedures and measurements at baseline, asphyxia, cardiac arrest and cardiopulmonary resuscitation, and reperfusion.

ABG indicates arterial blood gas; CA, cardiac arrest; CBF, cerebral blood flow; EtCO₂, partial pressure of end‐tidal carbon dioxide; HR, heart rate; MAP, mean arterial pressure; and ROSC, return of spontaneous circulation.

Figure 2. Physiological parameters during invasive monitoring.

The administration of inhaled CO₂ increased cerebral blood flow, end‐tidal CO₂, partial pressure of brain tissue oxygenation without compromising mean arterial pressure. Data are presented as the median and interquartile range. Pairwise comparisons of physiological parameters at 120 minutes post‐return of spontaneous circulation between experimental and control groups were performed by the post hoc Dunn test. The asterisk indicates statistical significance. CA indicates cardiac arrest; CBF, cerebral blood flow; ΔPbtO₂, delta of partial pressure of brain tissue oxygenation; EtCO₂, partial pressure of end‐tidal carbon dioxide; iCO₂, inhaled carbon dioxide; MAP, mean arterial pressure; and ROSC, return of spontaneous circulation.

Figure 3. Neurological outcomes and brain injury biomarkers at 24 hours post‐return of spontaneous circulation.

Compared with the control group, the neurological function score was higher in the 12% CO₂ group, suggesting better neurological recovery. The oxidative injuries, as indicated by the brain tissue malondialdehyde level, were lower in the 12% CO₂ group. Also, the neuronal injuries, as represented by the serum neuron‐specific enolase and S100β concentrations, were lower in the 12% CO₂ group. Data are presented as the median and third quartile. Pairwise comparisons of neurological outcomes or biomarker levels between experimental and control groups were performed using the post hoc Dunn test. Samples were randomly selected for measuring malondialdehyde, neuron‐specific enolase, and S100β levels. The asterisk indicates statistical significance. NSE indicates neuron‐specific enolase.

Figure 4. TUNEL staining in hippocampus.

The lower hippocampal TUNEL‐positive cell densities suggest reduced apoptotic cells in the 12% CO₂ group than the control group. Data are presented as the median and third quartile. Pairwise comparisons between experimental and control groups were performed by the post hoc Dunn test. Samples were randomly selected for TUNEL stain. A, Brain sections were stained for TUNEL (green) and 4’,6‐diamidino‐2‐phenylindole (bluish‐violet). Confocal microscopy was used to image hippocampal CA1 and CA3 regions at 50× magnification. Red rectangles indicate selected CA1 and CA3 regions for quantification. The scale bar represents 200 μm. B and C, Apoptotic cell densities in CA1 and CA3 regions. The asterisk indicates statistical significance. DAPI indicates 4’,6‐diamidino‐2‐phenylindole; and TUNEL, terminal deoxynucleotidyl transferase–mediated biotin–deoxyuridine triphosphate nick‐end labeling.

Figure 5. Western blotting of brain tissues.

Reactive oxygen species may activate the mitochondrial pathway leading to execution of apoptosis. Compared with the control group, no difference in the expression of cytochrome c, Bcl‐2, or Bax was observed. Caspase‐3 activation and PARP cleavage served as a proxy of execution of apoptosis pathway; also, no significant between‐group differences are noted. Data are presented as the median and third quartile. A through C, Comparisons of expression level of cytochrome c, Bcl‐2, and Bax in brain tissue by Western blotting. D and E, Comparisons of ratios between full‐length and cleaved caspase 3 or PARP in brain tissue by Western blotting.

Figure 6. Western blot analysis of beclin‐1, LC3B, p62, and LAMP2 expression.

Reduced expression levels of beclin‐1 and LAMP2 and lower LC3B‐II: LC3B:I ratio suggest downregulation of autophagy. Data are presented as the median and third quartile. Pairwise comparisons of protein expression levels or ratios between experimental and control groups were performed using the post hoc Dunn test. The asterisk indicates statistical significance. LAMP2 indicates lysosome‐associated membrane protein 2.