Low temperature, low metabolic rate, and high affinity hemoglobin

Naked mole-rats employ at least two mechanisms to conserve oxygen-dependent energy. The first mechanism involves a reduced generation of body heat when they are within their thermoneutral zone (reviewed in Buffenstein et al. 2021; Buffenstein and Craft, 2021). Within their equatorial burrows, naked mole rats can find a desirable ambient temperature by moving to different depths of the burrow as the surface of the earth heats and cools over the course of the day and night (Jarvis 1990). In other words, they can behaviorally thermoregulate. Holtze et al. (2017) found that burrow temperatures range over 24.6–48.8 °C.

Other mammals use physiological thermoregulation and expend a huge amount of energy to maintain their desired body temperature (Cannon and Nedergaard 2011). Hence, by favoring behavioral thermoregulation, naked mole-rats greatly reduce their need for oxygen-dependent energy (Buffenstein and Yahav 1991). Interestingly, in the laboratory, naked mole-rats exposed to cool temperatures (e.g. 20 degrees C) attempt to physiologically thermoregulate. However, they are unable to maintain their preferred temperature (Oiwa et al. 2020). Also interesting, within captive colonies maintained within their thermoneutral zone, naked mole-rats regulate their body temperatures within 1 degree C by huddling together (Yahav and Buffenstein, 1991). Another way that naked mole-rats behaviorally respond to hypoxia is that they progressively reduce activity in response to progressively more severe hypoxic atmospheres, presumably to save energy (Ilacqua et al. 2017).

Their second energy saver involves the naked mole-rat’s reduced resting metabolic rate, about 2/3rds that of similar-sized rodents (Buffenstein and Yahav 1991). In addition to these oxygen conserving mechanisms, naked mole-rats have very high affinity hemoglobin that can bind oxygen molecules in hypoxic environments that would be fatal to humans (Johansen et al. 1976).

Intrinsic brain tolerance to hypoxia

Naked mole-rats demonstrate intrinsic brain tolerance to hypoxia. This was shown in experiments with hippocampal brain slices (Larson and Park 2009). Figure 2 shows hippocampal slices in the recording chamber. Note that 95% O2 is used in the bath and atmosphere of the chamber, which is the standard for brain slice experiments.

Fig. 2

Hippocampal brain slices are pictured in a recording chamber where the concentration of O2 in the bath and atmosphere are precisely controlled. Stimulating and recording electrodes are positioned to generate and record evoked potentials (the summed activity of many neurons). The inset shows an example evoked potential. The initial, brief downward deflection is the stimulus artifact and indicates when the slice was stimulated. The large downward deflection is the evoked potential. Photo credit Thomas Park

Figure 3 shows the key results of that study. Figure 3 A shows the effects of exposing slices to 30 min of 15% O2 after achieving a stable baseline in 95% O2. The graph shows the amplitude of evoked field potentials triggered by a stimulating electrode. The stimulus and corresponding response took place every 20 s, and slices from mouse and naked mole-rat were tested side-by-side. The two example curves are from a mouse (open circles) and a naked mole-rat (NMR, closed circles) tested in the same bath solution. The horizontal bar labelled “15% O2” corresponds to the time when the O2 supply was changed from 95 to 15%. Exposure to 15% O2 caused the curve from the mouse to decline, indicating reduced synaptic function. Synaptic function recovered after 95% O2 was restored. The curve from the naked mole-rat was unaffected.

Figure 3B is in the same format as Fig. 3 A except that the period of hypoxia was more severe with only 10% O2. In this case, the slice from mouse completely lost function and did not recover. The slice from naked mole-rat declined in function but recovered after the hypoxic exposure. Figure 3 C shows the averaged decrease in evoked potential amplitude for various O2 concentrations, and Fig. 3D shows the percentage of slices that were able to recover from exposure to each O2 concentration. Clearly, the slices from naked mole-rat demonstrated significantly higher tolerance and higher recovery rate from hypoxia compared to slices from mouse.

Figure 3E shows the effects of exposure to zero O2 (anoxia). In this situation, functionality for the slice from naked mole-rat declined much slower than functionality for the mouse. Remarkably, the slice from naked mole-rat, but not mouse, recovered after O2 levels were restored. Figure 3 F shows the duration of 0% O2 exposure required to reach total loss of function (termed anoxic depolarization, AD) for mouse (MSE) and naked mole-rat (NMR). Data is shown for two bath temperatures. 35° C, which is close to mouse body temperature, and 30° C, which is close to naked mole-rat body temperature. In both cases, slices from naked mole-rats retained functionality significantly longer than slices from the mice.

Fig. 3

Intrinsic hypoxia tolerance in hippocampal brain slices from naked mole-rats compared to slices from mice. A and B, example curves from slices transiently exposed to 15% and 10% O2, respectively. C and D, summary data from naked mole-rat slices (n = 25) and mouse slices (n = 48). E, example curves from slices exposed to zero O2. F, time to complete loss of function during exposure to zero O2. Please see text for detail. This figure is from Larson and Park (2009)

Upregulated expression of GluN2D, which is usually associated with hypoxia tolerance in neonates

One factor likely contributing to intrinsic hypoxia tolerance in naked mole-rat brain is the high expression level of GluN2D in adults of this species. Hypoxia and the related decline in energy trigger a cascade of cellular processes resulting in most NMDA-type calcium channels opening which allows toxic levels of calcium into the cells (Peterson et al. 2012b and references therein). Brain cells from neonates lessen this effect because they express a substantial number of NMDA channels containing the 2D subunit (GluN2D), which open much less during hypoxia (Bickler et al. 2003). In most mammals, the expression of GluN2D decreases precipitously postnatally (Laurie et al. 1997). Peterson et al. (2012b) found that adult naked mole-rats retain 66% of neonatal levels of GluN2D whereas adult mice only retain 13% of neonatal levels (Fig. 4 A).

NMDA-type calcium channels open when glutamate binds to them. Interestingly, Cheng et al. (2022) found that overall levels of glutamate decrease significantly in the brain of naked mole-rats exposed to hypoxia compared to mice. Decreasing glutamate levels is likely another way that naked mole-rats reduce calcium entry into cells to achieve tolerance to hypoxia.

Reduced intracellular calcium accumulation from hypoxia

Subsequent experiments using calcium imaging showed that hippocampal cells from naked mole-rats displayed much less intracellular calcium from hypoxia compared to cells from mice (Fig. 4B,C). The bars in Fig. 4B indicate the average percent change in calcium-mediated fluorescence, with negative values corresponding to an increase in calcium (calcium decreases the fluorescent signal). An example from one experiment is shown in Fig. 4 C., which displays hippocampal cells before (top) and during (bottom) hypoxia.

Fig. 4

Retention of GLuN2D and reduced intracellular calcium from hypoxia in naked mole-rats. A. The bars show the percent of GLuN2D retained into adulthood for mice and naked mole-rats. This panel is from Peterson et al. (2012b). B. Florescent imaging results are shown for hippocampal cells from young (post-natal day 6, P6) and adult-like (post-natal day 20, P20) mice and naked mole-rat. Both ages of mice showed more intracellular calcium from hypoxia compared to naked mole-rats. Note that more negative values correspond to more calcium. C. Example data from one experiment. Panels B and C are from Peterson et al. (2012a)

Whole animal tolerance to hypoxia and anoxia

The previous findings – reduced need for physiological thermoregulation, reduced resting metabolic rate, hemoglobin with a high affinity for O2, and intrinsic brain tolerance to hypoxia – set the stage for testing the whole animal. Testing whole animals revealed that naked mole-rats were many times more tolerant to both hypoxia and anoxia compared to mice (Park et al. 2017). Figure 5 shows the results of testing naked mole-rats and mice in atmosphere chambers where O2 concentrations were precisely controlled. Figure 5 A illustrates the similarity in size of a mouse and a naked mole-rat. Figure 5B shows the results of testing mice and naked mole-rats in hypoxia (5% O2). On average, the mice did not survive more than 12 min whereas the naked mole-rats were able to survive for 5 h. This experiment was arbitrarily stopped at the 5-hour point.

The results from testing animals in anoxia (0% O2) are shown in Fig. 5 C. During anoxia, both the mice and the naked mole-rats lost consciousness and ceased voluntary movements in about 30 s. Respiration attempts were monitored, and an animal was removed from the atmosphere chamber if there was no respiration attempt for 20 s for mice or 60 s for naked mole-rats. On average, the mice made their last respiration attempt after about 40 s into the exposure, whereas the naked mole-rats made their last respiration attempt after about 240 s (4 min). Remarkably, all of the naked mole-rats survived, even after an additional minute under anoxia, whereas none of the mice survived a much shorter exposure.

Fig. 5

In vivo (whole animal) exposure to hypoxia (5% O2) and anoxia (0% O2) showed that naked mole-rats were extremely tolerant to these conditions compared to mice. Please see text for details. This figure is from Park et al. (2017)

The experiments described above showed that naked mole-rats could survive 5 min of exposure to anoxia (4 min to the last breath plus an additional minute before they were returned to room air (normoxia)). However, the next set of experiments revealed that 5 min was an underestimate of their capabilities. For the next experiments, respiration rate and heart rate were recorded, and naked mole-rats were exposed to anoxia for either 10, 18, or 30 min. Figure 6 A shows the results for respiration rate from the 18-minute exposure. The data is represented as average number of breaths per 10 s intervals. Respiration rate declined precipitously within the first few minutes of exposure, remained very low during the exposure, and then increased slowly when the atmosphere was returned to normoxia.

During these experiments, heart rates were simultaneously recorded from the same animals using an ECGenie recording system (Mouse Specifics, Inc.) where the animal’s feet were placed onto a platform with embedded electrode pads. Figure 6B shows average heart rate data from naked mole-rats and mice. For mice, the amplitude of the electrocardiogram trace declined into the noise after about 6 min in anoxia. In contrast, the average heart rate for the naked mole-rats declined from about 200 beats per minute in normoxia to about 50 beats per minute in anoxia. The average heart rate for naked mole-rats remained remarkably stable for the duration of the anoxia exposure, and then when the atmosphere was returned to normoxia, the heart rate gradually returned to baseline over the next 20 min. The end point of this experiment was when each naked mole-rat righted itself and walked off of the electrocardiogram contact pads.

Figure 6 C shows electroencephalography (EEG) data from a different cohort of naked mole-rats. Recordings were made in normoxia for 20 min prior to anoxia (-20 to 0 on the x-axis). In normoxia, brain activity was robust and quite variable. However, when the animals were exposed to anoxia, activity became extremely reduced (time 0 to 10 on the x-axis). The inset shows an expanded view of the time when they were in anoxia. Above the graph is an example EEG trace from one animal.

Cessation of voluntary movements and greatly reduced respiration rate, heart rate, and brain activity are reminiscent of a state called suspended animation (Blackstone et al. 2005; Blackstone and Roth, 2007), where biological functions are slowed to preserve physiological capabilities (Safar et al. 2000). This idea is consistent with the finding that naked mole-rats suppress their metabolism by up to 85% in acute severe hypoxia (Pamenter et al. 2018). The ability to go into a suspended animation state should be a great benefit to naked mole-rats for conserving energy during times of very low oxygen availability.

Fig. 6

Physiological responses to anoxia. A and B. Average respiration rate (A) and average heart rate (B) dramatically decreased during 18 min of anoxia (N = 3). The red bar on the X-axis corresponds to when the animals were exposed to anoxia. These figures are from Park et al. (2017). C. EEG data shows that average brain activity also dramatically decreased during anoxia (N = 3). This figure is from Park et al. (2021)

Metabolic re-wiring

In order to sustain even the most minimal heart and brain function (Fig. 6), tissues demand energy in the form of ATP, which is primarily generated via the glycolytic pathway or oxidative phosphorylation in the mitochondria. When oxygen is limited, ATP can be generated via anaerobic glycolysis, a process far less efficient but circumventing the need for oxygen. Metabolite analysis in various tissues and blood in the naked mole-rat revealed an unexpected finding. When exposed to an atmosphere with low or no oxygen the naked mole-rat activates the synthesis and utilization of fructose and sucrose (Fig. 7 A). Sucrose (a disaccharide composed of fructose and glucose) and fructose accumulated to high levels in the blood and tissues in the naked mole-rat exclusively under hypoxia. These metabolites were not detected in mouse tissue undergoing a similar hypoxic challenge. To support the idea that naked mole-rats switch to fructose metabolism, fructose-1-phosphate, a metabolite specific to fructose metabolism was also detected (Park et al. 2017). These findings suggested that in response to low oxygen, naked mole-rats are able to produce fructose in tissues such as liver and kidney, export this metabolite into circulation and deliver it to vital organs such as the brain and heart where it is used in glycolysis to generate ATP.

Fructose metabolism is largely limited to several tissues in most mammals, primarily the liver, kidney and intestine. This is due to restricted expression of the fructose transporter GLUT5 (Km = 6mM) and the less efficient transporter GLUT2 (Km = 11mM) (Hannou et al. 2018), as well as fructolytic enzymes fructokinase (KHK), aldolase reductase (ALDOB and C) and triokinase (TKFC) (Hannou et al. 2018). In contrast to mice and humans, naked mole-rats express high levels of Glut5 and Khka and Khkc mRNA across all tissues (Park et al. 2017) (Fig. 7B). Surprisingly, high expression of fructose-specific genes was observed in tissues from animals exposed to chronic normoxia suggesting that the naked mole-rat transcriptome is remodeled to enable efficient fructose uptake and utilization at all times and is not regulated by oxygen levels. Enhanced fructose metabolism under hypoxia seems therefore to be dependent on fructose availability rather than ability to transport and metabolize it. Indeed, increase in fructose and sucrose occurred only under low oxygen conditions. Under normoxic conditions, the levels of fructose and sucrose were very similar between mouse and naked mole-rat in all tissues tested.

Glucose is initially metabolized by the glycolytic pathway. Phosphofructokinase (PFK1) is the most critical regulatory enzyme in glycolysis, allosterically inhibited by ATP, citrate and lactate (Costa Leite et al. 2007; Kemp and Foe 1983). Under hypoxia, as lactate accumulates due to anaerobic glycolysis, PFK1 enzyme becomes inhibited leading to a slowing down of glycolytic flux and limiting the supply of ATP and glycolytic intermediates. Under these conditions, utilization of fructose as opposed to glucose can become beneficial to the cell. Fructolysis circumvents PFK1 inhibition by diverting metabolism towards the KHK enzyme which phosphorylates fructose to fructose-1-phosphate (F1P). F1P in turn is catalyzed to Dihydroxyacetone phosphate (DHAP) and Glyceraldehyde (GA) which enter glycolysis at a step downstream of the PFK1 enzyme (Fig. 7 C). The metabolism of fructose, unlike glucose is therefore an unchecked process, independent of the energy or metabolic state of the cell, enabling the cell to continue to attain ATP and glycolytic intermediates. A secondary advantage of fructose utilization has recently been demonstrated in APC−/− tumor cells (Goncalves et al. 2019). Because the KHK enzyme has a very fast kinetic and is unregulated, the initial phosphorylation of fructose as it enters the cell depletes ATP dramatically. This manifests in a short relief of PFK1 inhibition, promoting glycolytic flux from glucose to generate sufficient ATP and glycolytic intermediates for cellular growth and maintenance (Van Den Berghe et al. 1977; Johnson et al. 2020; Mirtschink et al. 2018). The effect of fructose on glucose metabolism has not yet been investigated in the naked mole-rat however it is possible that similar to what was found in cancer, fructose may be relieving the block on glucose uptake and metabolism allowing for a greater glycolytic flux which would be advantageous under low oxygen conditions.

In a recent study, Hadj-Moussa et al. (2021) reported a downregulation in both Glut5 and Khk protein in the brain of naked mole-rats 4 h post-hypoxia at 7% oxygen. Furthermore, Khk was predicted to be a target of a differentially expressed miRNA (miRNA365) which was upregulated under hypoxia. Whether or not the downregulation of Glut5 and Khk protein observed in this study results in a reduction of fructose uptake or metabolism in the brain under hypoxia is yet undefined since neither of these parameters were examined in the study (Hadj-Moussa et al., 2021). Further studies need to be done to examine whether different grades of hypoxia elicit unique adaptive metabolic responses in the naked mole-rat.

Fig. 7

Switch to fructose metabolism under hypoxia. A. A chromatogram showing a fructose peak in the naked mole-rat under normoxic conditions (black) and its increase under anoxic conditions (red). This panel is from Park et al. (2017) (B) Level of Glut5 transcript is over 10-fold higher in the naked mole-rat heart compared to the mouse but its level of expression is unchanged under hypoxia. This panel is from Reznick et al. (2021). (C) Differences between fructose and glucose metabolism. Fructose is transported into the cell by the fructose-specific transporter GLUT5 whereas glucose enters the cell via other transporters including insulin-independent GLUT1 and insulin-dependent GLUT4. Fructose is phosphorylated by fructokinase (KHK) to generate fructose-1-phosphate (F1P), which consumes ATP and phosphate. F1P is then cleaved to glyceraldehyde (GA) and dihydroxyacetone phosphate (DHAP) by aldolase B or C (AldoB and C) after which the metabolism of glucose and fructose converge. Unlike glycolysis, fructolysis bypasses phosphofructokinase (PFK1), a rate-limiting step in glycolysis, to bypass the feedback inhibition at PFK1. PFK1 is a rate-limiting enzyme of glycolysis and its activity is blocked by high levels of ATP, citrate and low pH. KHK is much faster than hexokinase (HK) in phosphorylating its substrate thus leading to rapid ATP depletion and phosphate consumption. This relieves the blockade at PFK1 and increases glucose uptake and metabolism