Evolved changes in breathing and CO₂ sensitivity in deer mice native to high altitudes

Abstract

We examined the control of breathing by O₂ and CO₂ in deer mice native to high altitude to help uncover the physiological specializations used to cope with hypoxia in high-altitude environments. Highland deer mice (Peromyscus maniculatus) and lowland white-footed mice (P. leucopus) were bred in captivity at sea level. The first and second generation progeny of each population was raised to adulthood and then acclimated to normoxia or hypobaric hypoxia (12 kPa O₂, simulating hypoxia at ~4,300 m) for 6–8 wk. Ventilatory responses to poikilocapnic hypoxia (stepwise reductions in inspired O₂) and hypercapnia (stepwise increases in inspired CO₂) were then compared between groups. Both generations of lowlanders appeared to exhibit ventilatory acclimatization to hypoxia (VAH), in which hypoxia acclimation enhanced the hypoxic ventilatory response and/or made the breathing pattern more effective (higher tidal volumes and lower breathing frequencies at a given total ventilation). In contrast, hypoxia acclimation had no effect on breathing in either generation of highlanders, and breathing was generally similar to hypoxia-acclimated lowlanders. Therefore, attenuation of VAH may be an evolved feature of highlanders that persists for multiple generations in captivity. Hypoxia acclimation increased CO₂ sensitivity of breathing, but in this case, the effect of hypoxia acclimation was similar in highlanders and lowlanders. Our results suggest that highland deer mice have evolved high rates of alveolar ventilation that are unaltered by exposure to chronic hypoxia, but they have preserved ventilatory sensitivity to CO₂.

INTRODUCTION

Animals native to high altitude can provide insight into the evolution of complex physiological systems because they have often adapted to the stressors associated with this challenging environment. High altitude is both cold and hypoxic, which challenges the ability of endotherms to maintain O₂ supply for thermoregulation and exercise. However, many human, other mammal, and bird populations live, reproduce, and exercise at high altitude, and the emerging evidence suggests that they have overcome the challenges of this environment through evolved changes in the O₂ transport cascade (32, 59). The function of this cascade, composed of ventilation, pulmonary diffusion, circulation, tissue diffusion, and cellular O₂ utilization, relies on adequate rates of ventilation to maintain tissue O₂ supply; therefore, increasing breathing is critical for O₂ uptake in hypoxic environments (59).

Breathing is stimulated by reductions in arterial O₂ levels at high altitude. Ventilation increases in response to acute hypoxia challenge, a process termed the hypoxic ventilatory response (HVR). Peripheral chemoreceptors in the carotid bodies sense reductions in the partial pressure of O₂ (Po₂) in arterial blood, which initiates the hypoxic chemoreflex that results in the HVR (12, 46). Breathing and ventilatory O₂ chemosensitivity is further enhanced with prolonged exposure to hypoxia over days to weeks, a process termed ventilatory acclimatization to hypoxia, (VAH). VAH is believed to result from increases in chemosensitivity of the carotid bodies and from increases in central gain of the afferent signals transmitted from the carotid bodies to the brainstem (40, 41, 48, 62). The resulting increases in ventilation improve O₂ uptake by increasing alveolar and arterial Po₂ (17).

Breathing at high altitude is also modulated by arterial CO₂ levels. The increases in ventilation that act to minimize the fall in arterial Po₂ also lead to a decline in the partial pressure of CO₂ (Pco₂) (respiratory hypocapnia) (44, 52). This can reduce the CO₂ chemoreflex drive to breathe, acting as feedback that inhibits the ventilatory response to environmental hypoxia (44). As a result, the HVR measured in poikilocapnic (uncontrolled CO₂) conditions is generally lesser in magnitude than when the HVR is measured under isocapnic conditions (when arterial Pco₂ is experimentally maintained) (34, 53). Furthermore, exposure to chronic hypoxia can increase ventilatory CO₂ sensitivity (9, 10, 29, 49, 57). Therefore, CO₂ sensitivity can have a strong influence on breathing and O₂ uptake in individuals at high altitude.

How has the control of breathing been adjusted in high-altitude natives? The isocapnic HVR as a measure of O₂ chemosensitivity has been examined in some studies of highland human populations, but the HVR of most other highland species has been examined in poikilocapnic conditions. Nevertheless, the literature suggests that highland natives can differ from lowland natives in divergent ways, with some highlanders exhibiting similar or enhanced ventilatory responses (24, 42, 53) and others exhibiting a blunted HVR (20, 23, 50). However, in these studies, it has often been difficult to distinguish uniquely evolved differences in highland taxa from developmental or multigenerational effects of exposure to hypoxia (2, 33). Much less is known about CO₂ sensitivity of breathing in highland taxa. Highland humans appear to have a reduced ventilatory sensitivity to CO₂ (54, 55), but it is unclear whether other high-altitude taxa exhibit a similar or distinct pattern of CO₂ sensitivity.

The objective of this study was to examine how the control of breathing by O₂ and CO₂ has evolved in high-altitude populations of deer mice (Peromyscus maniculatus). Deer mice are broadly distributed across North America and can be found from sea level to over 4,300 m elevation in the Rocky Mountains (16, 38, 56). High-altitude populations must sustain high metabolic rates in the wild (13), and they have evolved a higher aerobic capacity (V̇o_2max) in hypoxia than their low-altitude counterparts (3, 4, 27, 60) in association with changes in hemoglobin-O₂ affinity, cardiac function, muscle capillarity and metabolic phenotype, and tissue gene expression (6, 27, 28, 39, 51, 56, 58, 60, 61). We recently found that the control of breathing also differs in high-altitude native deer mice compared with a congeneric species from low altitude (white-footed mouse, P. leucopus) in a study of animals that were born and raised in captivity at sea level but were the first generation progeny of wild parents (18). Specifically, we found that highlanders do not appear to exhibit VAH in contrast with the robust VAH exhibited by lowlanders but that highlanders have a fixed breathing pattern that is similar to hypoxia-acclimated lowlanders (18). However, because these observations were made in a first generation progeny, it was unclear whether they resulted from an evolved difference in highlanders or from persistent multigenerational effects of the wild parents being born and raised in different native environments. Here, we sought to examine these possibilities by studying mice from both the first and second generations raised in captivity. We also sought to determine whether variation in CO₂ sensitivity has evolved in high-altitude mice and whether this might contribute to the apparent differences in breathing during poikilocapnic hypoxia and in VAH.

MATERIALS AND METHODS

Mouse populations.

Wild adult mice were live-trapped at low altitude on the Great Plains (Nine Mile Prairie, Lancaster County, NE, at 40°52′12″ N, 96″48′20.3″ W, 430 m above sea level) (P. leucopus) and at high altitude on the summit of Mount Evans (Clear Creek County, CO, at 39°35′18″N, 105°38′38″ W, 4,350 m above sea level) (P. maniculatus rufinus) and were then transported to McMaster University (Hamilton, Canada; ~50 m above sea level) and held in captivity. Mice were bred within each population in common conditions to produce a first generation (G1) progeny. This was accomplished by putting breeding pairs in individual cages, removing the male when the female was visibly pregnant, and weaning the born pups and moving them to a separate cage at 21 days after birth. G1 mice were similarly bred within each population to produce a second generation (G2) progeny. The experiments here were conducted on several independent families of G1 mice (3 lowland and 5 highland families) and G2 mice (4 lowland and 4 highland families). We did not determine the relatedness of the wild mice used to establish our breeding colonies, but deer mice were abundant and population sizes are large at each trapping site, so it is likely that the breeders in our colony were unrelated and represented a good proportion of the genetic diversity of the wild parental populations. All captive-born progenies were held in standard holding conditions (24°C–25°C, 12:12 light-dark photoperiod) with unlimited access to food and water and were raised in ambient conditions (sea level normoxia) until 6 mo of age before experiments were conducted. All animal protocols followed guidelines established by the Canadian Council on Animal Care and were approved by the McMaster University Animal Research Ethics Board.

Acclimation groups.

To assess the influence of the native population and acclimation environment, mice were chronically exposed to 1) standard normobaric normoxia or 2) hypobaric hypoxia, simulating the pressure at an elevation of ~4,300 m (barometric pressure of 60 kPa, Po₂ ~12.5 kPa). Specially designed hypobaric chambers were used for exposure to chronic hypoxia, as previously described (18, 30). Mice in hypobaric hypoxia were temporarily returned to normobaric conditions twice per week for <20 min for cage cleaning. Ventilatory measurements were carried out after 6–8 wk of exposure.

Acute hypoxia responses.

We examined the effects of hypoxia acclimation on the response to acute hypoxia in both G1 and G2 mice from each population. Breathing and O₂ consumption rates (V̇o₂) were measured in unrestrained mice using barometric plethysmography and respirometry techniques that are consistent with our previous studies (18, 19). Mice were placed in a whole body plethysmograph with normoxic air (21 kPa O₂, balance N₂) supplied at a rate of 600 ml/min and were given 20–60 min to adjust to the chamber until relaxed and stable breathing and metabolism were observed. Measurements were then recorded for an additional 20 min at 21 kPa O₂, after which mice were exposed to stepwise reductions in inspired Po₂ at 16, 12, 10, 9, and 8 kPa for 20 min at each step. Incurrent gas composition was set by mixing dry compressed gases using precision flow meters (Sierra Instruments, Monterey, CA) and a mass flow controller (MFC-4, Sable Systems, Las Vegas, NV), such that the desired Po₂ was delivered to the chamber at a constant rate of 600 ml/min. Body temperature (T_b) was measured at the end of the experiment using a mouse rectal probe (RET-3-ISO, Physitemp). T_b was also measured exactly 24 h later to determine normoxic T_b (this was used as a proxy for the normoxic T_b at the start of the experiment, which was not measured to prevent stress to the animal).

Breathing and V̇o₂ were determined during the last 10 min at each Po₂. Incurrent and excurrent air flows were subsampled at 200 ml/min. For incurrent air, O₂ fraction was continuously measured using a galvanic fuel cell O₂ analyzer (FC-10, Sable Systems). For excurrent air, we first measured water vapor using a thin-film capacitive water vapor analyzer (RH-300, Sable Systems) then dried the gas stream with prebaked drierite and measured O₂ fraction as above and CO₂ fraction using an infrared CO₂ analyzer (CA-10, Sable Systems). These data were used to calculate V̇o₂, expressed at standard temperature and pressure, using appropriate equations for dry air as described by Lighton (26). Chamber temperature was continuously recorded with a thermocouple (TC-2000, Sable Systems). Breathing frequency and tidal volume were measured from changes in flow across a pneumotachograph in the plethysmograph wall, detected using a differential pressure transducer (Validyne DP45, Cancopass, Mississauga, Canada). Tidal volume was calculated using established equations (7, 21) assuming a constant rate of decline in T_b with declining Po₂, which we have previously shown results in similar tidal volumes to those calculated using direct T_b measurements at each Po₂ (19). Total ventilation was determined as the product of breathing frequency and tidal volume. Total ventilation and tidal volume data are expressed at standard temperature and pressure. Ventilatory O₂ equivalent is the quotient of total ventilation and V̇o₂. All data was acquired using a PowerLab 16/32 and Labchart 8 Pro software (ADInstruments, Colorado Springs, CO).

Mice were returned to their acclimation environment after completing the above protocol and allowed at least 2 days to recover and were then subjected to one of two protocols to measure either the acute hypoxia response in the presence of elevated inspired CO₂ or acute hypercapnia responses, as described below.

Acute hypoxia responses with elevated inspired CO₂.

We measured responses to acute hypoxia in the presence of modestly elevated levels of inspired CO₂ in G1 mice to examine the HVR under conditions in which respiratory hypocapnia was reduced. We used the same whole body plethysmograph and the same stepwise hypoxia conditions as were used for the acute hypoxia responses that are described above, except that mice were also exposed to a constant incurrent Pco₂ of 2 kPa across all acute hypoxia steps. Breathing and metabolism were determined as described above, except that CO₂ fraction was measured in dry incurrent air for a few minutes at the beginning of each step (to assure a constant incurrent CO₂ baseline at the desired level), after which it was measured in dry excurrent air for the remaining time at each step.

Acute hypercapnia responses.

We measured responses to acute stepwise hypercapnia in G2 mice to assess ventilatory CO₂ sensitivity. We used the same whole body plethysmograph and the same conditions as were used for the measurements of acute hypoxia response that are described above, except that breathing and metabolism were measured during acute step-wise increases in environmental Pco₂ at 0, 2, 4, and 6 kPa CO₂. These measurements were made once in normoxia (21 kPa O₂) and once in hypoxia (12 kPa O₂) in two separate experiments that were conducted in a random order and separated by at least 2 days. Breathing and metabolism were determined as described above, except that CO₂ fraction was continuously measured in dry incurrent air, and incurrent and excurrent air was dried and scrubbed free of CO₂ with soda lime and ascarite before O₂ fraction was measured. We therefore calculated V̇o₂ using appropriate equations for dry and CO₂-free air as described by Lighton (26).

Statistical analysis.

Two-factor ANOVA and Holm-Sidak posttests were used throughout. The main effects of acclimation environment (normoxia vs. hypoxia) and inspired gas composition (repeated measure) were evaluated within each population to determine the impacts of hypoxia acclimation on O₂ or CO₂ sensitivity of breathing. Values are reported as means ± SE. All statistical analysis was conducted with SigmaStat software (version 3.5) with a significance level of P < 0.05.

RESULTS

Acute hypoxia responses.

Hypoxia acclimation altered breathing in lowland mice but not in highland mice (Figs. 1 and 2, Tables 1 and 2). Among G1 mice, lowlanders exhibited a robust ventilatory response to hypoxia that was primarily driven by increases in breathing frequency, offset slightly by small decreases in tidal volume, and led to a rise in the ventilatory O₂ equivalent (Fig. 1, A, C, and E, Tables 1 and 2). Hypoxia acclimation had an appreciable effect on breathing, reflected primarily by strong increases in tidal volume and reductions in breathing frequency (Fig. 1, C and E, Table 1), which offset each other such that there was no significant change in total ventilation and only a decrease ventilatory O₂ equivalent at 8 kPa O₂ after hypoxia acclimation (Fig. 1A, Tables 1 and 2). V̇o₂ and T_b declined in response to hypoxia challenge, but hypoxia acclimation reduced the declines in T_b (Tables 1 and 2). In contrast, hypoxia acclimation had very little effect on breathing or metabolism in G1 highland mice (Fig. 1, B, D, and F, Tables 1 and 2).

Fig. 1.

Hypoxia acclimation has very little effect on breathing in highland deer mice (B, D, and F) from the first generation (G1) raised in captivity, unlike G1 mice from low altitude (A, C, and E). *Significant pairwise difference between acclimation groups within each Po₂ using Holm-Sidak posttests (n as follows: 13 normoxia-acclimated lowlanders, 15 hypoxia-acclimated lowlanders, 15 normoxia-acclimated highlanders, 13 hypoxia-acclimated highlanders). Acc, acclimation.

Fig. 2.

Hypoxia acclimation has no effect on breathing in highland deer mice (B, D, and F) from the second-generation (G2) raised in captivity, unlike G2 mice from low altitude (A, C, and E). *Significant pairwise difference between acclimation groups within each Po₂ using Holm-Sidak posttests (n as follows: 10 normoxia-acclimated lowlanders, 9 hypoxia-acclimated lowlanders, 14 normoxia-acclimated highlanders, 14 hypoxia-acclimated highlanders). Acc, acclimation.

Table 1.

Statistical results of two-way ANOVA of acute hypoxia responses

	Acclimation Environment		Acute Po2		Interaction
	F	P	F	P	F	P
First generation mice
Total ventilation
Lowlander	2.612	0.118	78.61	<0.001	1.889	0.101
Highlander	1.149	0.294	74.62	<0.001	0.416	0.837
Breathing frequency
Lowlander	11.67	0.002	179.9	<0.001	12.07	<0.001
Highlander	1.160	0.291	286.4	<0.001	3.841	0.003
Tidal volume
Lowlander	18.74	<0.001	10.12	<0.001	0.555	0.735
Highlander	0.092	0.764	31.17	<0.001	0.954	0.449
O2 consumption
Lowlander	1.233	0.277	12.4	<0.001	1.989	0.084
Highlander	0.794	0.381	22.92	<0.001	2.148	0.064
Body temperature
Lowlander	6.047	0.021	196.8	<0.001	8.878	0.006
Highlander	12.92	0.001	61.55	<0.001	6.532	0.017
Ventilatory O2 equivalent
Lowlander	0.065	0.801	101.3	<0.001	4.133	0.002
Highlander	0.015	0.904	161.9	<0.001	1.168	0.328
Second generation mice
Total ventilation
Lowlander	1.418	0.251	17.78	<0.001	2.212	0.061
Highlander	0.043	0.838	24.67	<0.001	3.467	0.006
Breathing frequency
Lowlander	5.799	0.028	59.74	<0.001	5.089	<0.001
Highlander	0.486	0.492	236.6	<0.001	1.111	0.358
Tidal volume
Lowlander	4.830	0.043	6.258	<0.001	6.296	<0.001
Highlander	0.562	0.460	71.05	<0.001	0.979	0.433
O2 consumption
Lowlander	0.827	0.377	17.84	<0.001	0.945	0.457
Highlander	0.016	0.900	27.12	<0.001	4.900	<0.001
Body temperature
Lowlander	0.057	0.815	143.0	<0.001	2.486	0.134
Highlander	6.126	0.020	32.53	<0.001	4.640	0.041
Ventilatory O2 equivalent
Lowlander	0.079	0.782	95.14	<0.001	0.366	0.871
Highlander	0.012	0.913	85.07	<0.001	0.562	0.729

One degree of freedom for acclimation environment, 5 degrees of freedom for acute Po₂ and their interaction, 130 degrees of freedom for the residuals of first generation lowland and highland mice and second generation highland mice, and 80 degrees of freedom for the residuals of second generation lowland mice. Boldface denotes significant P values.

Table 2.

Rate of O₂ consumption and body temperature during acute hypoxia exposure in first generation laboratory-raised mice

	Lowland Peromyscus leucopus		Highland Peromyscus maniculatus
Acute Po2, kPa	Normoxia-acclimated	Hypoxia-acclimated	Normoxia-acclimated	Hypoxia-acclimated
O2 consumption rate, ml·g−1·min−1
21	0.045 ± 0.004	0.045 ± 0.003	0.048 ± 0.003	0.045 ± 0.003
16	0.043 ± 0.004	0.040 ± 0.002	0.039 ± 0.003	0.039 ± 0.004
12	0.039 ± 0.003	0.041 ± 0.003	0.035 ± 0.002	0.037 ± 0.002
10	0.035 ± 0.003	0.037 ± 0.002	0.032 ± 0.002	0.038 ± 0.002
9	0.031 ± 0.002	0.038 ± 0.002	0.031 ± 0.002	0.037 ± 0.002
8	0.027 ± 0.002	0.036 ± 0.002	0.030 ± 0.002	0.033 ± 0.002
Ventilatory O2 equivalent, ml air/ml O2
21	33.97 ± 2.58	41.20 ± 1.79	35.94 ± 1.49	39.16 ± 2.22
16	40.43 ± 2.09	48.95 ± 2.23	44.93 ± 1.95	47.77 ± 2.27
12	55.86 ± 2.74	58.28 ± 2.24	57.13 ± 2.66	59.31 ± 1.77
10	71.97 ± 3.32	67.79 ± 3.18	74.18 ± 2.68	70.28 ± 3.96
9	81.87 ± 4.53	73.46 ± 2.51	84.98 ± 2.69	79.22 ± 4.17
8	92.26 ± 5.50	82.31 ± 2.81*	97.92 ± 3.84	95.21 ± 5.16
Body temperature, °C
21	37.29 ± 0.29	37.25 ± 0.24	36.08 ± 0.24	36.49 ± 0.23
8	33.52 ± 0.21	34.80 ± 0.21*	33.60 ± 0.23	35.20 ± 0.34*

Values are means ± SE. Po₂, partial pressure of O₂.

^*Significant pairwise difference from normoxia-acclimated mice of the same population.

The diminished effects of hypoxia acclimation in G1 highland mice were also observed in G2 highland mice (Fig. 2, Tables 1 and 3). Hypoxia acclimation had a strong effect on breathing in G2 lowlanders, as it did in G1 lowlanders, as reflected by significant increases in tidal volume (Fig. 2E) and reductions in breathing frequency (Fig. 2C) during acute hypoxia. Interestingly, there appeared to be some variation in breathing across generations in lowlanders, which tended to have lower total ventilation, tidal volume, and ventilatory O₂ equivalent in G2 than in G1. Nevertheless, even though there was some subtle variation in the magnitude of breathing and the HVR across generations, the effects of hypoxia acclimation on breathing pattern in lowlanders was generally preserved. In contrast, hypoxia acclimation had no effect on breathing or metabolism in G2 highlanders (Fig. 2, Table 2), as was observed in G1 highlanders, and the magnitude of breathing and the HVR was very similar across generations.

Table 3.

Rate of O₂ consumption and body temperature during acute hypoxia exposure in second-generation laboratory-raised mice

	Lowland Peromyscus leucopus		Highland Peromyscus maniculatus
Acute Po2, kPa	Normoxia-acclimated	Hypoxia-acclimated	Normoxia-acclimated	Hypoxia-acclimated
O2 consumption rate, ml·g−1·min−1
21	0.045 ± 0.004	0.044 ± 0.004	0.061 ± 0.002	0.051 ± 0.003
16	0.040 ± 0.003	0.039 ± 0.003	0.055 ± 0.003	0.052 ± 0.003
12	0.038 ± 0.003	0.039 ± 0.004	0.051 ± 0.003	0.049 ± 0.003
10	0.032 ± 0.003	0.037 ± 0.003	0.045 ± 0.002	0.049 ± 0.004
9	0.031 ± 0.002	0.035 ± 0.003	0.039 ± 0.001	0.043 ± 0.003
8	0.028 ± 0.002	0.031 ± 0.002	0.037 ± 0.001	0.042 ± 0.002
Ventilatory O2 equivalent, ml air/ml O2
21	31.03 ± 2.28	29.97 ± 3.99	31.87 ± 1.73	34.78 ± 1.71
16	35.75 ± 1.49	38.67 ± 2.47	34.84 ± 1.51	35.82 ± 2.12
12	44.96 ± 1.58	45.53 ± 2.53	43.67 ± 1.75	42.85 ± 2.40
10	53.45 ± 1.72	52.37 ± 2.10	51.37 ± 2.64	48.84 ± 1.81
9	57.85 ± 1.88	58.20 ± 1.56	55.67 ± 2.76	55.31 ± 3.10
8	64.07 ± 2.32	64.62 ± 2.06	61.38 ± 3.47	62.34 ± 4.55
Body temperature, °C
21	37.63 ± 0.33	37.45 ± 0.22	36.23 ± 0.29	36.58 ± 0.30
8	34.30 ± 0.19	34.70 ± 0.30	34.50 ± 0.23	35.80 ± 0.31*

Values are means ± SE. Po₂, partial pressure of O₂.

^*Significant pairwise difference from normoxia-acclimated mice of the same population.

Hypercapnic ventilatory response.

We next sought to determine whether hypoxia acclimation increases CO₂ sensitivity of breathing and whether this effect of hypoxia acclimation is altered in highlanders. We examined these possibilities by measuring the ventilatory response to stepwise hypercapnia in normoxia and hypoxia. We found that hypoxia acclimation enhanced the hypercapnic ventilatory response in both lowland and highland mice when tested in normoxic conditions (Fig. 3, Table 4). Both lowlanders and highlanders exhibited similar robust ventilatory responses to increasing CO₂ that were driven by increases in both tidal volume and breathing frequency (Fig. 3, Table 4). This response was augmented similarly after hypoxia acclimation in both populations, particularly at higher CO₂ levels, because of further increases in tidal volume in both populations and breathing frequency in lowlanders. V̇o₂ and T_b were not altered by acute hypercapnia or by hypoxia acclimation in either population (Table 4, data not shown).

Fig. 3.

Hypoxia acclimation increased ventilatory sensitivity to CO₂ in both lowland (A, C, and E) and highland (B, D, and F) mice when measured in normoxic conditions (21 kPa O₂). Measurements were made on second generation (G2) mice. *Significant pairwise difference between acclimation groups within each Pco₂ using Holm-Sidak posttests (n as follows: 10 normoxia-acclimated lowlanders, 9 hypoxia-acclimated lowlanders, 13 normoxia-acclimated highlanders, 14 hypoxia-acclimated highlanders).

Table 4.

Statistical results of two-way ANOVA of acute hypercapnia responses, measured in normoxia or hypoxia

	Acclimation Environment		Acute Pco2		Interaction
	F	P	F	P	F	P
Normoxic conditions (21 kPa O2)
Total ventilation
Lowlander	7.437	0.014	92.07	<0.001	9.496	<0.001
Highlander	11.68	0.002	249.3	<0.001	20.26	<0.001
Breathing frequency
Lowlander	2.535	0.130	51.60	<0.001	3.006	0.039
Highlander	0.223	0.641	181.2	<0.001	3.477	0.020
Tidal volume
Lowlander	3.451	0.081	106.0	<0.001	5.843	0.002
Highlander	12.55	0.002	120.8	<0.001	15.61	<0.001
O2 consumption
Lowlander	3.022	0.100	1.741	0.170	0.755	0.524
Highlander	0.579	0.454	1.404	0.248	0.315	0.815
Body temperature
Lowlander	0.045	0.835	5.147	0.037	0.572	0.460
Highlander	0.055	0.817	39.52	<0.001	0.374	0.546
Hypoxic conditions (12 kPa O2)
Total ventilation
Lowlander	9.610	0.007	87.13	<0.001	11.52	<0.001
Highlander	4.372	0.047	118.6	<0.001	4.831	0.004
Breathing frequency
Lowlander	0.088	0.770	16.40	<0.001	8.088	<0.001
Highlander	1.925	0.178	45.09	<0.001	0.723	0.541
Tidal volume
Lowlander	4.916	0.041	195.6	<0.001	5.279	0.003
Highlander	14.87	<0.001	210.6	<0.001	14.00	<0.001
O2 consumption
Lowlander	0.001	0.991	1.020	0.392	1.262	0.297
Highlander	0.095	0.761	3.722	0.015	1.835	0.148
Body temperature
Lowlander	1.087	0.312	9.683	0.006	0.184	0.673
Highlander	0.444	0.511	24.77	<0.001	0.776	0.387

One degree of freedom for acclimation environment, 3 degrees of freedom for acute Pco₂ and their interaction, and 51 and 75 degrees of freedom for the residuals of lowland and highland mice, respectively, in normoxic and hypoxic conditions. Boldface denotes significant P values.

We found that hypoxia acclimation resulted in comparable increases in the hypercapnic ventilatory response when it was tested in hypoxic conditions (12 kPa O₂) (Fig. 4, Table 4). The response to acute hypercapnia was similar to that observed in normoxic conditions, except that ventilation was higher overall due to hypoxia. Hypoxia acclimation augmented the hypercapnic ventilatory response measured in hypoxia in both populations, as observed for the hypercapnic ventilatory response measured in normoxia, but in this case, the increases in ventilation were entirely caused by increases in tidal volume (Fig. 4, E and F). These findings suggest that hypoxia acclimation increases the CO₂ sensitivity of breathing in both highland and lowland mice, but in general, there were no apparent differences in the hypercapnic ventilatory response of highlanders compared with lowlanders.

Fig. 4.

Hypoxia acclimation increased ventilatory sensitivity to CO₂ in both lowland (A, C, and E) and highland (B, D, and F) mice, when measured in hypoxic conditions (12 kPa O₂). Measurements were made on second generation (G2) mice. *Significant pairwise difference between acclimation groups within each Pco₂ using Holm-Sidak posttests (n as in Fig. 3).

Acute hypoxia responses in 2 kPa CO₂.

We sought to examine whether the apparent lack of VAH in highlanders could be a by-product of increases in ventilatory sensitivity to respiratory hypocapnia after chronic hypoxia. Given that hypoxia acclimation appears to augment CO₂ sensitivity (Figs. 3 and 4), this could foreseeably augment the restraining influence of respiratory hypocapnia on the poikilocapnic HVR and thus offset other effects of chronic hypoxia that stimulate breathing and would tend to cause VAH. We examined this possibility by measuring the HVR during moderately elevated inspired CO₂ (2 kPa CO₂) to offset respiratory hypocapnia with the prediction that increases in CO₂ would amplify the effects of chronic hypoxia on breathing. There was some modest support for this prediction, as reflected by the apparent increase in the magnitude of the effects of hypoxia acclimation on average in both populations (compare the variation in tidal volume in Fig. 5 with Fig. 1). In highlanders in particular, hypoxia acclimation increased tidal volume at the higher Po₂ values when measured in the presence of 2 kPa CO₂ (Fig. 5F, Table 5). However, hypoxia acclimation still had a much smaller effect on breathing in highlanders than in lowlanders, and there were still no significant main effects of hypoxia acclimation on total ventilation, breathing frequency, or tidal volume in highlanders (Table 5). Otherwise, the effects of acute hypoxia on breathing and metabolism in the presence of 2 kPa CO₂ were quite similar to those observed without CO₂ in the inspired gas (Tables 5 and 6). Therefore, the apparent lack of VAH in highlanders cannot be explained by variation in the CO₂ sensitivity of breathing.

Fig. 5.

Hypoxic ventilatory responses measured in the presence of moderately elevated levels of inspired CO₂ (2 kPa). Measurements were made on first generation (G1) lowlander (A, C, and E) and highlander (B, D, and F) mice. *Significant pairwise difference between acclimation groups within each Po₂ using Holm-Sidak posttests (n as in Fig. 1).

Table 5.

Statistical results of two-way ANOVA of acute hypoxia responses, measured with elevated inspired CO₂

	Acclimation Environment		Acute Po2		Interaction
	F	P	F	P	F	P
Total ventilation
Lowlander	4.082	0.054	41.97	<0.001	0.451	0.812
Highlander	1.672	0.207	37.58	<0.001	0.800	0.552
Breathing frequency
Lowlander	10.09	0.004	93.50	<0.001	10.64	<0.001
Highlander	0.001	0.987	130.7	<0.001	0.428	0.828
Tidal volume
Lowlander	17.93	<0.001	1.419	0.222	1.454	0.209
Highlander	3.223	0.084	11.74	<0.001	1.422	0.220
O2 consumption
Lowlander	0.656	0.425	1.226	0.300	0.237	0.946
Highlander	3.637	0.068	10.99	<0.001	0.898	0.484
Body temperature
Lowlander	7.929	0.009	97.30	<0.001	0.244	0.625
Highlander	7.939	0.009	36.55	<0.001	8.415	0.007

One degree of freedom for acclimation environment, 5 degrees of freedom for acute Po₂ and their interaction, and 130 degrees of freedom for the residuals of lowland and highland mice. Boldface denotes significant P values.

Table 6.

Rate of O₂ consumption and body temperature during acute hypoxia exposure, measured with elevated inspired CO₂

	Lowland Peromyscus leucopus		Highland Peromyscus maniculatus
Acute Po2, kPa	Normoxia-acclimated	Hypoxia-acclimated	Normoxia-acclimated	Hypoxia-acclimated
O2 consumption rate, ml·g−1·min−1
21	0.039 ± 0.003	0.041 ± 0.003	0.052 ± 0.005	0.054 ± 0.003
16	0.037 ± 0.003	0.038 ± 0.002	0.042 ± 0.003	0.046 ± 0.002
12	0.037 ± 0.002	0.041 ± 0.002	0.039 ± 0.003	0.048 ± 0.003
10	0.039 ± 0.003	0.039 ± 0.002	0.039 ± 0.002	0.046 ± 0.003
9	0.037 ± 0.002	0.040 ± 0.002	0.039 ± 0.002	0.047 ± 0.004
8	0.035 ± 0.002	0.038 ± 0.002	0.036 ± 0.002	0.044 ± 0.004
Body temperature, °C
21	36.94 ± 0.35	37.92 ± 0.28	36.49 ± 0.22	36.36 ± 0.17
8	34.11 ± 0.35	34.79 ± 0.24	34.30 ± 0.25	35.60 ± 0.23*

Values are means ± SE. Po_2, partial pressure of O₂.

^*Significant pairwise difference from normoxia-acclimated mice of the same population.

DISCUSSION

Effective control of ventilation is critically important for small endotherms in the O₂-limited environment at high altitude to maintain adequate tissue O₂ supply for thermogenesis and exercise. Previously, we observed that hypoxia acclimation had little effect on breathing and the HVR of high-altitude native deer mice at levels of chronic hypoxia that did induce a VAH response in lowland mice (18). Here, we show that the apparent blunting of VAH is observed across multiple generations of laboratory-raised highland mice, suggesting that this blunting has evolved in response to the challenges of life at high altitude. This blunting was not associated with any evolved change in the effects of hypoxia acclimation on CO₂ sensitivity of breathing in highland mice. As a consequence, variation in the ventilatory sensitivity to respiratory hypocapnia does not appear to contribute to the attenuation of VAH in highlanders.

VAH is attenuated in high-altitude native deer mice.

Our findings suggest that the apparent lack of VAH in high-altitude native deer mice may result from an evolved change in the magnitude of hypoxia-induced plasticity of breathing. In previous studies, it has been challenging to establish whether changes in the control of breathing in highland taxa are evolved and genetically based because it has often been difficult to exclude the influence of developmental and/or parental exposure to different environments (2, 33). The blunted VAH we previously reported in first generation highlanders raised in captivity suggested that this blunting cannot be explained by differences in the developmental environment (18). However, it was possible that these differences between populations of first generation mice could be explained by exposure of parents and/or germ cells to different environments. For example, exposure of parents and their germline cells to hypoxia has persistent effects on hypoxia tolerance in offspring in zebrafish (15). However, the persistent lack of VAH in the second generation of highlanders raised in captivity cannot be attributed to the exposure of parents and germline cells to the high-altitude environment. There might have been some effect of the native low-altitude environment in lowlanders, based on our observations that breathing tended to decline slightly from G1 to G2, but this did not affect the expression of VAH in lowlanders. The attenuation of VAH in highlanders is more likely to be an evolved trait and could also be influenced by transgenerational epigenetic effects.

Our understanding of the potential importance of transgenerational epigenetic effects on cardiorespiratory physiology at high altitude is still in its infancy (1, 17, 37). There is known to be epigenetic regulation of cardiorespiratory physiology and carotid body function in response to chronic exposure to intermittent hypoxia in adulthood or early development (35, 36), but it is unknown whether such effects can persist across generations. Effects of some other environmental stressors are known to persist across multiple generations [e.g., transgenerational effects of pollutants on survival and development in zebrafish (5)], but it is unclear whether chronic hypoxia can have a similarly persistent effect. Disentangling the relative importance of genetically and epigenetically based changes in adaptive phenotypes at high altitude will be an exciting area for future research.

What are the mechanisms that account for blunting of the VAH in high-altitude native deer mice? In lowlanders, VAH arises from adjustments in the carotid bodies and the central nervous system. Chronic hypoxia enhances O₂ chemosensitivity of the carotid bodies, which appears to be associated with neovascularization and growth of the organ, and changes in O₂ signaling by O₂-sensitive glomus cells (22, 47, 62). Some of these adjustments appear to be attenuated in highland deer mice, as reflected by our previous observation that highlanders do not exhibit carotid body growth in response to chronic hypoxia (18). Chronic hypoxia also leads to increases in central gain of the afferent signals from the carotid body in lowlanders (e.g., changes in glutamatergic signaling in the nucleus of the solitary tract) (40, 48), and it is possible that these mechanisms are also attenuated in highland mice. However, before carrying out the current study, we could not exclude the possibility that the apparent blunting in VAH arose from a variation in the effects of CO₂ on breathing because the HVR was measured under poikilocapnic conditions. Our results here suggest that this is not the case because highlanders still exhibited a blunted VAH when respiratory hypocapnia was alleviated by exposure to moderately elevated, inspired CO₂ (Fig. 5). This likely implies that VAH and its underlying peripheral and/or central mechanisms are indeed blunted in highland mice. This does not appear to have altered the effects of chronic hypoxia on hypoxic anapyrexia because the magnitude of T_b depression in response to acute hypoxia was still reduced after hypoxia acclimation in highland mice (Tables 2 and 3).

Effects of chronic hypoxia on the CO₂ sensitivity of breathing.

Hypoxia acclimation increased the CO₂ sensitivity of breathing, driven primarily by larger increases in tidal volume in response to high CO₂. This observation is consistent with previous observations in humans in which chronic hypoxia increases ventilatory CO₂ sensitivity and/or lowers the recruitment threshold above which blood CO₂ stimulates ventilation (9, 10, 54). Chronic hypoxia is well known to induce mechanisms of acid-based compensation to counter respiratory hypocapnia and alkalosis (45), so it is possible that apparent changes in the CO₂ sensitivity of breathing arise from changes in the relationship between CO₂ and pH or in pH buffering of the blood (9, 10). It is also possible that increases in ventilatory CO₂ sensitivity in response to chronic hypoxia arise from increases in the chemosensitivity of peripheral or central CO₂/pH chemoreceptors, as suggested by previous studies in humans (11). If this is also the case in deer mice, then the mechanisms likely do not depend upon hypoxia-induced growth of the carotid bodies, which occurs in lowlanders but not in highlanders (18).

Ventilatory sensitivity to CO₂ appeared to be similar in high-altitude mice and their low-altitude counterparts, both before and after hypoxia acclimation. This contrasts previous findings in some other high-altitude taxa. Ventilatory sensitivity to CO₂ and ventilatory recruitment threshold are lower in Himalayans residing at high altitude than in lowlanders at sea level (55). Similarly, ventilatory sensitivity to CO₂ is lower in Andeans residing at high altitude than in lowlanders acclimatized to the same altitude for 10 days (54). In bar-headed geese, a species that flies over the Himalayas during its biannual migration, ventilatory sensitivity to hypercapnia is unaltered, but sensitivity to respiratory hypocapnia is reduced, such that bar-headed geese breathe more than low-altitude birds when exposed to poikilocapnic hypoxia (53). Therefore, there appears to be differences across highland taxa, with ventilatory sensitivity to CO₂ having been either unaltered or reduced.

Perspectives and Significance

The emerging evidence suggests that there are a number of changes in the control of breathing by hypoxia in high-altitude deer mice. Our results here and in previous studies suggest that VAH and hypoxia-induced growth of the carotid bodies are attenuated in highlanders, which may represent an evolved loss of plasticity associated with high-altitude adaptation. Highlanders instead exhibit a fixed breathing pattern, characterized by deep but less frequent breaths, which should improve effective (alveolar) ventilation and thus help increase arterial O₂ saturation in hypoxia (18). These changes exist without any apparent alterations in ventilatory CO₂ sensitivity. An intriguing question to consider is why these putatively evolved changes have taken place. VAH increases ventilation and thus improves respiratory gas exchange, so why have high-altitude mice not maintained the VAH response that is typical of lowlanders? One possibility is that highland mice have undergone the evolutionary process of genetic assimilation, in which a phenotype that originally exhibits adaptive plasticity becomes genetically fixed (assimilated) (8, 25, 43). Another possibility that we and others have discussed previously is that there may have been an overall restructuring of the hypoxic chemoreflex in high-altitude deer mice (18, 31). It is possible that by fixing a breathing pattern that is beneficial for O₂ uptake, highlanders may avoid some costs associated with plasticity in response to chronic hypoxia at high altitude (e.g., chronic sympathetic activation, etc.). Given the harshness of high-altitude environments and the correspondingly strong selection favoring respiratory performance (14), evolved changes in control of breathing may help safeguard O₂ uptake and contribute to the success and high abundance of deer mice in high-altitude environments.

GRANTS

The equipment and operational costs of this research were supported by funds from McMaster University, the Canadian Foundation for Innovation, the Ontario Ministry of Research and Innovation, and a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant to G. R. Scott. C. M. Ivy was supported by a NSERC Postgraduate Scholarship and an Ontario Graduate Scholarship. G. R. Scott is supported by the Canada Research Chairs Program.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.M.I. and G.R.S. conceived and designed research; C.M.I. performed experiments; C.M.I. analyzed data; C.M.I. and G.R.S. interpreted results of experiments; C.M.I. and G.R.S. prepared figures; C.M.I. drafted manuscript; C.M.I. and G.R.S. edited and revised manuscript; C.M.I. and G.R.S. approved final version of manuscript.