Opportunities and barriers to translating the hibernation phenotype for neurocritical care

Abstract

Targeted temperature management (TTM) is standard of care for neonatal hypoxic ischemic encephalopathy (HIE). Prevention of fever, not excluding cooling core body temperature to 33°C, is standard of care for brain injury post cardiac arrest. Although TTM is beneficial, HIE and cardiac arrest still carry significant risk of death and severe disability. Mammalian hibernation is a gold standard of neuroprotective metabolic suppression, that if better understood might make TTM more accessible, improve efficacy of TTM and identify adjunctive therapies to protect and regenerate neurons after hypoxic ischemia brain injury. Hibernating species tolerate cerebral ischemia/reperfusion better than humans and better than other models of cerebral ischemia tolerance. Such tolerance limits risk of transitions into and out of hibernation torpor and suggests that a barrier to translate hibernation torpor may be human vulnerability to these transitions. At the same time, understanding how hibernating mammals protect their brains is an opportunity to identify adjunctive therapies for TTM. Here we summarize what is known about the hemodynamics of hibernation and how the hibernating brain resists injury to identify opportunities to translate these mechanisms for neurocritical care.

Keywords: NIRS; TTM; cerebral ischemia; ground squirrel; ischemia/reperfusion; neurocritical care; therapeutic hypothermia; torpor.

Figure 1

Body temperature in relation to life cycle and season of the hibernating arctic ground squirrel (AGS). The hibernation season, also referred to as hibernation, is marked by repeated bouts of prolonged torpor, termed here as hibernation torpor. Hibernation torpor is interrupted by spontaneous interbout arousals (IBA) noted by rapid increases in body temperature. Spontaneous IBAs are observed in all species of hibernators when core body temperature falls below 30°C (26). During arousal from hibernation, the nadir of mammalian metabolism is over-ridden by the high energetic costs of warming core body temperature from near 0°C to ~35°C in 2–3 h. During repeated recovery from hibernation oxygen consumption surges 300 fold from about 0.01 mLO₂g⁻¹h⁻¹ to about 3 mLO₂g⁻¹h⁻¹ (27). Arousal from torpor may be spontaneous or induced by external stimuli such as gentle handling. Once initiated, evidence suggests induced arousals proceed in the same way as natural arousals with the exception that induced arousals are faster and may be more energetically demanding [Adapted from Drew et al. (28)].

Figure 2

After entrance into hibernation torpor in Syrian hamsters, systolic blood pressure (SBP) increases to a new plateau at about 50% of euthermic systolic blood pressure (SBP). By contrast, heart rate (HR) remains at a steady minimum throughout the torpor bout. SBP and HR were measured in unanesthetized animals by telemetry with the catheter of a pressure transmitter inserted into the abdominal aorta [Horwitz et al. (33)].

Figure 3

Open loops in hysteresis plots shown for SBP, HR and baroreceptor sensitivity (BRS) illustrate that cardiovascular control operates in fundamentally different ways during entrance and arousal. Importantly, the hysteresis between HR and SBP (A) illustrates that BP increases prior to HR during arousal and declines at the same rate as HR during entrance. During arousal, SBP (B) and HR (C) both increase to near maximal levels before BRS begins to increase [Horwitz et al. (33)].

Figure 4

A gradual increase in HR is the first sign of arousal although a rapid increase in systolic blood pressure (SBP) precedes the subsequent, rapid rise in HR. Maximal SBP during arousal overshoots SBP measured during euthermia. Minimal baroreceptor sensitivity (BRS) at the onset of arousal may allow for the rapid rise in SBP and HR that are needed to support the metabolic demands of heart and brain as animals rewarm from hibernation torpor seen as a change in core body temperature (T_c) [Horwitz et al. (33)].

Figure 5

Use of a novel miniaturized near infrared spectroscopy (NIRS) device for quantifying Hb and HbO₂ in small animals shows that HbO₂ decreases during arousal from hibernation in hind leg (A) and brain (B) tissue. Hibernating AGS fit with sensors on the head and leg were placed in a metabolic cage at 0 min. Handling-induced arousal produced tissue hypoxia in both tissues. Rate of O₂ consumption increased from 0.06 mLg⁻¹h⁻¹ at 0 min to a maximum of 3.1 mLg⁻¹h⁻¹ between 131 and 231 min. Data shown are mean ± SEM (A, B) and median with Q2 and Q3 defined by box and range shown as whiskers (C) (n = 8; 4M, 4F AGS, 10–11 months old).

Figure 6

Oxygen concentration in AGS brain tissue does not decrease during arousal, despite a decrease in oxyhemoglobin. Data shown are representative graphs from a single AGS collected during an induced arousal from hibernation torpor at an ambient temperature of 2°C. First and second dashed line from left shows the time when arousal and euthermia started. (A) Changes in brain temperature during arousal. (B) Changes in PtO₂ without temperature correction (P O₂, meas). (C) Changes in calculated PtO₂ with temperature correction (P O₂, Cal). (D) Changes in the rate of oxygen consumption (VO₂) measured by open flow respirometry [Ma and Wu (51) with permission].

Figure 7

Western blots show higher expression of the 17kD neuroglobin monomer in cortex of AGS compared to cerebral ischemia sensitive rat and in cortex of euthermic AGS in winter compared to euthermic AGS in summer; 100 μg of protein was resolved on 10% SDS-PAGE and membranes were incubated with anti-Ngb (Ngb PolyAntibody (FL-151), 1:200, Santa Cruz Biotechnology, sc-30144) overnight followed by incubation with HRP-conjugated secondary antibody (Gt anti-rabbit IgG, 1:5,000, Santa Cruz Biotechnology). Optical density was normalized to actin. Rats were male, 3–4 months. AGS were male and female, adult (>1 year of age). ^*P < 0.0003, t-test, n = 6 AGS vs rat; ^*p < 0.0045, t-test, n = 4 summer vs. winter euthermic AGS [adapted from Bhowmick (52)].

Figure 8

AGS brain tolerates OGD better than rat regardless of hibernation state or season. Cell death was measured from LDH released into the perfusion fluid in acute hippocampal slices from rat and euthermic AGS during the summer season (seAGS), during hibernation torpor (hAGS) and during early arousal (4 h ibeAGS) and late arousal (20 h ibeAGS), 4 and 20 h after initial handling to induce arousal. (A) LDH in perfusates increased in rat hippocampal slices exposed to OGD (rat, OGD), but not in rat slices exposed to artificial cerebral spinal fluid (aCSF) (rat, aCSF), nor in slices harvested from summer euthermic AGS and exposed to aCSF (seAGS, aCSF). A small amount of cell death is noted in slices collected from seAGS and exposed to OGD (seAGS, OGD). *p < 0.05 rat aCSF vs. rat OGD, ⁺p < 0.05 rat OGD vs. seAGS OGD, ^#p < 0.05 seAGS aCSF vs. seAGS OGD. (B) As a positive control TritonX increased LDH release in seAGS slices (*p < 0.05 0.1% TritonX vs. aCSF). (C) AGS hippocampal slices are most vulnerable to OGD when collected from AGS 20 h into an interbout arousal (20 h ibeAGS). Insert shows the sum of LDH in perfusates collected 15–210 min from onset of OGD. *p < 0.05 seAGS vs. 20 h ibeAGS, ⁺p < 0.05 4 h ibeAGS vs. 20 h ibeAGS, ^#p < 0.05 hAGS vs. 20 h ibeAGS, t-test with Bonferroni correction. (D) Exposure of slices from the same groups of animals as in (C) to aCSF has no effect on LDH release. Gray bar indicates 30 min treatment period. Data shown are means ± SEM, n = 4 slices in B, 25–30 slices per treatment in (A, C, D). The novel microperfusion method, an improvement over previous use of propidium iodide as an indicator of cell death, replicated results obtained with propidium iodide [Bhowmick et al. (20)].