Evolved changes in reflex control of the cardiovascular system in deer mice native to high altitude Open Access

Cardiovascular function is imperative for the appropriate transport of respiratory gases, nutrients and waste products throughout the body across a broad range of environmental and metabolic conditions. Two vital mechanisms of cardiovascular control in vertebrates are the arterial baroreflex and the hypoxic chemoreflex. The arterial baroreflex exhibits negative feedback control of arterial blood pressure by adjusting cardiac output and systemic vascular resistance. This baroreflex control of arterial blood pressure is vital for maintaining tissue perfusion pressure and O₂ delivery in a range of conditions. The hypoxic chemoreflex controls the response to reductions in arterial O₂ tension, including adjustments in tissue blood flow through changes in cardiac output and systemic vascular resistance. These reflexes act and intersect via efferent pathways involving the autonomic nervous system. Although much is known about how these reflexes work, less is known about how they may by modulated by environmental acclimation/acclimatization or evolutionary (genetically based) adaptation to help overcome physiological challenges.

Studies of high-altitude natives present a valuable opportunity to examine this issue. The cold and hypoxic conditions at high altitude require that endotherms maintain high metabolic rates in air with low O₂ availability, which imposes high demands for circulatory O₂ transport (McClelland and Scott, 2019; Scott et al., 2024; Storz and Cheviron, 2021). Cardiovascular function is essential for overcoming these challenges, but the plastic and evolved adjustments in cardiovascular function in high-altitude natives remain poorly understood in most taxa. Emerging evidence suggests that the baroreflex can be modified by acclimatization and/or adaptation to high-altitude hypoxia in humans. During acute hypoxia, systemic tissues faced with O₂ limitation produce local factors that tend to cause vasodilation and reduce peripheral resistance (Kulandavelu et al., 2015; Premont et al., 2020; Weisbrod et al., 2001). Baroreflex stimulation of cardiac output via sympathetic activation is then critical for maintaining arterial blood pressure and meeting the increased demand of tissues for blood flow (Bernardi, 2007). However, chronic activation of the baroreflex can impair baroreflex sensitivity to additional modulators of blood pressure. Indeed, humans native to low altitude exhibit reduced baroreflex sensitivity when exposed to chronic hypoxia at high altitude (Alvarez-Araos et al., 2024; Bourdillon et al., 2017, 2018; Fisher et al., 2022; Steinback et al., 2009) and reduced sympathetic neurovascular transduction (Berthelsen et al., 2020; Simpson et al., 2019). Similarly, in domestic rats, acute hypoxia results in sympathoexcitation and parasympathetic withdrawal and thus impairs baroreflex control (Beltrán et al., 2020). In contrast, Sherpa adapted to high altitudes have greater baroreflex sensitivity in hypoxia than lowlanders, in association with lower baseline sympathetic activity along with greater neurovascular transduction and/or α-adrenergic vasoconstrictive reserve (Bernardi et al., 1998; Berthelsen et al., 2020; Simpson et al., 2019, 2021). Whether similar evolved changes are exhibited in adults of other animal species at high altitude is poorly understood. This is particularly true of small mammals, for which the demands on the circulatory system are expected to be greater than for larger endotherms due to the increased metabolic costs of thermogenesis to cope with the cold temperatures at high altitude.

It also remains unclear whether small high-altitude mammals differ in the overall cardiovascular response to hypoxia, which results from the combined effects of the baroreflex, hypoxic chemoreflex and local factors in peripheral tissues, along with centrally controlled adjustments in metabolism and body temperature. However, important differences in the cardiovascular response to hypoxia between high- and low-altitude taxa have been previously reported in birds. In the low-altitude barnacle goose (Branta leucopsis), acute hypoxia reduces peripheral vascular resistance (likely resulting from vasodilation), which is only partially offset by increases in cardiac output such that arterial blood pressure declines (Lague et al., 2016). In contrast, in high-altitude bar-headed geese (Anser indicus) and Andean geese (Chloephaga melanoptera), reductions in peripheral resistance in acute hypoxia are associated with sufficient increases in cardiac output that arterial pressure is well maintained (Lague et al., 2017). Interestingly, systemic vasodilation in response to acute hypoxia appears to be attenuated over more prolonged exposures to chronic hypoxia in both rats and humans, in concert with sympathetic overactivity and hypertension (Hansen and Sander, 2003; Huang et al., 2002; Lundby et al., 2018; Narvaez-Guerra et al., 2018; Simpson et al., 2024). Whether high-altitude mammals exhibit similar responses to chronic hypoxia as those of their low-altitude counterparts remains poorly understood.

In this study, we address these knowledge gaps by examining reflex control of cardiovascular function in the deer mouse (Peromyscus maniculatus), the species with the broadest altitudinal range of any North American mammal. High-altitude populations of this species maintain higher field metabolic rates than their low-altitude counterparts (Hayes, 1989a,b), likely because of the increased costs of thermogenesis in the cold environment at high altitude. Correspondingly, aerobic capacity for thermogenesis is increased in high-altitude deer mice by plastic and evolved adjustments across the O₂ transport pathway and in thermogenic tissues (Cheviron et al., 2013; Dane et al., 2018; Lui et al., 2015; Rezende et al., 2004; Shirkey and Hammond, 2014; Tate et al., 2017, 2020; Van Sant and Hammond, 2008; Velotta et al., 2016). Key to this increase in thermogenic capacity is corresponding variation in cardiac output and vascular O₂ transport (Scott et al., 2024; Tate et al., 2020), but potential differences in cardiovascular control remain poorly understood. We previously reported that high-altitude deer mice maintain the capacity for adrenergic vasoconstriction in chronic hypoxia, in contrast to white-footed mice (P. leucopus; a congeneric found exclusively at low altitude), in which the vascular responsiveness to α-adrenergic stimulation is blunted (Wearing et al., 2022). These findings provide indirect evidence that there may be evolved changes in baroreflex control in high-altitude mice, but direct tests of baroreflex function were not performed. Furthermore, previous measurements were made in mice held at warm temperatures, at which metabolic demands are low, and did not examine how cardiovascular function and control may be adjusted by chronic exposure to cold combined with hypoxia. In the present study, we tested the hypothesis that reflex control of cardiovascular function is amplified in high-altitude deer mice to better maintain systemic blood pressure and circulatory O₂ transport in cold hypoxia.

Experiments were conducted on deer mice, Peromyscus maniculatus (Wagner 1845), raised in captive breeding colonies that we have used in past research (Wearing and Scott, 2022a). The colonies are established from wild mice that are live-trapped at low altitude in the Great Plains (Buffalo County, NE, USA, 40°41′58″N, 99°04′53″W; approximately. 660 m above sea level) or at high altitude near the summit of Mount Blue Sky (formerly ‘Mount Evans’, Clear Creek County, CO, USA, 39°35′18″N, 105°38′38″W; 4350 m above sea level). Wild mice are transported to McMaster University and bred in captivity within their respective population to produce lab-born progeny, which are generally bred and raised in common laboratory conditions (25°C, approximately 20 kPa O₂). The experiments here used second-generation progeny of 6–10 months of age from nine families of low-altitude mice and 10 families of high-altitude mice (often referred to as ‘lowlanders’ and ‘highlanders’, respectively). Male and female mice from each population were divided at random into two acclimation groups that were exposed to the following environmental conditions for 6–8 weeks: (1) warm normoxia (25°C, ∼20 kPa O₂), or (2) cold hypoxia (5°C, 12 kPa O₂) simulating ambient O₂ levels and winter burrow temperatures at high altitude (Hayward, 1965). Cold hypoxic conditions were maintained using custom-made hypobaric chambers (Lui et al., 2015; McClelland et al., 1998) housed in a temperature-controlled room. All mice were provided with unlimited access to water and standard rodent chow (Teklad 22/5 Rodent Diet formula 8640; Envigo). Cages were cleaned twice weekly, which required that cold hypoxic mice be briefly returned to normobaria (<20 min). All animal procedures followed guidelines established by the Canadian Council on Animal Care and were approved by the McMaster University Animal Research Ethics Board (Animal Use Protocol 20-01-02).

We measured the reflex control of cardiovascular function in mice for which we have previously reported routine body temperature and cardiovascular function throughout the diel cycle (Wearing and Scott, 2022a). Following 6 weeks in the respective acclimation condition, each animal was surgically instrumented with a physiological telemeter (HD-X11; Data Sciences International, St Paul, MN, USA) to measure arterial blood pressure (in the ascending aortic arch, via the left common carotid artery), heart rate (f_H) and body temperature (T_b). Isoflurane anaesthesia, telemeter placement in the interscapular subcutaneous space, post-surgical pain management (carprofen) and 7-day recovery were conducted as previously described in full detail (Wearing and Scott, 2022a,b). Normal circadian oscillations of mean arterial blood pressure (P_mean), f_H and T_b in awake and freely behaving mice became stable after 3–4 days post-surgery. Routine measurements that have been previously reported (Wearing and Scott, 2022a) were then conducted over the next 2–3 days. We then made the two series of in vivo measurements reported here: (i) cardiovascular responses to progressive stepwise hypoxia in awake mice; then (ii) pharmacological assessment of baroreflex function in anaesthetized mice.

Responses to progressive stepwise hypoxia were made at 25°C using methods that we have previously described in detail in a study of CD-1 strain domestic house mice (Wearing and Scott, 2022b). Measurements were made in each of the four experimental groups of deer mice: warm normoxic lowlanders, n=13 (6 males, 7 females; body mass=25.5±1.6 g mean±s.e.m.); cold hypoxic lowlanders, n=12 (5 males, 7 females; 23.6±1.1 g); warm normoxic highlanders, n=9 (5 males, 4 females; 21.7±0.6 g); and cold hypoxic highlanders, n=12 (6 males, 6 females; 23.3±0.7 g). Briefly, mice were weighed and placed in an open-flow respirometry chamber (530 ml) supplied with 600 ml min⁻¹ normoxic air (21% O₂, balance N₂). Mice were left for 20–60 min until calm and physiological measurements became stable. The chamber then continued to be supplied with normoxic air for another 20-min measurement period, after which the incurrent O₂ partial pressure (P_O2) was reduced stepwise to 16, 12, 10, 9 and 8 kPa O₂, with a period of 20 min at each P_O2 step. The first 10 min at each step allowed the mouse to adjust to the new P_O2, and steady-state resting measurements were then recorded for analysis during the last 10 min of each step. Occasional brief periods of heightened activity during this recording window were excluded from the selections used for analysis. Incurrent chamber flow rate, incurrent and excurrent O₂ and CO₂ fractions, arterial blood pressure and subcutaneous temperature (a proxy for T_b) were acquired at 1 kHz using a PowerLab 16/32 data acquisition unit and Labchart 8 Pro software (ADInstruments, Colorado Springs, CO, USA) with PhysioTel Connect (ADInstruments) for acquisition of telemetry data via an MX2 data acquisition system (Data Sciences International). Oxygen consumption rate (V̇_O2) was calculated at standard temperature and pressure (STP) using established equations for incurrent flow measurement (Lighton, 2018), and is expressed relative to body mass. P_mean and f_H were calculated from the systolic–diastolic oscillations in arterial pressure.

Measurements of baroreflex function were then conducted the day after measuring cardiovascular responses to progressive stepwise hypoxia. Mice were anaesthetized to a surgical plane using 2–3% isoflurane (delivered in 100% O₂), and the right external jugular vein was catheterized with a 5-cm length of saline-filled heparinized microrenathane tubing (MRE025, Braintree Scientific, Braintree, MA, USA) for direct intravenous administration of pharmacological agents (as described below). Once surgery was completed, blood pressure and heart rate data were recorded via the telemeter, and the isoflurane dose was held at the minimum level at which there was a <10-mmHg P_mean response to toe pinch (1.5–2.0% isoflurane in O₂). This low dose of isoflurane is widely considered the best choice of anaesthesia for studies measuring cardiovascular function and regulation in rodents (Janssen et al., 2004; Lorenz, 2002; Wearing et al., 2025), minimizing the potential influence of anaesthetics on cardiovascular function and control, while avoiding the confounding effects of variation in animal activity and metabolism. At this isoflurane level, P_mean and f_H remained stable in most individuals at levels resembling resting conscious values, suggesting that autonomic control of arterial pressure remained functionally intact across experimental groups. However, values did not stabilize in a small subset of animals that were thus excluded from the final dataset, such that the final sample sizes for each group were: warm normoxic lowlanders, n=11 (5 males, 6 females); cold hypoxic lowlanders, n=10 (4 males, 6 females); warm normoxic highlanders, n=8 (4 males, 4 females); and cold hypoxic highlanders, n=8 (3 males, 5 females).

Baroreflex function was assessed by measuring the changes in f_H and P_mean after intravenous injections of phenylephrine (an α-adrenergic vasopressor that increases P_mean and causes a baroreflexive decrease in f_H) followed by sodium nitroprusside (SNP; a vasodilatory nitric oxide donor that decreases P_mean and causes a baroreflexive increase in f_H). Baseline P_mean and f_H data were collected for 5 min to ensure stable measures. Serial doses of phenylephrine (phenylephrine hydrochloride; Sigma-Aldrich, Oakville, ON, Canada) of ascending concentrations (0.15, 1.5, 15 and 150 µg ml⁻¹) were then injected intravenously in 3.33-µl injections per gram of animal mass, such that mice received 0.5, 5, 50 and 500 µg kg⁻¹ doses of phenylephrine in series, with 1- to 2-min washout periods between doses before P_mean and f_H returned to pre-injection values. Similar increasing concentrations (0.3, 3, 30 and 300 µg ml⁻¹) of SNP (Sigma-Aldrich) were subsequently injected at 3.33 µl g⁻¹ to final doses of 1, 10, 100 and finally 1000 µg kg⁻¹ (with 3- to 5-min washout periods between each dose). The highest doses of each pharmacological agent were generally found to be saturating, as reflected by a lack of response compared to the second highest dose. Mice were then euthanized by cervical dislocation, weighed and dissected for organ mass measurements that have been reported previously (Wearing and Scott, 2022a).

Here, we report G₅₀, the overall maximum and minimum f_H and P_mean values obtained at the highest doses of phenylephrine and SNP, and the difference between these overall maximum and minimum values (Δf_H and ΔP_mean) for each individual mouse.

We also determined an average baroreflex curve for each experimental group to visually compare mean baroreflex function between groups. This was accomplished using the baroreflex curve for each individual to calculate its curve-predicted f_H values across the full P_mean range from 0 to 200 mmHg, and then fitting an asymmetrical baroreflex curve to the entire set of resulting data from all the mice in the group. These average baroreflex curves as well as the baroreflex curves for each individual mouse are shown in Fig. S1.

Statistical comparisons were made using linear mixed models with the lme4 package (Bates et al., 2015) in R v.4.4.1 (https://www.r-project.org/). Data from males and females were included in our analyses, with both sex and family included as random factors in all linear models. For baroreflex data, we tested for the main effects of population (lowland versus highland) and acclimation environment (warm normoxia versus cold hypoxia) as well as the interactions between these factors. Holm-adjusted Tukey's HSD post hoc tests were performed to test for the pairwise differences between acclimation environments within each population, or between populations within each acclimation environment. For the responses to progressive stepwise hypoxia, we tested for effects of and interactions between population, acclimation environment and inspired P_O2. Body mass was included as a covariate when P<0.10 (or excluded in final models if P>0.10 in initial tests). For each population, Holm-adjusted Tukey's HSD post hoc tests were performed to test for the pairwise differences between acclimation environments within each P_O2, and between each P_O2 versus the initial value at 21 kPa within each acclimation group. For wire myography experiments, data from males and females were combined, and two-tailed unpaired t-tests with Welch's correction were performed to test for effects of population on equation constants. Absolute values were used for these analyses, with the exception of EC₅₀ data, which were log₁₀ transformed before analysis (though we report absolute values). Extra sum-of-squares F-tests were also performed to compare dose–response curves between populations using an α value of 0.05. Statistical analyses were generally carried out on absolute values of traits that were not corrected for body mass (because effects of body mass were accounted for as a covariate in statistical models when P<0.10), but the V̇_O2 data presented here are expressed relative to body mass as is conventional in the literature. Data are presented as individual values, sometimes along with box and whisker plots, or as means±s.e.m.

We measured the responses to progressive stepwise hypoxia to consider blood pressure maintenance and heart rate responses under the combined influence of the baroreflex, hypoxic chemoreflex, local factors in peripheral tissues and changes in metabolism (Fig. 1). In general, exposure to acute hypoxia had a strong effect on all measured traits (main effects of P_O2, P<0.001; Table 1). Mean arterial pressure (P_mean; Fig. 1A) was maintained and heart rate (f_H; Fig. 1B) tended to increase in mild to moderate hypoxia, but P_mean declined in severe hypoxia as f_H returned towards baseline values. The declines in P_mean in severe hypoxia were often associated with reductions in oxygen consumption rate (V̇_O2; Fig. 1C) and body temperature (T_b; Fig. 1D).

Acclimation to cold hypoxia attenuated hypotension in severe hypoxia, in contrast to the pronounced hypotension observed in warm normoxic mice, but this effect seemed to be driven primarily by variation in highlanders. There was a significant effect of acclimation on P_mean (acclimation effect, P=0.003) along with a significant acclimation×P_O2 interaction (P<0.001). This was associated with cold hypoxic highlanders exhibiting significantly greater P_mean than warm normoxic highlanders when P_O2 was ≤10 kPa, such that P_mean was maintained at normoxic levels (i.e. values at 21 kPa O₂) in all but the most severe level of hypoxia (Fig. 1A). In contrast, there were no significant effects of cold hypoxia acclimation on P_mean in lowlanders. The improved maintenance of P_mean in highlanders after acclimation was associated with a pronounced increase in f_H in severe hypoxia (Fig. 1B). This response likely drove the significant acclimation×P_O2 interaction on this trait (P<0.001), because there was no effect of acclimation on f_H in lowlanders. Acclimation to cold hypoxia also appeared to reduce the depression of V̇_O2 and T_b in severe hypoxia (Fig. 1C,D). Differences in the effects of acclimation between highlanders and lowlanders were less apparent for these traits, although only cold hypoxic highlanders maintained V̇_O2 at normoxic levels at 10 kPa O₂. There was also a significant population effect on T_b (P=0.010), which was lower in highlanders than in lowlanders in normoxia and moderate hypoxia.

We measured the f_H and P_mean responses to phenylephrine followed by SNP in anaesthetized mice to uncover potential evolved and/or environmentally plastic changes in baroreflex function. We measured the maximum and minimum P_mean and f_­H values recorded throughout the baroreflex assessments, and used these to calculate the scope of change (Δ) in each of these indices of cardiovascular function (ΔP_mean and Δf_H, respectively). Population differences were greatest after saturating phenylephrine administration, when highlanders exhibited lower values of maximum P_mean (population effect, P=0.003; Table 2) and minimum f_H (P=0.041). Minimum P_mean and maximum f_H elicited by saturating SNP were similar between populations (P=0.738 and P=0.940, respectively). As such, the difference between maximum and minimum values of P_mean (ΔP_mean) was 17–21% lower on average in highlanders versus lowlanders (population effect, P=0.001). The population effect on the difference between maximum and minimum values of f_H (Δf_H) neared significance (P=0.071) owing to 2.1-fold higher average values in highlanders versus lowlanders in warm normoxia.

Consideration of the combined changes in P_mean and f_H suggested that baroreflex sensitivity was greater in highland mice (Fig. 2). There was an upward and leftward shift in the relationship between Δf_H and ΔP_mean in highlanders compared with lowlanders, reflecting an exaggerated f_H response and reduced change in pressure in response to a baroreflex disturbance (Fig. 2A). To visualize the full baroreflex-dependent relationship between blood pressure and heart rate, we generated baroreflex curves using absolute P_mean and f_H data recorded throughout the pharmacological baroreflex assessment for each individual (Fig. S1). Representative curves for a high-altitude mouse and a low-altitude mouse are shown in Fig. 2B. Consistent with the aforementioned population differences in Δf_H and ΔP_mean, these baroreflex curves show a greater and steeper f_H response to changes in P_mean in the highlander compared with the lowlander (Fig. 2B). From these curves, baroreflex gain (G₅₀) was calculated as the slope of the curve at the midpoint of the P_mean response range (i.e. dotted lines in Fig. 2B). This index of baroreflex sensitivity was greater in magnitude (i.e. more negative) in highlanders than in lowlanders (population main effect, P=0.049; Fig. 2C), again indicating that highland mice exhibit greater baroreflex sensitivity than lowland mice.

We examined vasomotor sensitivity to α₁-adrenoceptor stimulation as a potential mechanism for population differences in baroreflex function. We focused on population comparisons in warm normoxic mice because population effects were the dominant source of variation in baroreflex function (Fig. 2, Table 2). Isolated mesenteric arteries contracted in response to phenylephrine in a dose-dependent manner (Fig. 3A). We also examined endothelium-dependent relaxations in response to acetylcholine (Fig. 3B). We found that the responses to each of these drugs were very similar between highlanders and lowlanders, with a single curve for each drug being sufficient to describe the dose response of both populations (see black curves in Fig. 3). As a result, we also saw no population differences in either the maximal response (F_max or F_min) or the effective concentration at 50% of the maximal response (EC₅₀) for either pharmacological treatment (Table 3). As such, sum-of-squares F-tests showed that there was no population difference in either the contraction or relaxation response curves.

Our findings show that highland deer mice have altered reflex control of cardiovascular physiology compared with low-altitude conspecifics. Highland deer mice mounted an augmented heart rate response to progressive hypoxia to better maintain blood pressure after acclimation to cold hypoxia, and had a more responsive and effective baroreflex. Vascular sensitivities to α₁-adrenoceptor activation and capacity for endothelium-dependent relaxation of small calibre mesenteric arteries did not differ between highland and lowland mice, reflecting preserved responsiveness to regulators of vascular tone in these vessels. These findings suggest that evolved changes in reflex control may arise from adjustments in autonomic control of the heart and/or other resistance vessels.

The cardiovascular responses to hypoxia appear to have evolved in highland deer mice to help support circulatory O₂ delivery and better maintain O₂ demands during chronic hypoxia exposure (Fig. 1). With tissue hypoxia, paracrine factors released by peripheral tissues favour vasodilation to increase local blood flow and help restore the balance between O₂ supply and demand (Kulandavelu et al., 2015; Premont et al., 2020; Weisbrod et al., 2001). However, with global hypoxaemia, systemic vasodilation may reduce P_mean to such an extent that tissue perfusion pressure and therefore O₂ delivery (and aerobic metabolism) is constrained. The overall cardiovascular response to hypoxia results from the combined influence of these peripheral factors along with hypoxic chemoreflex activation, the baroreflex, and centrally controlled adjustments in metabolism and body temperature. Here, we observed that mild hypoxia at 16 kPa O₂ tended to increase f_H, likely reflecting the stimulatory effects of local vasodilatory factors combined with the hypoxic chemoreflex and baroreflex, and P_mean was maintained at normoxic values (Fig. 1). Declines in P_mean became apparent in moderate to severe hypoxia (8–12 kPa O₂), particularly in mice acclimated to warm normoxia, suggesting that the reflex stimulation of cardiac output (including f_H) was no longer able to fully offset the effects of tissue vasodilatory factors. However, cold hypoxic highlanders maintained greater f_H at 9–10 kPa O₂, in association better maintenance of blood pressure and aerobic metabolism (i.e. V̇_O2). These levels of hypoxia are representative of what could realistically be experienced within underground burrows. The ability to maintain blood pressure to support circulatory O₂ transport at these levels of hypoxia could therefore confer marked advantages in the wild, given the vital importance of aerobic metabolism for thermogenesis and fitness at high altitude (Hayes and O'Connor, 1999; Scott et al., 2024; Wearing and Scott, 2021). Multiple circulatory and metabolic adjustments likely contribute to the variation in cardiovascular function observed here (Chappell et al., 2007; Tate et al., 2020). For example, we have previously found that cold hypoxia acclimation increases maximal stroke volume in highlanders but not in lowlanders (Tate et al., 2020). This increase may enable highlanders to better maintain cardiac output and circulatory O₂ transport in their cold hypoxic environment, helping avoid hypoxia-induced reductions in P_mean that could limit tissue perfusion and aerobic metabolism.

The most severe levels of hypoxia led to larger declines in f_H and P_mean across groups, and were associated with pronounced declines in aerobic metabolism and body temperature. Reductions in metabolism and body temperature are a common response to acute hypoxia in small mammals, and likely help avoid a mismatch between O₂ supply and demand during hypoxia-induced hypoxaemia (Frappell et al., 1992; Gu and Jun, 2018). These reductions may have had a significant influence on f_H here, because hypometabolism likely reduced circulatory O₂ demands whereas T_b depression may have reduced intrinsic heart rate by slowing cardiac pacemaker currents (Joyce and Wang, 2022; Senges et al., 1979). Furthermore, studies in ectotherms show that baroreflex sensitivity is strongly affected by body temperature (Sandblom et al., 2016; Zena et al., 2015), so the reductions in T_b observed here could contribute to why barostatic regulation failed to maintain blood pressure in severe hypoxia. However, metabolic depression did not occur in cold hypoxic highlanders at 10 kPa O₂, in contrast to all other groups, and highlanders also appeared to exhibit less T_b depression at this P_O2 (Fig. 1). This is consistent with previous findings that hypoxia acclimation leads to greater attenuation of T_b depression in severe hypoxia in highlanders than in lowlanders (Ivy and Scott, 2017), and that this difference may be partly underlain by a non-synonymous mutation in Epas1, the gene that encodes hypoxia-inducible factor 2α (Ivy et al., 2022).

The greater ability of highlanders to maintain P_mean and f_H in severe hypoxia could be partly attributed to their greater baroreflex sensitivity (i.e. gain), characterized by greater swings in heart rate in association with smaller changes in arterial pressure compared with low-altitude conspecifics (Fig. 2, Table 2). These measurements necessitated the use of anaesthesia, which can affect baroreflex function, but the low level of isoflurane used here is commonly accepted as the best choice for assessing cardiovascular physiology in rodents (Janssen et al., 2004; Lorenz, 2002; Wearing et al., 2025). At this level of anaesthesia (in absence of phenylephrine/SNP), blood pressure and heart rate are maintained at resting conscious values, and the confounding effects of variation in activity and metabolism are avoided. Furthermore, the modest levels of anaesthesia were consistent across groups, and could not explain the observed differences in baroreflex function between populations. These population differences in baroreflex sensitivity could be partly explained by variation in β₁-adrenergic responsiveness, as suggested by previous findings that highland deer mice have a greater capacity to increase heart rate in response to β₁-adrenergic stimulation than lowland white-footed mice (Wearing et al., 2022). This observation could result from increased expression and/or sensitivity of cardiac β-adrenoceptors in highland deer mice. If so, this difference would contrast findings in high-altitude plateau pika, which have lower cardiac β₁-adrenoceptor expression than domestic rats (a species that is generally found only at low altitudes) (Pichon et al., 2013), but it is unclear whether these findings reflect a specific adaptation to high altitude or a broad taxonomic difference between rodents and lagomorphs such as pika. Chronic hypoxia typically results in reduced expression and desensitization of cardiac β₁-adrenoceptors in low-altitude taxa, likely in response to prolonged sympathetic activation via the hypoxic chemoreflex, with reductions in the chronotropic effects of pharmacological receptor stimulation that would impair baroreflex potency (Antezana et al., 1992; Browne et al., 1997; Richalet, 2016; Richalet et al., 2024, 1988; Voelkel et al., 1981). Whether similar responses occur during combined exposure to cold and hypoxia is unknown, but such responses did not elicit any significant acclimation-induced changes in the baroreflex here (Fig. 2).

Population differences in the baroreflex could also result from variation in vagal (parasympathetic) control. Our previous work in conscious deer mice under routine conditions showed no population difference in routine vagal chronotropic tone, but chronic exposure to cold hypoxia reduced vagal tone in both populations, leading to increased heart rates to support the metabolic demands of thermogenesis at low temperatures (Wearing and Scott, 2022a). Nevertheless, highlanders could experience greater vagal inhibition during baroreceptor loading compared with lowlanders, which could help explain the trend for highlanders to exhibit greater changes in heart rate. Future assessments of intrinsic heart rate and different components of the baroreflex (e.g. baroreceptor sensitivity, sympathetic and parasympathetic activity, etc.) would provide further clarity on the mechanisms underlying evolved changes in baroreflex control.

Population differences in baroreflex control did not appear to be associated with any variation in vascular sensitivity to α-adrenergic stimulation. This is supported by our finding that mesenteric arteries from highlanders exhibit similar contraction responses to phenylephrine as those from lowlanders in comparisons among warm normoxic mice (Fig. 3). However, it is possible that our findings in first-order mesenteric branches are not fully representative of all other resistance vessels, as vasomotor sensitivity can vary with vessel size and branch order (Bund et al., 1994; Hilgers and De Mey, 2009). Whether higher-order mesenteric resistance vessels or resistance vessels supplying other peripheral capillary beds (e.g. in skeletal muscle) have differences in vasomotor sensitivity in highlanders remains to be seen.

We also found that the capacity for endothelium-dependent relaxation of small mesenteric arteries did not differ between highland and lowland deer mice (Fig. 3, Table 3). Further exploration of the mechanisms underlying endothelium-dependent relaxation is warranted because different local mediators of these responses (e.g. nitric oxide and endothelium-dependent hyperpolarization) have been shown to be altered by hypoxia and may compensate when other vasodilator mechanisms are impaired (Spilk et al., 2013). Heterogeneity in the contribution of different endothelium-dependent mechanisms to muscarinic receptor agonist-induced relaxations of vascular smooth muscle has been described in mouse mesenteric arteries (Zhang et al., 2025). Accordingly, our experiments with first-order mesenteric arteries assessed contributions to endothelium-dependent relaxations by pathways including and not including nitric oxide (e.g. endothelium-dependent hyperpolarization). It is possible that the relative importance of these distinct mechanisms varies in smaller diameter mesenteric vessels. Furthermore, many studies using stimuli other than acetylcholine (e.g. shear stress, other G-protein coupled receptor agonists) have reported variation in endothelium-dependent vasodilation mechanisms. Therefore, a wider selection of endothelial cell agonists and comparative studies across microvasculature beds would enable broader generalization about the mechanisms of endothelium function. There could also be intrinsic population differences in vascular network structure that affected responses to vasomotor substances in vivo, based on findings that highlanders have greater capillary densities than lowlanders in some tissues (skeletal muscle) (Lui et al., 2015). Whether such differences affect systemic vascular resistance is unclear, particularly when considering that routine f_H and P_mean are similar between populations (Wearing and Scott, 2022a).

High-altitude deer mice exhibit several physiological adjustments that help circumvent the challenges presented by the cold and hypoxic conditions in their native environment, adjustments that arise via the interactive effects of evolutionary adaptation and environmentally induced plasticity (McClelland and Scott, 2019; Scott et al., 2024; Storz and Cheviron, 2021). Highlanders thus have a greater aerobic capacity for thermogenesis than lowlanders, underlain by greater maximal cardiac output along with several other evolved changes across the O₂ transport pathway and in the metabolic function of thermogenic tissues (Chappell et al., 1988; Coulson et al., 2021; Dane et al., 2018; Hayes and O'Connor, 1999; Lui et al., 2015; Mahalingam et al., 2017; Rezende et al., 2004; Shirkey and Hammond, 2014; Snyder et al., 1982; Storz et al., 2010; Tate et al., 2020; Van Sant and Hammond, 2008; Velotta et al., 2016; Wearing et al., 2021, 2022; Wearing and Scott, 2022a; West et al., 2021). Our current findings add to growing evidence that high-altitude deer mice have also evolved changes in cardiovascular control by the sympathoadrenal system. The present study suggests that highland deer mice have evolved a more pronounced baroreflex, likely in association with greater cardiac responsiveness to β₁-adrenergic receptor stimulation (Wearing et al., 2022), and they better maintain arterial blood pressure in hypoxia. The precise maintenance of blood pressure is essential for safeguarding tissue perfusion pressure and O₂ delivery in hypoxia, and may thus contribute to fitness given the vital importance of thermogenic capacity to survival at high altitude (Hayes and O'Connor, 1999). However, highland deer mice also have lower circulating catecholamines and their adrenal medulla releases fewer catecholamines in response to stimulation (Pranckevicius et al., 2025; Scott et al., 2019).

Taken together, these findings suggest that high-altitude adaptation has led to a nuanced fine-tuning of the neural and humoral components of the sympathoadrenal system. Specifically, highland deer mice appear to have reduced the broad systemic effects of adrenal catecholamine release into the blood in chronic hypoxia, reducing potential issues related to prolonged sympathoexcitation (e.g. general systemic vasoconstriction and blood flow restriction to peripheral tissues) (Alvarez-Araos et al., 2024), while enhancing the autonomic neural control of blood pressure and cardiac output when needed to support high metabolic rates. These evolved changes in cardiovascular function may be especially important for coping with the metabolic challenges of living in the cold and hypoxic environment at high altitude.

The authors thank Joselia Carlos and Angela Wang for their valuable technical assistance with developing the wire myography protocol employed. The authors also thank the veterinary staff in McMaster's Central Animal Facility for assistance with animal care and husbandry.