Animals.

The male and female NMRs (ages 2–4 yr) used in this study were second- or third-generation captive-born animals, descended from animals captured in Kenya in 1980. Animals were maintained at the University of Texas Health Science Center at San Antonio (UTHSCSA) in normoxic conditions in interconnected systems consisting of tubes and cages of varying sizes to approximate the multichambered burrow and tunnel systems that NMRs inhabit in the wild. NMRs were housed under climatic conditions approximating their native habitat (30°C; ≥50% relative humidity). They met all their nutrient and water needs through an ad libitum supply of fruit and vegetables, supplemented with a protein- and vitamin-enriched cereal (Pronutro; Bokomo, Rosendal, South Africa). The male and female mice in this study, of the C57BL/6J strain (ages 3–5 mo), were group housed at 25°C on a 12:12-h light-dark cycle and were given ad libitum access to food and water. The ages selected for this study allowed for physiological age matching between species such that both were at equivalent percentages of maximum life span. This study was approved by the Institutional Animal Care and Use Committee (protocol 11042x) at UTHSCSA.

Heart rate and two-dimensional echocardiography.

Heart rates from awake, unanesthetized mice and NMRs were measured using a modified blood pressure tail cuff (Hatteras Instruments, Cary, NC). These measurements were acquired after a minimum of three training sessions. For echocardiography, mice and NMRs were anesthetized using 1–2% isoflurane in a 100% oxygen mix and placed in the supine position with paws taped to electrodes on a temperature-controlled electrocardiogram board. Heart rate, respiratory rate, and temperature were monitored to ensure the data collected were physiologically relevant. Measurements were acquired with the Vevo 2100 Imaging System (VisualSonics, Toronto, ON, Canada) to obtain two-dimensional parasternal LV B-mode recordings at the long axis and M-mode recordings at the short-axis midpapillary region according to the recommendations set out by Lang et al. (24). Stress was induced by an intraperitoneal injection of dobutamine (3 μg/g body mass) for each animal. This dose was chosen because it was the minimum at which heart rate and fractional shortening were maximally stimulated in both species after dose-response experiments were conducted (data not shown). Echocardiograms were recorded at baseline before injection and 30 min after injection.

Biaxial tensile testing.

To investigate the possibility of species LV functional differences being due to differences in mechanics, biaxial testing was performed. NMRs and mice were anesthetized with isoflurane, and hearts were excised. LVs were separated from right ventricles and atria and then horizontally sectioned and cut in half along the circumference such that a midpapillary region of the free wall was isolated. Samples were submerged in PBS at 25°C in a tissue bath and mounted to racks connecting to actuators in a CellScale BioTester (Waterloo, ON, Canada) that stretch tissue simultaneously in the longitudinal and circumferential directions. Testing was conducted with samples following three different loading protocols: 20% equibiaxial stretch over 30 s, 20% circumferential stretch and 10% longitudinal stretch over 30 s, and 10% circumferential and 20% longitudinal stretch over 30 s. To reduce viscoelastic effects, each protocol was conducted slowly six times for tissue preconditioning, and only the final cycle was considered for analysis. Forces and displacements were determined from actuator output data, and Cauchy stresses were calculated following a protocol described by Fomovsky and Holmes (14).

Stress and displacement data were fitted to a two-dimensional stress-strain relationship based on a four-parameter Fung-type strain energy density equation:

Specifically, it yields a two-dimensional orthotropic nonlinear stress-strain relationship (27): where W is the strain energy density; E_θ and E_z are the circumferential and longitudinal components of the Green strain, respectively; and c, b₁, b₂, and b₃ are material constants determined by fitting Eq. 2 to the experimental data. To directly compare the mechanical properties under physiological conditions, the stretch ratios under diastolic pressures were calculated using the Law of Laplace and diastolic dimensions.

Histological analyses.

After tensile testing, the LV midpapillary slices were fixed in 10% zinc formalin and embedded in paraffin. Two 5-μm-thick sections were collected every 50 μm through the thickness of the sample. For each pair, one of the slides was stained with hematoxylin and eosin (H&E) for cardiomyocyte alignment and the other with picrosirius red (PSR) for collagen density. For alignment and collagen analyses, slides were imaged at three locations along the circumferential midline of the specimen at ×10 magnification. Image processing to calculate alignment and density was conducted using custom Matlab-based codes (version 7.14; The MathWorks, Natick, MA). Briefly, cardiomyocyte alignment was calculated using a gradient-based edge-finding routine. Collagen density was calculated by counting the percentage of pixels in the sample that were positively stained with PSR.

Tissue from a separate population of animals was taken for cardiomyocyte cross-sectional area measurements. For these, animals were euthanized with isoflurane anesthesia, and hearts were flushed with cardioplegic solution before excision to arrest cardiomyocytes in diastole. LV midpapillary regions were isolated as before but here fixed and embedded without circumferential sectioning before being stained with H&E. To calculate cardiomyocyte area, 30 myocytes per section were randomly scanned at ×40 magnification according to methods previously described such that quantifiable cardiomyocytes were rounded and had central nuclei (28). Areas were quantified using NIS-Elements imaging software (Nikon, Melville, NY).

Statistical analyses.

Five male and five female NMRs in addition to five male and five female mice were used for in vivo cardiac physiology comparisons and echocardiographic assessments. For all remaining analyses, two additional animals of each sex and species were included to provide sufficient tissue for the various morphometric measurements undertaken. Samples were randomly selected for histological evaluations, with the number of animals per experiment denoted in all figure legends. For all assessments, sex differences were evaluated using unpaired t-tests. Since no sex differences were evident in either species, male and female data were combined. Data are expressed as means ± SE. Unpaired t-tests (with Welch's corrections if the variances were significantly different) were used to compare values between species for basic measurements, echocardiography, biaxial testing, collagen deposition, and cardiomyocyte cross-sectional area. Linear regression was used to determine the rotation of cardiomyocytes through the thickness of the samples. Samples with correlation coefficients <0.5 were excluded from the data set because they may have been damaged during biaxial testing. A circular statistics test based on the Rayleigh's test was used to determine if cardiomyocytes had an alignment that was significantly different from a random uniform distribution (38). GraphPad Prism 5 (GraphPad Software, San Diego, CA) or Matlab were used for all statistical calculations.

Observed and predicted species differences in basic cardiac parameters.

NMRs had both larger body and heart masses compared with mice, but given the interspecies body size differences, the ratio of heart mass to body mass was significantly smaller for NMRs (Table 1). Allometric analyses based on the species' respective average weights revealed that both species had heart weights lower than predicted (32). Average unanesthetized NMR heart rate was less than half that of the mouse and approximately half the rate predicted for a 45-g mammal. In contrast, the mice had a 35% higher heart rate than predicted by mass (5). NMR cardiac output, in keeping with the low heart rate, was also less than half that of the mouse and less than half that predicted by mass. Conversely, mouse cardiac output was much higher than its allometrically predicted value (36).

Assessment of basal LV dimensions and function.

Pronounced species differences in LV dimensions and function were evident under basal conditions. M-mode echocardiograms (Fig. 1) revealed that NMRs displayed larger LV wall thicknesses than mice in diastole (Table 2). However, it must be noted that NMRs have significantly larger hearts overall compared with mice (Table 1). Conversely, end-diastolic dimension was not different between species, but end-systolic dimension in the NMR was significantly larger. All LV function parameters (fractional shortening, ejection fraction, cardiac output, and stroke volume) were significantly lower in the NMR than the mouse, indicating reduced cardiac function in the NMR under basal conditions.

LV dimensions and function under exercise-like stress.

Given these results, it was important to determine whether the low cardiac function of young, healthy NMRs under nonstressed conditions was indicative of an underlying unhealthy heart condition or was a species-specific trait. If the former was true, NMR cardiac function under stress would also be compromised (37). Since dobutamine is routinely used to induce an exercise-like stress response, treatment with this drug was also used to evaluate the cardiac reserve of the species. Dose-response data revealed that intraperitoneal injection of dobutamine at 3 μg/g body mass elicited maximal inotropic and chronotropic responses in both species. End-systolic dimension with dobutamine was no longer significant between species, indicative of a greater increase in NMR cardiac function (Table 2). Furthermore, NMR ejection fraction and stroke volume were no longer significantly lower than the mouse parameters with dobutamine treatment.

Determination of species-specific responses to dobutamine further elucidated NMR heart function. There was no difference in chronotropic responses to dobutamine, as shown by the similar percentage change in heart rate from baseline to 30 min after dobutamine treatment (Fig. 2A). The NMR inotropic response, however, was significantly higher than that of the mouse: NMR fractional shortening increased by 72 ± 10% compared with 34 ± 4% in the mouse (Fig. 2B). The NMR increase (76 ± 8%) in ejection fraction was about twofold greater than that of the mouse (41 ± 7%; Fig. 2C). Percent change in cardiac output was again significantly higher in the NMR (77 ± 12%) compared with the mouse (45 ± 4%; Fig. 2D). Following the same trend, NMRs greatly increased their stroke volume (42 ± 9%), whereas mice did not (2 ± 3%; Fig. 2E). The stroke volume data are in keeping with previous findings that this measurement does not change with dobutamine treatment in mice (37). Together, these results indicate that the NMR heart did not display any apparent dysfunction, suggesting that the observed low basal cardiac function was a natural healthy phenotype for this species. Furthermore, NMRs possess an enhanced cardiac reserve such that the work of the NMR heart can be increased substantially beyond what is necessary for the animal's basic functions of living.

Ventricular mechanical properties.

Predicted circumferential wall stresses as calculated by the Law of Laplace were not different between species in both systolic and diastolic states (Fig. 3A). The lack of a species difference in systolic stress indicates that reduced systolic loading is likely not a means of limiting myocardial wear and tear. Biaxial tensile testing under physiologically relevant pressures and stresses showed no differences in circumferential or longitudinal stretch ratios under estimated diastolic conditions, indicating similar passive mechanical properties (Fig. 3B).

Histological examination of ventricular composition.

To further probe the mechanisms for species differences in contractility and LV structure, we set out to determine cellular differences in LV structure. The cross-sectional area of cardiomyocytes was increased in NMRs (216 ± 10 vs. 178 ± 7 mm²) (Fig. 4). Interestingly, it was more difficult to find myocytes properly aligned for quantification in NMR tissue than in that of mice. This may be evidence that NMR cardiomyocytes are less longitudinally arranged than those of mice.

Cardiomyocyte alignment is known to vary linearly through the thickness of the wall. However, it is also dependent on the size of the heart such that smaller hearts have greater myocardial fiber rotation to develop greater torsion (17, 29). In the present study, NMR LV tissue showed a more overall circumferential alignment of cardiomyocytes than that of the mouse due to reduced rotation through the thickness of the myocardium (Fig. 5A). Furthermore, NMRs show reduced rotation of fiber angle despite having cardiomyocyte arrangement at the epicardial surface that is comparable to that of mice (Fig. 5B).

Interstitial collagen content was quantified because collagen provides structural support for cardiomyocytes and contributes to the mechanical properties of the LV. NMRs exhibited a nonsignificant trend for increased interstitial collagen deposition (P = 0.054) as evidenced by PSR staining (Fig. 6).

Species lifestyle explains low cardiac function in the NMR.

Supporting our hypothesis, we found that NMRs have much lower heart rate, fractional shortening, ejection fraction, cardiac output, and stroke volume at baseline compared with mice (Tables 1 and 2). Lower cardiac function under basal conditions in the NMR likely reflects the lower metabolic requirements associated with ecophysiological adaptations to life underground (15). Gas exchange in the burrows of the NMR is poor, because air movement is largely restricted to diffusion through soil. The reduced oxygen availability is further exacerbated by the large number of animals, microorganisms, and plant roots respiring together. NMRs have adapted to such forbidding conditions by evolving extreme tolerance of variable oxygen atmospheres, particularly hypoxia (6). For example, NMR brain slices maintain synaptic activity for three times longer than mouse slices kept under identical anoxic conditions (25). Low levels of oxygen consumption and consequent low resting metabolic rates, which are 66–75% of those expected allometrically in the NMR (9, 15), are also considered an adaptive trait to life in a low-oxygen environment (31). With this low resting metabolism, the NMR's low cardiac function at baseline is likely adequate to meet the basal energy needs of this animal.

Additionally, NMRs have high hematocrit levels, and their hemoglobin has a higher affinity for oxygen compared with aboveground-dwelling mammals (6, 20). Even after being housed under normoxic conditions in captivity for more than 30 years, NMRs maintain hematocrits of ∼50%. Logically, with this greater oxygen uptake per heartbeat, the NMR heart should not have to pump as hard to maintain adequate oxygen and nutrient supply to other tissues, especially under normoxic conditions. Low resting body temperature (∼32–35°C), stemming from both reduced metabolism and lack of an insulating pelage, could also promote low NMR cardiac function by decreasing cardiac output. A decrease in temperature by 1°C can cause a 2% increase blood viscosity (1, 21). This, coupled with the already higher hematocrit levels, means the NMR heart must pump against a much more viscous fluid than those of aboveground-dwelling mammals.

Both tension in the LV wall and the velocity of contraction are important contributors to myocardial oxygen consumption (3). Based on our findings of low cardiac function, the tension and contraction velocity should be low in the NMR heart. Thus it is plausible that the low cardiac function serves as an energy-saving strategy for the NMR heart. Furthermore, a lower heart rate prolongs the diastolic filling period, and basal oxygen consumption of the heart is 20% of that during systole (3). Therefore, a lower resting heart rate may promote overall metabolic efficiency in the NMR. A strategy that limits basal energy expenditure is thus adaptive given both the NMR's harsh underground living conditions and its energetically costly foraging process.

Increased cardiac reserve in the NMR.

Despite our knowledge of reduced basal metabolic requirements in the NMR, it was important to ascertain whether the NMR heart was truly “idling” at a lower basal level or functionally compromised due to pathological reasons. We thus used a β-adrenergic receptor agonist, dobutamine, to mimic exercise-induced cardiac stress, since it was not possible to undertake echocardiography in conscious animals running on a treadmill. Dobutamine, like exercise, induces an increase in both heart rate and contractility and is commonly used as a stress test to determine cardiac functional reserve. If NMR heart function was compromised, dobutamine stress would be unable to induce greater contractility (37).

It is doubtful that the NMR heart is less healthy than that of the mouse, because dobutamine treatment caused a greater increase in cardiac function in NMRs. LV end-systolic dimensions were similar between species with dobutamine treatment, whereas they were significantly greater in NMRs at baseline (Table 2). This is evidence of a greater increase in cardiac function in the NMR that is further supported by the echocardiographic measurements of LV function. NMRs show significantly higher percentage changes in fractional shortening, ejection fraction, cardiac output, and stroke volume compared with mice (Fig. 2). Still, both species exhibit similar chronotropic stimulation with dobutamine, because heart rate changes were not significantly different. Despite this, the NMR undergoes a near doubling of most parameters of its cardiac function under exercise-like conditions, unlike the mouse. This is evidence of a large cardiac reserve, which may allow NMRs to meet the obligatory high energy needs associated with burrowing (30) despite having low basal cardiac function.

Having low basal metabolism allows NMRs to be economical with their energy expenditure. In sub-Saharan Africa, the NMR’s food sources are sparse and the soil is densely compacted. The energetic cost of digging to find food in this environment is thus extremely high, causing metabolic rates to increase more than fivefold from resting levels (30). Collectively, adaptations to a burrowing lifestyle and hypoxic environment likely contribute to the low basal cardiac function and heart rate of NMRs and provide a reason why the NMR displays so large an increase in cardiac reserve.

Pathology unlikely despite low NMR cardiac function.

Diminished basal NMR cardiac function (Table 2) was accompanied by morphological traits commonly associated with cardiac pathology in both mice and humans (10). Histological assessments revealed significantly greater cardiomyocyte cross-sectional area and a trend, albeit not significant (P = 0.054), toward increased LV collagen content in NMRs relative to mice (Figs. 4 and 6). Both increased cell size and increased collagen deposition are commonly seen in LV dysfunction and are associated with cardiomyocyte cell death and fibroblast proliferation (22). Nonetheless, it is unlikely that these histomorphometric features allude to a pathological state in this species, for the young NMRs examined in the present study displayed increased cardiac reserve. Cardiac reserve is commonly diminished in disease states (23). Increased cardiomyocyte cross-sectional area, rather, is likely to be a hallmark of the species, since NMRs have larger hearts and greater LV wall thicknesses than mice. Although increased collagen limits LV relaxation capabilities (22), the trend toward increased LV collagen deposition in young NMRs is not detrimental (Fig. 6), because we have seen no difference in diastolic relaxation between species (Fig. 3B).

It has been suggested that the smaller cardiomyocyte size and lower collagen content seen in aged Ames dwarf mouse hearts are beneficial and allow this mutant strain to achieve greater longevities compared with wild-type mice (18). Although this premise may hold true within a species, it does not do so across species of disparate longevities. Indeed, we have observed traits in NMR hearts opposite to those of the Ames dwarf mice despite the fact that the extraordinary longevity of NMRs is seven- to eightfold greater than the maximum life span of Ames dwarf mice (2, 8, 11). Furthermore, we have previously shown that NMRs display attenuated age-related declines in cardiac diastolic function (16). These histomorphometric traits must then reflect species-specific qualities that do not hamper evolutionary fitness or hinder the NMR from meeting its physiological requirements. Interestingly, exceptionally high cardiac collagen deposition (13–18%) does not keep the Burmese python from markedly boosting its metabolism after a large meal. This high collagen deposition is drastically different from the normal 1–2% commonly found in most mammalian hearts, and yet this reptile does not display cardiac dysfunction (33). On the contrary, the Burmese python is capable of extreme physiological cardiac hypertrophy, which is accompanied by a 3.3-fold postprandial augmentation of cardiac output (33, 34). Clearly, the python's cardiac function is not hindered by such high collagen. It is instead a unique nondetrimental cardiac feature of the species like those we have described presently in the NMR.

The lower basal cardiac function observed in NMRs might reflect species differences in cardiomyocyte fiber rotation. Mice have much greater fiber rotation in their LVs compared with NMRs (Fig. 5). This promotes increased contractile efficiency, because greater fiber rotation is linked to improved torsion, allowing a heart to better eject blood with each contraction (29). However, mice also have significantly smaller cardiomyocyte cross-sectional area and smaller diastolic wall thickness in their overall smaller hearts (Fig. 4, Table 1). It is possible that the greater fiber rotation is a compensation for the thinner walls and fewer cardiomyocytes in the smaller mouse heart (17). In contrast, the reduced fiber rotation contributes to lower contractility in the NMR heart. However, low contractility is in line with the low-energy, basal state in which NMRs exist until they have to meet the high energy demands of burrowing. Despite this, low basal contractility does not necessarily limit mechanical wear and tear on the NMR heart, because the mathematically predicted systolic stress does not differ between species (Fig. 3A). However, the species' reduced heart rate may limit the effects of cyclic loading, thereby reducing cardiac tissue fatigue.

In summary, this first study characterizing the hearts of young NMRs has revealed reduced baseline cardiac function yet enhanced cardiac reserve, in keeping with the species' inherent ecophysiology. Assessing these unique cardiac features of the NMR was an essential step in establishing the groundwork for understanding how the heart of this novel species responds to aging and other stressors. Future studies will assess NMR cardiac function with age and explore how this species may employ protective mechanisms in its heart to withstand both oxidative stress and the vagaries of time.