The vertebrate brain is generally highly vulnerable to damage when deprived of oxygen (anoxia); without oxygen, mitochondrial energy production ceases,and ATP-dependent processes fail. Some vertebrates, however, including certain freshwater turtles, the goldfish and the Crucian carp, can survive for weeks or months without oxygen by decreasing ATP demand low enough for anaerobic metabolism alone (lactate production) to provide sufficient energy for survival. One key energy-consuming process known to be downregulated for energy conservation in anoxia-tolerant turtles is ion flux; decreases in ion leakage across cell membranes (channel arrest) reduce the need for ATP-dependent ion pumping, while decreases in calcium influx reduce the likelihood of glutamate release, which would in turn allow calcium into the neuron to trigger cell death. But while turtles essentially enter a`reversible coma', with little electrical activity or movement, the goldfish and carp remain at least slightly active even in anoxic waters (presumably enabling them to seek out water with more oxygen), and avoid potential lactate poisoning during anaerobic metabolism by converting lactate into ethanol and excreting it across the gills. These differences led Michael Wilkie and his colleagues at Wilfrid Laurier University and the University of Toronto to investigate whether goldfish exhibit channel arrest to conserve energy like anoxia-tolerant turtles.

The team measured the activity of the key glutamate (NMDA) receptor in normoxia and found that NMDA receptor currents in the brain slice decreased by 40–50% within 20 min of anoxic exposure, the first direct evidence of channel arrest in oxygen-starved goldfish.

The team also made other goldfish anoxic in nitrogen-bubbled water, to measure the activity of the main neuronal ATP-dependent ion pump(Na+/K+ ATPase) in the brain, as a decrease in ion flux should be mirrored by a decrease in ion pumping. While the team did find that the fish's Na+ and K+ ion pumping activity dropped, it did so more slowly than the NMDA calcium current changes. Immediate changes in NMDA activity and long-term changes in ion pumping together may save considerable ATP and thus allow the fish to survive anoxia, whereas reduced cellular activity in live anoxic fish is reflected by decreased activity and ventilation rate, and loss of balance.

Knowing that the anoxic fish produce ethanol and ethanol reduces neuronal activity in mammals, the team decided to test the activity of the NMDA receptor in the presence of ethanol. They found that it did not change receptor current amplitude, suggesting that ethanol does not change in vivo ion pump activity. Thus the reduction in NMDA currents is due to factors other than the general neurodepressant effects of ethanol.

Given that anoxic Crucian carp (a close relative of the goldfish) have shown a general metabolic decrease of about 40%, Wilkie and his colleagues suggest that the behavioural changes associated with anoxia may be explained,at least in part, by alterations in NMDA receptor currents, as decreasing NMDA currents would result in decreased neurotransmitter release and altered neuronal activity. However, the mechanism by which these receptor currents is altered is as yet unknown. Brief bouts of anoxia may raise tissue ethanol levels as high as 5–7 mmol l–1, enough to depress mammalian neurons, yet Wilkie and his colleagues found that 10 mmol l–1 did not alter NMDA activity, and others report that ethanol does not alter goldfish activity levels. The combined data suggest that goldfish neurons are resistant not just to anoxia but also to the effects of ethanol – so the blind, tipsy fish you meet on Saturday night gets to claim, `It's just my NMDA receptors! What's your excuse?'

Wilkie, M. P., Pamenter, M. E., Alkabie, S., Carapic, D., Shin,D. S. H. and Buck, L. T. (
). Evidence of anoxia-induced channel arrest in the brain of the goldfish (Carassius auratus).
Comp. Biochem. Physiol.