It is generally accepted that cardiac sympathetic tone dominates the control of heart rate (HR) in mice. However, we have recently challenged this notion given that HR in the mouse is responsive to ambient temperature (Ta) and that the housing Ta is typically 21–23°C, well below the thermoneutral zone (∼30°C) of this species. To specifically test the hypothesis that cardiac sympathetic tone is the primary mediator of HR control in the mouse, we first examined the metabolic and cardiovascular responses to rapid changes in Ta to demonstrate the sensitivity of the mouse cardiovascular system to Ta. We then determined HR in 1) mice deficient in cardiac sympathetic tone (“β-less” mice), 2) mice deficient in cardiac vagal tone [muscarinic M2 receptor (M2R−/−) mice], and 3) littermate controls. At a Ta of 30°C, the HR of β-less mice was identical to that of wild-type mice (351 ± 11 and 363 ± 10 beats/min, respectively). However, the HR of M2R−/− mice was significantly greater (416 ± 7 beats/min), demonstrating that vagal tone predominates over HR control at this Ta. When these mice were calorically restricted to 70% of normal intake, HR fell equally in wild-type, β-less, and M2R−/− mice (ΔHR = 73 ± 9, 76 ± 3, and 73 ± 7 beats/min, respectively), suggesting that the fall in intrinsic HR governs bradycardia of calorically restricted mice. Only when the Ta was relatively cool, at 23°C, did β-less mice exhibit a HR (442 ± 14 beats/min) that was different from that of littermate controls (604 ± 10 beats/min) and M2R−/− mice (602 ± 5 beats/min). These experiments conclusively demonstrate that in the absence of cold stress, regulation of vagal tone and modulation of intrinsic rate are important determinants of HR control in the mouse.
- intrinsic heart rate
- cold stress
- caloric restriction
- sympathetic nervous system
the frequency of cardiac contraction in mammals is achieved through modulation of heart rate (HR) around its intrinsic rate (IHR). HR is slowed by parasympathetic nervous system activity via the muscarinic M2 receptor (M2R) (17) and elevated by sympathetic nervous system activity via the β1-adrenergic receptor (39). Blockade of these autonomic inputs indicates that murine IHR is typically about 500 beats/min, well below the generally reported resting HR of about 600 beats/min [reviewed in Ref. 27, supporting the concept that cardiac sympathetic tone predominates at rest in the mouse (19, 27)]. However, some recent studies indicate that resting mouse HR is lower than 500 beats/min at rest, calling into question the sympathovagal balance controlling HR in mice (6, 11, 35).
Our laboratories (48, 54–56) and others (8, 52) have demonstrated that mouse HR is markedly reduced when experiments are conducted in thermoneutral conditions [ambient temperature (Ta) = 30–32°C]. At this temperature, HR is generally in the range of 350–400 beats/min, in contrast to the range of 550–600 beats/min when studied at the typical Ta used for in vivo mouse physiology (21–23°C). Our prevailing hypothesis has been that the tachycardia associated with a cool Ta reflects the recruitment of sympathetically mediated nonshivering thermogenesis (13, 23, 33, 51). We have shown that when female Swiss mice are acclimated to various ambient temperatures for 48 h, a strong positive and linear relationship between HR and Ta prevails (48). One purpose of this study was to test the hypothesis that very rapid changes in Ta would produce concurrent rapid responses in metabolic rate and HR in mice. The finding would provide important additional evidence, supporting the need to carefully monitor and control Ta during the assessment of cardiovascular function in mice.
Caloric restriction is both a strategy voluntarily undertaken by millions of Americans attempting to lose weight and an experimental intervention known to extend life span in worms and rodents (43). Caloric restriction elicits a complex set of physiological effects consistent with a general homeostatic survival mechanism in response to reduced energy availability (9). These responses include reduced sympathetic activity to most organs including the heart (36, 58), increased cardiac vagal activity (1, 24), and decreased HR and blood pressure (15, 57, 59). An additional purpose of this study was to further analyze the relative contributions of the sympathetic and parasympathetic nervous systems to the bradycardia of caloric restriction.
In addition to alterations in autonomic tone, it is also possible that Ta could modulate IHR. Several endocrine/paracrine signals, including thyroid hormones (18), prostanoids (50), angiotensin (11), and oxyntomodulin (44), influence IHR. Thus we determined IHR in both thermoneutrality and at standard laboratory temperatures. To determine the respective roles of sympathetic and parasympathetic control in murine HR, we determined HR, HR variability, and IHR for mice lacking M2 and β-adrenergic receptors at both thermoneutral and standard temperatures. After observing that Ta can influence IHR, we determined whether the bradycardia of caloric restriction is also associated with a decrease in IHR. Therefore, the primary objective of this work was to gain an enhanced understanding of the autonomic control of HR in mice. The results reveal that, in addition to increased sympathetic tone, reduced cardiac vagal activity and increased IHR contribute to the increase in mouse HR when studied at Ta below thermoneutrality.
MATERIALS AND METHODS
Mice and Surgical Instrumentation
Male C57BL/6J (n = 7; Charles River) were used for temperature ramp studies. Male mice lacking all three β-adrenergic receptors (“β-less” mice), maintained on a C57Bl/6J background (β-less, n = 8; and wild-type controls; n = 12) were shipped to Florida State University for physiological assessment (generously provided by Eric Bachman, Harvard University). Mice lacking M2 receptors maintained on a C57Bl/6NTac background (M2R−/−, n = 8) and wild-type controls (n = 8) were shipped to Williams College for physiological assessment (provided by J. Wess, National Institute of Diabetes and Digestive and Kidney Diseases). Mice were studied at 3 to 4 mo of age. Animal experimentation was approved by local Institutional Animal Care and Use Committees and was in accord with the American Physiological Society's “Guiding Principles in the Care and Use of Animals.”
Mice were anesthetized (halothane, 1–2% in 95% O2-5% N2 mixture) and instrumented with either an ECG telemetry device (TA10-F20; Data Sciences, St. Paul, MN) or with a blood pressure telemetry device, with the catheter placed in the left common carotid artery (TA11PA-C20; Data Sciences) for cardiovascular assessment (48, 56). For the implantation into the carotid artery, we used the procedures described by Butz and Davisson (10). Mice were individually housed using a 12-h:12-h light-dark cycle and were allowed to recover from the surgery for a minimum of 10 days. Mice were provided with ad libitum access to water and standard rodent chow during all testing with the exception of caloric restriction studies.
Indirect Calorimetry and Activity Monitoring
The apparatus and procedures used for determination of metabolic rate and locomotor activity have been described previously (57). Metabolic cages were placed within environmental chambers that provided computer-controlled Ta and lighting conditions for four individual cages (37). Individually metabolic cage Ta was monitored by a thermister mounted on the inside of the polycarbonate cage lid. Briefly, a computer controls individual cage heaters and the cooling compressor of the environmental chamber to achieve the desired individual cage Ta. Oxygen consumption (V̇o2) and carbon dioxide production (V̇co2) were measured every 2.5 min by open circuit respirometry using a modification of the approach described by Bartholomew et al. (5) to isolate successive samples. The flow rate of air into the chamber was set at 1 l/min. V̇o2 was calculated every 4 min. Locomotor activity was measured using a custom-designed force platform as described previously (57). Total activity, measured in meters, was accumulated in 30-s periods and stored with a 1-cm resolution.
The radiotelemetry signal was continuously detected (1,000-Hz sample rate) and either analyzed with Data Sciences analysis software, with HR determined by R-R intervals, or were routed to a pulse detection unit connected to a computer. Custom-written software determined arterial blood pressure, and the time between diastolic nadirs, measured to the nearest millisecond, was used to determine the interbeat interval and HR. Both methods of HR assessment result in equivalent measurements of HR.
To assess cardiovascular variability in the time domain, the standard deviation of the interbeat interval (SDNN) was calculated from home cage recordings obtained throughout the light phase of the circadian cycle. The light phase was selected because it coincides with a period of generally low locomotor activity and is consistent with the time period during which analysis of cardiovascular variability is typically performed in mice (19, 26, 30). Light-phase SDNN was determined from a minimum of 10 measurements/h calculated over intervals of 5–30 s each hour. The reported value for total light-cycle SDNN represents an average of at least 110 determinations per animal.
IHR was measured by either intraperitoneal injection of a cocktail containing metoprolol (10 mg/kg; β1-receptor blocker) and atropine (4 mg/kg; muscarinic receptor blocker), or with the ganglionic blocker, chlorisondamine (2.5 mg/kg). The IHR was taken as the minimum HR over the next 30–60 min. Both of these methods have been used to determine IHR, and each method results in an identical IHR measurement (S. J. Swoap, personal observations).
Experiment 1: cardiovascular responses of B6 mice to rapid changes in Ta.
Studies were performed using B6 mice instrumented with arterial catheters and housed in indirect calorimeters to allow assessment of mean arterial pressure (MAP), HR, V̇o2, and locomotor activity. After mice completely recovered from surgery and acclimated to initial housing conditions (Ta = 30°C), a rapid temperature ramp protocol was initiated without handling or disturbing the mice. The temperature changes were performed during the middle of the light phase and were completed in 1.5 to 2 h. The first ramp consisted of cooling from 30° to 20°C. The Ta was kept at this level for 22 h, and the ramp was reversed by warming back to 30°C.
Experiment 2: influence of Ta on resting HR and IHR in control, β-less, and M2R mice.
All mice were housed at 23°C for 4 days. Resting HR was derived from a 10-h average of light-cycle data obtained from the final day of housing at 23°C. IHR was then measured, using either double autonomic blockade or ganglionic blockade, 4 h before the onset of the dark cycle of the fifth day. Mice were then housed at thermoneutrality (Ta = 30°C) for 4 days. Light-cycle HR was recorded, and IHR was again measured.
Experiment 3: influence of caloric restriction on resting HR and IHR.
To determine the relative contribution of decreased sympathetic and increased vagal tone to caloric restriction-induced bradycardia in mice, we measured the HR, IHR, and SDNN in the same groups of mice before and after caloric restriction (consuming 70% of ad libitum food intake) at 30°C. Baseline caloric intake was measured during the 4 days at Ta = 30°C. Mice were then calorically restricted to 70% of ad libitum food intake for 10 days. A final measurement of light-cycle HR and IHR was obtained 4 h before the onset of the dark cycle.
Data Analysis and Statistics
The physiological parameters measured of the littermate controls for the β-less mice were not significantly different from those measured from the littermate controls of the M2R−/− mice (both control groups were B6 mice). Hence, both sets of control mice were pooled into a single control group. All results are reported as means ± SE. Differences in baseline variables between knockout mice and wild-type controls were assessed by Student's t-tests. The effects of increased Ta and subsequent caloric restriction were statistically assessed by repeated-measures ANOVA. Tukey post hoc tests were used to determine significant differences between means. Significance levels of P < 0.05 were accepted.
Experiment 1: Mouse HR and MAP Exhibit Rapid Response to Changing Ta
In the first phase of this experiment, the Ta within the metabolic cage was cooled from 30° to 20°C in 1 h. Figure 1 shows that HR (Fig. 1A), V̇o2 (Fig. 1C), and MAP (Fig. 1E) responded rapidly to decreasing Ta. Within 90 min, HR increased from 300 to 600 beats/min, V̇o2 increased from 0.75 to 1.5 ml/min, and MAP increased from 75 to 105 mmHg. These responses were generally independent of locomotor activity (Fig. 1G). The elevated levels of HR, V̇o2, and MAP achieved promptly after cooling were sustained for the subsequent 20 h as evident in Figs. 1, B, D, and F. In the second phase of this experiment, Ta within the metabolic cage was warmed from 20° to 30°C in 2 h. HR, V̇o2, and MAP rapidly decreased in these mice as the Ta was warmed from 20° to 30°C, with those changes being independent of locomotor activity (Fig. 1H).
Experiment 2: Vagal Tone Suppresses HR Below Intrinsic Levels at Thermoneutrality
Figure 2A depicts HR over dark and light cycles at a Ta of 23°C in control B6 mice, mice lacking sympathetic influence to the heart (β-less), and in mice lacking parasympathetic influence to the heart (M2R−/−). At this cool Ta, control mice exhibited the expected high light-cycle HR of 604 ± 10 beats/min. M2R−/− mice also displayed a high light-cycle HR at this same Ta (602 ± 5 beats/min). However, β-less mice had significantly lower light-cycle HR (442 ± 14 beats/min), consistent with other reports of lower resting HR in mice lacking β-receptors (14). Dark-cycle and 24-h HR averages were also significantly lower in β-less mice at 23°C (data not shown). When Ta was raised to 30°C, the light-cycle HR fell significantly in all groups (control mice, 351 ± 11; β-less, 363 ± 10; and M2R−/−, 416 ± 7 beats/min). The light-cycle HR of β-less mice was identical to that of control mice (P > 0.05). However, the light-cycle HR in M2R−/− mice was significantly elevated (Fig. 2B). These data suggest that β-adrenergic receptor signaling is required at cool housing temperatures (e.g., 23°C) to achieve a normal light-cycle HR, whereas muscarinic receptor signaling is required at 30°C to achieve a normal light-cycle HR.
We also examined IHR after housing mice for 4 days at either 23° or 30°C to determine whether there was a fundamental shift in autonomic balance of HR control at thermoneutrality. A typical HR tracing for a B6 control mouse with IHR measurement is shown in Fig. 3A. When Ta was raised from 23° to 30°C, the IHR of control mice fell significantly from 486 ± 14 to 403 ± 7 beats/min (Fig. 3B). As Fig. 3B shows, the IHR of M2R−/− mice was significantly lower than that of control mice at 23°C (436 ± 2 beats/min). Similar to that in control mice, IHR in the M2R−/− mice fell significantly with exposure to thermoneutrality (377 ± 5 beats/min). The percent drop in IHR with the increase in Ta was not significantly different between control and M2R−/− mice (17 ± 3% vs. 14 ± 2%, respectively). The IHR of β-less mice at 30°C was 418 ± 9 beats/min, which was not significantly different from that in control mice (Fig. 3B). However, when we attempted to measure IHR in β-less mice at 23°C in a pilot study, we were unable to assess an IHR at this cool temperature because the HR of these mice did not stabilize after ganglionic blockade. These mice experienced continuous and progressive bradycardia to well below 300 beats/min. In fact, these mice are cold sensitive (28), and we presume the mice became hypothermic due to a disruption of sympathetic activity to brown adipose tissue.
The difference between resting light-cycle HR and IHR reflects the autonomic control of HR and, therefore, the balance between sympathetic and parasympathetic control of the heart. As Fig. 3C shows, light-cycle HR at 30°C in control and β-less mice was well below IHR (∼40 and 70 beats/min, respectively). In contrast to the relationship between HR and IHR in control mice, the light-cycle HR of M2R−/− mice was well above IHR (∼40 beats/min; Fig. 3C). In fact, we never observed a HR at any time in these knockout mice that was below IHR.
Since assessment of HR variability in the time domain may provide insight into cardiac autonomic activity, we calculated SDNN in response to increasing Ta in all mice. As reported previously (48, 54), increasing Ta toward thermoneutrality increased SDNN in control mice (Fig. 4). SDNN in β-less mice was not significantly different from SDNN in control mice at either 23° or 30°C (Fig. 4). In contrast, the SDNN of M2R−/− mice was significantly lower than control mice at 23°C and did not rise in response to elevated Ta.
Experiment 3: IHR is Decreased by Caloric Restriction
Caloric restriction resulted in an ∼10% decrease in body weight and similar magnitude of light-cycle bradycardia (∼75 beats/min) in all groups (Fig. 5A). Decreased caloric intake in control mice increased SDNN in control mice (Fig. 5B), as seen previously (54). This increase in HR variability was significantly blunted in β-less mice and completely absent in M2R−/− mice (Fig. 5B). Furthermore, we observed that caloric restriction-reduced IHR by ∼30–40 beats/min in all groups of mice (Fig. 5A).
The results support new and important concepts in our understanding of HR control in mice. We have previously demonstrated that Ta is an important determinant of murine HR and blood pressure (48, 56). In this study, we report for the first time concurrent measures of metabolism, HR, and blood pressure during rapid shifts in Ta. The findings provide further strong support for the concept that housing conditions of mice are an important determinant of cardiac sympathetic and parasympathetic outflow, which in turn has a major impact on HR and HR variability. In addition, novel findings from mutant mice lacking M2R and β-receptors convincingly demonstrate that cardiac vagal tone dominates autonomic control of resting HR in mice studied at thermoneutrality. Finally, we provide new data demonstrating that both Ta and caloric balance modulate IHR in mice. The finding adds to a growing list of factors that can modulate IHR.
Ta and Cardiovascular Physiology in Mice
It is generally accepted that normal mice have a resting HR of about 550- 600 beats/min, exhibit minimal vagal tone, and have an IHR that is below resting HR (19, 27). Our new findings support previously published work and convincingly demonstrate that these descriptors of the mouse physiological state are strongly influenced by Ta. The physiological demands of thermoregulation are substantial in mice studied at a Ta comfortable for humans. While housing factors such as type of bedding and number of mice per cage are important considerations (21), we and others have clearly documented the metabolic and cardiovascular consequences of the mild cold stress that exist under standard laboratory conditions (8, 48, 51, 52, 56). Nonshivering thermogenesis is activated at temperatures cooler than the lower critical limit of the thermoneutral zone, which is about 30°C for mice (22). Sympathetic control of brown adipose tissue metabolic activity is an important mechanism of heat generation in small rodents (see Ref. 42 for review). Indeed, V̇o2 rapidly and promptly changes inversely with Ta between 20–30°C in a range from 0.75–1.5 ml/min.
We also report for the first time the rapid responsiveness of mouse blood pressure and HR to changes in Ta. In C57 mice, we find that MAP decreases from about 100 to about 80 mmHg when Ta is warmed from 20–30°C. The finding reinforces the need to monitor and control for Ta during blood pressure measurements in rodents (32). Not surprisingly, HR is directly related to metabolic rate during changes in Ta. However, the sensitivity of mouse HR to small changes in Ta is striking. The magnitude of change in HR is 250–300 beats/min between the Ta of 20–30°C. The HR at thermoneutrality (∼350 beats/min) is similar to values reported previously (56). These HR responses are clearly not due to alterations in locomotor activity (Fig. 1, G and H) and thus are likely regulated by the autonomic nervous system.
Cardiac Autonomic Activity in Mice
To further analyze the autonomic mechanisms regulating HR, we measured HR, IHR, and HR variability in the time domain at multiple Ta values and during caloric restriction in β-less mice and in mice lacking the M2R. β-less mice are known to exhibit cold intolerance (28) and increased susceptibility to diet-induced obesity (2), indicating impaired sympathetically mediated facultative thermogenesis. At standard housing temperatures, β-less mice exhibit a 10–15% reduction in metabolic rate and reduced body temperature but have normal thyroid hormone levels and responsiveness to hyperthyroidism (2, 3, 28). Although these mice tolerate standard laboratory conditions well, the possibility of unrecognized compensatory adaptations that contribute to the phenotype in unconditional knockout mice must be acknowledged (53).
Mice lacking the β1-receptor have markedly impaired exercise- and isoproterenol-induced elevation in HR (38, 39) but were reported to have normal resting HR (Ta was not reported in these studies). More recent findings (14) are consistent with our observations (Fig. 2) of significantly lower resting HR in β-less mice studied at Ta = 23°C. Furthermore, brief, noninvasive ECG recordings indicated that that β-less mice have a HR of ∼150 beats/min lower than wild-type mice (3), although HR levels obtained with the noninvasive ECG method (550 beats/min) were about 100 beats/min greater than we observed using telemetry (450 beats/min, Fig. 2). The magnitude of tachycardia associated with decreasing Ta below thermoneutrality is markedly diminished in β-less (Fig. 2), suggesting that increased sympathetic activity produces the elevated HR observed in at Ta = 23°C. Nonetheless, β-less mice still exhibit significant bradycardia upon exposure to thermoneutrality. The finding suggests that increased vagal tone or decreased IHR are also involved in the thermoneutrality-induced slowing of HR.
The bradycardia induced by vagal efferent stimulation or by methacholine administration is mediated by the M2 receptor (17). We found that M2R knockout mice have a normal resting HR at Ta = 23°C (Fig. 2) yet exhibit minimal HR variability (Fig. 4). This is consistent with the prevailing view that the major determinant of beat-to-beat variability in HR is vagal activity (16, 19). At Ta = 30°C, we observed that in contrast to control and β-less mice, the IHR of M2R mice was always less than resting HR (Fig. 3). This is clear evidence that control of HR in mice when housed near thermoneutrality requires the M2 receptor, and thus, HR at thermoneutrality is mediated predominantly by vagal tone. In control and β-less mice, we observed an increase in HR variability with exposure to thermoneutrality (Fig. 4). In contrast, M2R−/− mice failed to increase HR variability with either elevated Ta (Fig. 4) or with caloric restriction (Fig. 5). The interpretation of these findings is somewhat limited by a lack of adequate sampling to perform variability analysis in the frequency domain. However, even with this information, it is recognized that caution is required when extrapolating from measures of HR variability to activity of the autonomic nervous system (45, 46).
The bradycardia of negative energy balance is observed in both humans (1, 9, 31) and mice (47, 49, 56) undergoing weight loss. Studies using pharmacological blockade and analysis of HR variability suggest that the primary mechanism for caloric restriction-induced bradycardia is increased vagal tone (1). In this study, caloric restriction reduced HR in all mice but did so using different mechanisms (Fig. 5). In control and β-less mice, increased vagal tone (inferred from the increase in SDNN) and reduced IHR were engaged (Fig. 5). Caloric restriction-induced bradycardia in M2R−/− mice was not accompanied by an increase in SDNN but was accompanied by a decrease in IHR. Collectively, these data suggest that there are two major components to bradycardia associated with caloric restriction: 1) a depressed IHR and 2) an elevation of vagal activity at the heart. The potential mechanisms of the decrease in IHR with caloric restriction are discussed below.
IHR is Influenced by Many Factors
An additional major concept we provide here is that IHR is heavily influenced by Ta and caloric availability. IHR is commonly determined by pharmacological blockade of sympathetic and parasympathetic nerve activity and thus presumably represents the spontaneous depolarization rate of sinoatrial myocytes. Regulation of the hyperpolarization-activated cyclic nucleotide-gated channel (HCN), which is responsible for the diastolic, depolarizing “funny” current, is clearly one mechanism by which IHR could be modulated. Interestingly, the HCN channel gene expression is known to be depressed in mice with depressed thyroid signaling and bradycardia (20). Clearly, thyroid hormone activity is regulated by nutritional status (42), thus this is one mechanism by which nutritional status could regulate IHR. However, it must be acknowledged that the ultimate pacemaker activity is determined by multiple ionic currents (41) and that much remains to be learned concerning the mechanisms by which IHR is determined (12).
Previously, only a few interventions have been shown to alter IHR, including nucleus ambiguus lesions or renal hypertension (34) that may be mediated by the actions of angiotensin II on IHR (7, 11). We show here that IHR fell by 20% with a simple elevation of Ta by 7°C. This fall in IHR accounts for a substantial portion of the drop in resting HR associated with housing at thermoneutrality. In addition, we provide evidence that reductions in IHR may be responsible for about 50% of the bradycardia associated with caloric restriction. Several hormones known to be altered by energy balance, including thyroid hormone (18, 29, 40), insulin (25), and glucagon-like peptide 1 (4), influence IHR. In addition, Sowden et al. (44) recently demonstrated that an acute administration of the anorexigenic hormone oxyntomodulin can dramatically increase IHR. Hence, it is likely that the circulating status of these hormones, as well as other hormones that reflect energy status (e.g., leptin or ghrelin), plays a major role in the reduction of IHR during a bout of caloric restriction. The relative contribution of each of these hormones remains to be determined. Taken together, our findings indicate that, in addition to vagal and sympathetic tone, IHR is an important component of the adaptive response to altered Ta and reduced energy availability and should be considered as an important regulatory arm of HR control.
These experiments demonstrate that mice exhibit sensitive metabolic and cardiovascular responses to variations in Ta below the thermoneutral zone. Furthermore, our findings indicate that when studied at thermoneutrality, cardiac autonomic function of mice more closely resembles that of humans in that both species exhibit dominant vagal tone and subsequent resting HR levels below IHR. In contrast, in the mild cold stress of standard laboratory conditions, mice exhibit tachycardia due to a withdrawal of cardiac vagal tone, increased sympathetic tone, low HR variability, and elevated IHR. The findings further emphasize the need for careful attention to controlling Ta during the assessment of cardiovascular physiology in mice.
This work was supported by National Heart, Lung, and Blood Institute Grants R15-HL-081101-01 (to S. J. Swoap) and R01-HL-56732 (to J. M. Overton).
We gratefully acknowledge the technical assistance of Charles Badland, Molly Gutilla, and Ashley Christman.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked ““advertisement”” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 by the American Physiological Society