The brain of most vertebrates is highly sensitive to hypoxia, whereby pathological activity ensues within minutes of exposure. Like most hypoxia-intolerant vertebrates, brainstem motor networks of the American bullfrog (Lithobates catesbeianus) exhibit hyperexcitability followed by loss of rhythmic activity in severe hypoxia, which outwardly resembles the “anoxic depolarization” in mammals. However, we recently identified that adult bullfrogs acclimated to an aquatic overwintering environment (cold-acclimated, CA) have increased tolerance to hypoxia when compared to non-acclimated controls (warm-acclimated, WA). This response involved avoidance of hyperexcitability and a nearly 30-fold increase in time until cessation of rhythmic brainstem motor output (~10 minutes to ~3.5 hours measured at 22 °C) supported by a shift to anaerobic glycolysis (Bueschke et al., Current Biology, accepted for publication). Although glycolysis sustains ATP production, a major question remains; how does the motor network maintain homeostasis with such a dramatic reduction in ATP turnover relative to oxidative phosphorylation? Glutamatergic synapses that use NMDA-type receptors (NMDARs) are energetically demanding due to their Ca2+ permeability. Indeed, NMDARs cause excitotoxicity in ischemic stroke models and contribute to respiratory-related synaptic transmission onto motoneurons in bullfrogs. Therefore, we hypothesized that NMDAR function is altered or reduced to transform the brainstem into a hypoxia-tolerant state. To test this hypothesis, we made extracellular cranial nerve recordings from in vitro brainstem preparations and compared the sensitivity of WA and CA to the block of NMDARs with an antagonist, AP5. If NMDAR function is reduced or altered following CA, we expected AP5 to improve motor function in WA, but not CA, brainstems during hypoxia. We first confirmed hypoxia tolerance after simulated overwintering. WA preparations stopped at 18 ± 13 min (mean±S.D.), with 10/11 preparations showing hyperexcitable activity. CA preparations functioned for gt;60 min in all experiments and 0/8 exhibited hyperexcitable activity (Fisher’s exact test to compare proportions of preparations with hyperexcitable activity, p=0.0001). In contrast to our hypothesis, WA and CA preparations did not show differences in sensitivity to APV for respiratory-related network variables (time until final burst and reduction in motor burst amplitude), suggesting no change in NMDAR function at respiratory synapses. However, AP5 reduced the probability of preparations exhibiting chaotic, hyperexcitable motor output in WA preparations, (3/8 compared to 10/11 in controls, Fisher’s exact test, p=0.040), while CA preparations had no apparent hyperexcitable motor responses over the entire 60-minute protocol with or without AP5. Thus, activation of non-respiratory NMDARs contributes to motor hyperexcitability during hypoxia, suggesting that extrasynaptic NMDAR function is reduced or altered in CA animals to constrain motor excitability during hypoxia. Ongoing experiments aim to assess the synaptic and extrasynaptic function of NMDARs (receptor density, calcium permeability, desensitization, hypoxia sensitivity) using whole-cell voltage clamp. These results, and future experiments, will provide new insights into the physiological and metabolic adjustments needed to transform neural circuits into a state of hypoxia tolerance.
Support or Funding Information
NIH R15 1R15NS112920-01A1
U.S. Department of Defense W911NF2010275
lt;pgt;NIH R15 1R15NS112920-01A1lt;/pgt; lt;pgt;U.S. Department of Defense W911NF2010275lt;/pgt;
