Cardiovascular and respiratory evaluation in adenosine A2A receptor knockout mice submitted to short‐term sustained hypoxia

Abstract Sustained hypoxia (SH) in mice induces changes in the respiratory pattern and increase in the parasympathetic tone to the heart. Among adenosine G‐protein‐coupled receptors (GPCRs), the A2A receptors are especially important in mediating adenosine actions during hypoxia due to their expression in neurons involved with the generation and modulation of the autonomic and respiratory functions. Herein, we performed an in vivo evaluation of the baseline cardiovascular and respiratory parameters and their changes in response to SH in knockout mice for A2A receptors (A2AKO). SH produced similar and significant reductions in mean arterial pressure and heart rate in both wild‐type (WT) and A2AKO mice when compared to their respective normoxic controls. Mice from WT and A2AKO groups submitted to normoxia or SH presented similar cardiovascular responses to peripheral chemoreflex activation (KCN). Under normoxic conditions A2AKO mice presented a respiratory frequency (f R) significantly higher in relation to the WT group, which was reduced in response to SH. These data show that the lack of adenosine A2A receptors in mice does not affect the cardiovascular parameters and the autonomic responses to chemoreflex activation in control (normoxia) and SH mice. We conclude that the A2A receptors play a major role in the control of respiratory frequency and in the tachypnoeic response to SH in mice.

Exposure to sustained hypoxia (SH) for hours or days is experienced by individuals ascending to high altitudes, and under this condition important adaptative responses are observed in the cardiovascular and respiratory functions (Calbet, 2003;Hansen & Sander, 2003;Powell, 2007;Powell et al., 1998).Recent studies from our laboratory documented that SH (F iO 2 0.1 for 24 h) in mice induced an increase in respiratory activity associated with an augmented parasympathetic tone to the heart (Machado, 2023;Rodrigues et al., 2021;Souza et al., 2022).However, the underlying mechanisms contributing to these changes observed in mice submitted to SH have not yet been evaluated.
Adenosine is an active player in the central nervous system during hypoxic challenges.Under physiological conditions the extracellular levels of adenosine in the brain are relatively low and varied in the range 20-300 nM.However, under challenging conditions such as intense physical activity or hypoxia, the extracellular concentration of adenosine increases into the micromolar range (Borea et al., 2018;Dale et al., 2000;Frenguelli et al., 2003;Lee & Venton, 2018;Liu et al., 2019;Nguyen et al., 2014).Adenosine is also considered one of the most important neuromodulators of synaptic transmission in the brain (Burnstock, 2017;Choudhury et al., 2019;Cunha, 2001;Snyder, 1985).
It is also important to note that an increase in the extracellular level of adenosine has been observed after systemic hypoxia in the nucleus tractus solitarii (NTS), the main synaptic station for processing the peripheral chemoreceptor afferents, as well as within regions containing neurons of the ventral respiratory group that are recruited during hypoxic challenges (Barraco et al., 1991;Gourine et al., 2002;Richter et al., 1999;Winn et al., 1981;Yan et al., 1995).Furthermore, it was described that hypoxia induces adenosine release by the CB of rats, which in turn stimulates the carotid chemoreceptors afferents (Conde & Monteiro, 2004;Drumm et al., 2004), representing another important physiological role for adenosine in mediating the autonomic and respiratory responses to chemoreflex activation.
However, there is no evidence about the involvement of A 2A receptors in the cardiovascular and respiratory adjustments in response to shortterm sustained hypoxia or to peripheral chemoreflex activation in conscious freely moving mice.
Taking into consideration the relevance of adenosine as a signaling molecule during hypoxic challenges and the evidence of an important role of A 2A receptors in the cardiovascular and respiratory neural networks under hypoxia, the aims of this study were to evaluate whether the lack of A 2A receptors in knockout mice submitted to SH affects (1) the changes in the baseline cardiovascular and respiratory parameters, and (2) the cardiovascular and respiratory responses to chemoreflex activation.To reach these goals, we used control (wild-

Highlights
• What is the central question of this study?
Are cardiovascular and respiratory parameters and their changes in response to sustained hypoxia (SH) altered in adenosine A 2A receptor knockout mice?
• What is the main finding and its importance?
Cardiovascular parameters and their changes in response to SH were not altered in A 2A KO mice.
The respiratory frequency in A 2A KO was higher than in WT mice.In response to SH the respiratory frequency increased in WT, while it was reduced in A 2A KO mice.A 2A receptors play a major role in the modulation of respiratory frequency and in the tachypnoeic response to SH in mice.type) and adenosine A 2A receptor knockout mice sujected to SH, and cardiovascular and respiratory recordings were performed in the conscious freely moving condition.

Ethical approval
All experimental protocols used in this study were approved by the Institutional Ethics Committee on Animal Experimentation of the School of Medicine of Ribeirão Preto, USP (CEUA no.029/2021).The experimental protocols are also in accordance with the animal ethics principles and regulations of Experimental Physiology (Grundy, 2015).

Animals
In this study 34 male adenosine A 2A receptor knockout mice (C;129S-

Arterial and venous catheterization
The surgery for implantation of catheters into femoral artery and jugular vein was performed as previously described by Rodrigues, After surgery, an antibiotic (Pentabiotic; Fort Dodge Saúde Animal Ltda., Campinas, SP, Brazil) was administered (0.2 ml of 1.2 million IU, I.M.).Mice were maintained under observation by the investigator for at least 2 h, and then were housed in individual cages for 4 days to recover from the anaesthetic and surgical stresses (Figure 1).

Sustained hypoxia
On the fourth day after the surgery for arterial and venous catheterization, A 2A KO and WT mice were submitted to SH or a normoxic protocol (Figure 1).

Cardiovascular and respiratory recordings in conscious freely moving mice
At the end of the SH or normoxic protocols, the arterial catheter was connected to a pressure transducer (MLT0380; ADInstruments, Bella Vista, NSW, Australia) attached to an amplifier (Bridge Amp, ML221; ADInstruments).Pulsatile arterial pressure (PAP), mean arterial pressure (MAP) and heart rate (HR) signals were acquired by a computerized system (PowerLab 4/25 ML845; ADInstruments) and recorded on a computer (sampling rate: 1 kHz) using an acquisition software (LabChart 5, ADInstruments).
Baseline cardiovascular parameters were recorded for 60 min in room air, but the first half hour of recordings was not considered in the data analyses due to possible stress of the animals in response to the manipulation for connecting the catheteres (Figure 1).
The respiratory parameters were evaluated using a whole-body plethysmography approach (Malan, 1973) in parallel to the baseline cardiovascular recordings.Under these conditions, for each mouse placed inside a sealed acrylic plethysmographic chamber (1 litre), the respiratory-related oscillations in the pressure inside the chamber were detected by a high-sensitivity differential pressure transducer (ML141 spirometer, ADInstruments).The signals were processed by a data acquisition system (PowerLab 4/25 ML845; ADInstruments) and recorded on a computer (sampling rate: 1 kHz) via LabChart software (v.5;ADInstruments).The respiratory volume calibration was performed using a syringe to inject 1 ml of air inside the chamber.Temperatures inside and outside the chamber were continuously monitored.After the 30 min of adaptation to the environment by the mice, the chamber was closed and the respiratory variables were recorded in two series of 10 min each, interspersed for periods of 10 min in which the chamber was opened to avoid a major increase of CO 2 (Figure 1).Tidal volume (V T ) and respiratory frequency (f R ) were calculated as described by Malan (1973), and ventilation ( VE ) was obtained offline as the product of V T and f R .The parameters were analysed using periods of respiratory recordings in which mice were quiet and not exploring the cage.
F I G U R E 1 Schematic representation of protocol for cardiovascular and respiratory recordings in conscious freely moving mice submitted to SH or normoxia.
F I G U R E 2 Schematic representation of protocol for arterial blood collections before and after SH in conscious freely moving mice.

Activation of peripheral chemoreceptors in conscious freely moving mice
After 60 min of baseline cardiovascular and respiratory recordings, potassium cyanide (KCN, 0.16 mg/kg) was injected (I.V.) to activate peripheral chemoreflex, as described by Franchini & Krieger (1993) KCN was injected twice with a 15-min time interval between injections (Figure 1).The maximum changes in HR and MAP were quantified as an average of the responses to two activations and the data between groups were compared.At the end of recordings, mice were killed using an injection of a high concentration of the anaesthetic urethane (Sigma-Aldrich, 2 g kg −1 , I.V.) (Figure 1).

Arterial blood gases and biochemical parameters analysis in conscious freely moving mice
In distinct groups of A 2A KO and WT mice, with the femoral artery previously catheterized, a sample of arterial blood (∼90 µl) was collected via the arterial catheter before and after exposure to SH for arterial blood gas and biochemical parameter analysis (Figure 2).Using the i-STAT CG4+ gasometry cartridge (REF 03P85-25) and its i-STAT analyser (Abbott, Chicago, IL, USA), we measured pH, partial pressure of oxygen (P O 2 ), partial pressure of carbon dioxide (P CO 2 ), oxygen saturation index (S O 2 ) and concentration of bicarbonate (HCO 3 − ) present in arterial blood.After the initial arterial blood gas analysis (before SH exposure), mice were submitted to the SH protocol, and at the end of the SH protocol a new sample of arterial blood was collected for arterial blood gas analysis as described above.The blood sample was collected before and after SH using a syringe (1 ml) attached to the arterial catheter, in which a small negative pressure was carefully applied to collect 90 µl of arterial blood.This procedure was performed in a room air environment with the mice inside the open plethysmographic chamber and after a period of acclimatization of the animals (∼1 h) to avoid any additional stress to the animal (Figure 2).
At the end of the blood sample collections, mice were killed using an injection of a high concentration of the anaesthetic urethane (2 g kg −1 , I.A.; Sigma-Aldrich; Figure 2).

Statistical analysis
Data are expressed as means ± standard deviation (SD).The data were analysed using two-way analysis of variance (two-way ANOVA).
Repeated-measures analysis of arterial blood gases and biochemical parameters data were performed by fitting a mixed-effects model.The two-way ANOVA results for the two individual factors (defined as 'mice' to determine the main effects of the absence of A 2A receptors, and defined as 'SH' to determine the main effects of SH exposure) and interaction (mice vs. SH) are reported.Bonferroni's post-hoc comparison test was used to report the differences among groups.
Differences were considered statistically significant when P ≤ 0.05.All graphical and statistical analysis was performed using GraphPad Prism program (version 8, GraphPad Software, La Jolla, CA, USA).

Respiratory parameters
The WT mice were similar (Mice effect P-values: 0.12 and 0.6239), as well as in the groups of mice submitted to SH when compared to those maintained under normoxia (SH effect P-values: 0.2498 and 0.0637).
However, the interaction of mice versus SH was different (Interaction P-values: <0.0001 and 0.0141).The parameter V T was similar among groups for all factors (Mice effect P = 0.2618, SH effect P = 0.2141, and Interaction P = 0.5280; Figure 5a-c).
The post-hoc comparison test shows that SH in Balb/c WT mice produced a significant increase in f R (241 ± 25 vs.

Cardiovascular responses to peripheral chemoreceptors activation
The increase in MAP (ΔMAP) and bradycardia (ΔHR) in response to peripheral chemoreflex activation with KCN (0.16 mg kg −1 , I.V.) were evaluated in Balb/c WT control (n = 7) and SH (n = 9) mice and in
No differences were observed in arterial blood pH (7.43 ± 0.02 vs.

DISCUSSION
In the present study we used knockout mice for adenosine A 2A Although under conscious freely moving condition we observed no differences in the cardiovascular responses to SH and to chemoreflex activation when comparing A 2A KO and their Balb/c WT control mice, there is evidence in favour of a modulatory role of these receptors in brainstem areas containing neurons directly involved with the generation and modulation of the autonomic activity such as NTS, nucleus ambiguus and rostral ventrolateral medulla (Barraco et al., 1990;Minic et al., 2015Minic et al., , 2018;;Phillis et al., 1997;Scislo et al., 2001;Scislo & O'Leary, 2005, 2006;Thomas et al., 2000), as well as the contribution of adenosine and its A 2 receptors in the signalling transduction that occurs in the CB (Conde et al., 2017;Lahiri et al., 2007;Monteiro & Ribeiro, 1987;Sacramento et al., 2018Sacramento et al., , 2019)).
Considering the cardiovascular parameters under control conditions (normoxia), our findings showed similar levels of MAP, SAP, DAP and HR in WT and A 2A KO mice.These findings are in accordance with previous studies showing no changes in the baseline cardiovascular parameters in the absence of expression of adenosine A 2A receptors in mice (Meriño et al., 2020;Sakata et al., 2005;Sehba et al., 2010).Considering the cardiovascular changes in response to SH, the significant reduction in the HR observed in both A 2A KO and Balb/c WT mice submitted to SH may indicate an important change in the parasympathetic tone to the heart, since we recently documented that C57BL/c mice submitted to SH presented a similar reduction in HR after SH, which was due to a significant increase in the parasympathetic tone to the heart (Machado, 2023;Rodrigues, Souza et al., 2021;Souza et al., 2022).
With respect to the reduction in MAP in both A 2A KO and WT mice submitted to SH we suggest that it was mainly due to the reduction in DAP (London & Guerin, 1999;London & Pannier, 2010;Vlachopoulos & O'Rourke (2000).The observed reduction in HR may also have made some contribution to this fall in MAP because it can impact on the cardiac output (CO).However, a study performed by Janssen et al. (2002) demonstrated that in mice the fluctuations in CO are determined more by changes in stroke volume (SV) than by changes in HR levels.The reduction in DAP of mice submitted to SH, in turn, may be the result of an overall vasodilatation rather than a reduction in the sympathetic outflow to the vessels.Adenosine acting on its A 2A receptors play an important role in the vascular smooth muscle modulating the vascular tone (Guieu et al., 2020;Iwamoto et al., 1994;Reiss et al., 2019;Shryock & Belardinelli, 1997), which could be linked to local mechanism responsible for the reduction in DAP observed in mice submitted to SH.However, the average values of DAP under normoxia as well as after SH were similar between WT and A 2A KO mice, suggesting the A 2A receptors in the vascular smooth muscle are not involved in the reduction of DAP following SH in WT mice.
In relation to the respiratory parameters, we observed that under normoxia (control) the A 2A receptor knockout mice presented a respiratory frequency significantly higher, indicating a key role of these receptors in the maintenance of the basal respiratory activity.
We suggest that A 2A receptors contribute to enhancing a possible inhibitory input to neurons generating the respiratory rhythm, which apparently is not active in mice lacking these receptors (A 2A KO).
A 2A receptors are stimulatory G protein (Gs)-coupled receptors and their expression in GABAergic neurons contributes to the inhibitory modulation of the neural network (Borea et al., 2018;Corsi et al., 1999;Cunha & Ribeiro, 2000;Ochi et al., 2000).Our suggestion is based upon previous pharmacological studies performed in rats and pigs showing that microinjection of the A 2A receptors agonist (CGS-21680) into the 4th ventricle reduced the respiratory frequency, which was blocked by previous microinjection of bicuculline, a GABA A receptor antagonist (Mayer et al., 2006;Wilson et al., 2004).Furthermore, the expression of A 2A receptors was documented in respiratory groups on the ventral surface of the brainstem containing GABAergic neurons, such as the Bötzinger complex, reticular formation, caudal and rostral ventrolateral medula (Zaidi et al., 2006).Zaidi et al. (2006) also described a subpopulation of GABAergic neurons projecting to the phrenic nerve motor nuclei and containing the mRNA for expression of A 2A receptors.
Further pharmacological studies are required to explore this possibility in control mice, which may contribute to clarifying the mechanisms underlying the increase in respiratory frequency in A 2A knockout mice.
It is also important to note that exposure to SH produced an increase in respiratory frequency in Balb/c WT mice, but not in A 2A KO mice.
Indeed, the respiratory frequency of A 2A KO SH mice was lower in comparison to those maintained under normoxia, suggesting that the adenosine A 2A receptors are important for the tachypnoeic response to SH.We may also consider an important role for adenosine and its A 2A receptors in the carotid body chemosensory activity in order to explain the absence of increased respiratory rate after SH in A 2A KO mice.Several studies indicate adenosine as an excitatory mediator of CB hypoxic signaling, and this role is dependent of the activation of A 2A and A 2B receptor subtypes, which are expressed in CB chemoreceptor cells (Conde et al., 2017;Gauda et al., 2000;Kobayashi et al., 2000;Monteiro & Ribeiro, 1987;Sacramento et al., 2018Sacramento et al., , 2019;;Xu et al., 2006).
Combined with the in vivo recordings of the respiratory parameters, the arterial blood gas analysis provides a full characterization of the respiratory profile in A 2A knockout mice and their Balb/c WT controls submitted to SH.In this study we measured arterial pH, P O 2 , P CO 2 and bicarbonate concentration in order to evaluate possible respiratory disturbances and changes in acid-base balance that may occur in knockout and control mice in response to SH.We observed that after SH both A 2A knockout and Balb/c WT control mice presented a decrease in P CO 2 and an increase in arterial P O 2 .In this case we suggest that peripheral chemoreceptors are activated in response to sustained hypoxia producing a tachypnoeic response in order to increase the oxygen uptake in a situation of low F iO 2 (Barros et al., 2002;Machado, 2001).It is important to highlight that the measurement of blood gases after SH was performed only after the animals were removed from the hypoxic chamber and the recording of respiratory parameters was completed, that is, several minutes after the animals returned to the normoxic condition.Therefore, the increase in pulmonary ventilation in response to SH for 24 h may explain the increase in P O 2 and reduction in F iO 2 in arterial blood of mice when the animals were back to normoxia (21% F iO 2 ).Although ventilation did not show a significant increase in A 2A knockouts after SH, the observed increase in tidal volume in these animals may impact pulmonary ventilation in order to contribute to the increase in CO 2 rate elimination and O 2 uptake.
The reduction in P CO 2 as a consequence of an increase in the ventilatory response to SH may cause an acid-base imbalance, characterizing respiratory alkalosis.However, no change in arterial pH was observed in A 2A knockout and Balb/c WT control mice after SH.The reduced arterial bicarbonate concentration of mice after SH may represent a metabolic compensatory response to the reduction in P CO 2 .We may suggest that an increase in bicarbonate excretion by the kidneys may compensate for the respiratory alkalosis produced by hyperventilation, which may explain the normal pH in mice after SH.
Another possibility is that the reduction in blood bicarbonate levels is due to the reduction in P CO 2 , since CO 2 is transported in the blood mainly in the form of bicarbonate ions.Therefore, further studies are required to confirm that the reduction in bicarbonate concentration in arterial blood of mice after SH represents a compensatory mechanism by the kidneys.
Using a knockout mouse model for A 2A receptors combined with cardiovascular and respiratory recordings in conscious freely moving mice, we documented that adenosine A 2A receptors play no major role in the modulatory control of the autonomic components to the cardiovascular system as well as in the cardiovascular responses to peripheral chemoreflex activation in mice.On the other hand, the data show a major role for these receptors in the modulation of respiratory frequency because in A 2A receptor knockout mice the respiratory frequency was higher than in control WT mice.It is interesting that WT mice submitted to SH presented an increase in the respiratory frequency to levels similar to that observed in A 2A KO mice under normoxia, suggesting that in SH the possible inhibitory role played by A 2A receptors is removed.In this case, we suggest that SH by a mechanism yet to be revealed is also removing the adenosine inhibitory modulation of the respiratory frequency.
Taking into consideration that adenosine is a signalling molecule commonly associated with challenging conditions, such as hypoxia, the increased respiratory frequency in normoxic A 2A knockouts actually draws attention to its relevance under physiological conditions, opening interesting possibilities for further studies with the purpose of revealing the mechanisms by which A 2A receptors modulate the respiratory frequency.

Figure 3
Figure3shows the genotype of a representative WT control mouse and a representative A 2A KO mouse.Gel bands of about 550 bases respiratory parameters f R , V T and VE were evaluated in Balb/c WT control (n = 11) and SH (n = 13), and in A 2A KO control (n = 13) and SH (n = 10) mice.The parameteres f R and VE in A 2A KO and Balb/c
Figure 7a-d).The post-hoc comparison test shows that SH produced similar effects on arterial gas blood levels and biochemical parameters in both receptors and Balb/c WT mice as their controls in order to evaluate to what extent the lack of A 2A receptors affects (1) the cardiovascular and respiratory parameters, (2) the changes in these parameters in mice submitted to SH and (3) the cardiovascular and respiratory Blood gases and biochemical parameters in arterial blood of conscious freely moving Balb/c WT and A 2A KO mice before and after SH.Average values of pH (a), bicarbonate concentration (HCO 3 − , b), partial pressure of carbon dioxide (P CO 2 , c), partial pressure of oxygen (P O 2 , d), and oxygen saturation (S O 2 , e) in arterial blood of WT mice before (n = 8) and after SH (n = 8), and A 2A KO mice before (n = 11) and after SH (n = 10).Mixed-effects model analysis for repeated-measures followed by Bonferroni post-hoc test to compare differences between groups.P-values for individual factors, interaction and post-hoc comparison test are indicated.*Different from the measurement before SH (P < 0.05).responses to peripheral chemoreflex activation.Our main findings are the following: (1) SH produced similar effects on cardiovascular parameters of both knockout and WT mice, indicating that the A 2A receptors play no major role in the cardiovascular changes observed in mice in response to SH, (2) under normoxia A 2A KO mice presented respiratory frequency significantly higher than in Balb/c WT controls, suggesting that these receptors are important for the generation and/or modulation of respiratory activity, (3) in contrast to the observed increase in WT mice, the exposure to SH reduced the respiratory frequency of A 2A KO mice, suggesting an important role of these receptors in the tachypnoeic response to SH, and (4) the absence of A 2A receptores produced no changes in the magnitudes of pressor and bradycardic responses to peripheral chemoreflex activation in control (normoxia) and SH mice, indicating these receptors are not playing a role in the neurotransmission/neuromodulation of autonomic responses of this reflex.