Sympathetic dysregulation induced by postnatal intermittent hypoxia

Abstract

Study Objectives

Exposure to postnatal chronic intermittent hypoxia (pCIH), as experienced in sleep-disordered breathing, is a risk factor for developing cardiorespiratory diseases in adulthood. pCIH causes respiratory instability and motor dysfunction that persist until adult life. In this study, we investigated the impact of pCIH on the sympathetic control of arterial pressure in rats.

Methods and Results

Neonate male Holtzman rats (P0–1) were exposed to pCIH (6% O₂ for 30 seconds, every 10 minutes, 8 h/day) during their first 10–15 days of life, while control animals were maintained under normoxia. In early adult life (P25–40), freely behaving pCIH animals (n = 13) showed higher baseline arterial pressure levels linked to augmented sympathetic-mediated variability than control animals (n = 12, p < 0.05). Using decerebrated in situ preparations, we found that juvenile pCIH rats exhibited a twofold increase in thoracic sympathetic nerve activity (n = 14) and elevated firing frequency of ventromedullary presympathetic neurons (n = 7) compared to control rats (n = 6–7, p < 0.05). This pCIH-induced sympathetic dysregulation was associated with increased HIF-1α (hypoxia-inducible factor 1 alpha) mRNA expression in catecholaminergic presympathetic neurons (n = 5, p < 0.05). At older age (P90–99), pCIH rats displayed higher arterial pressure levels and larger depressor responses to ganglionic blockade (n = 6–8, p < 0.05), confirming the sympathetic overactivity state.

Conclusions

pCIH facilitates the vasoconstrictor sympathetic drive by mechanisms associated with enhanced firing activity and HIF-1α expression in ventromedullary presympathetic neurons. This excessive sympathetic activity persists until adulthood resulting in high blood pressure levels and variability, which contribute to developing cardiovascular diseases.

Exposure to intermittent hypoxia during the first 2 weeks of life (pCIH) elevates resting arterial blood pressure of rats at juvenile and adult ages. In the ventral lateral medulla, pCIH increases the expression of the hypoxia-inducible factor 1 alpha (HIF-1α) mRNA and the firing frequency of presympathetic neurons, leading to higher levels of vasoconstrictor sympathetic activity. Created with BioRender.com.

intermittent hypoxia, postnatal, sympathetic activity, hypertension, brainstem

Statement of Significance

Postnatal chronic intermittent hypoxia (pCIH), as experienced by premature babies and infants with sleep-disordered breathing, is a risk factor for developing cardiovascular dysfunctions in adulthood. Here, we identify in rodents that pCIH causes persistent activation of the brain circuitry that regulates blood vessel diameter, leading to elevated arterial pressure levels. This long-term effect of pCIH on arterial pressure is associated with increased expression of a protein named hypoxia-inducible factor 1 alpha (HIF-1α) in neurons that control the sympathetic activity to the cardiovascular system. HIF-1α can alter the transcription of other genes and regulate cellular activity, suggesting that genomic-associated mechanisms may underly the effects of pCIH on brain circuitries controlling arterial pressure.

Introduction

Cardiovascular diseases (CVD) are among the leading causes of death worldwide. Hypertension, which affects approximately one-third of the global male and female adult population [1], is a significant risk factor for CVD and premature death. Despite identifying conditions contributing to the chronic elevation of arterial pressure levels (e.g. obesity, physical inactivity, and genetics), the underlying etiology of primary hypertension remains under investigation. Consequently, many patients suffer from uncontrolled arterial hypertension despite using five or more antihypertensive drugs within optimal doses [2, 3], putting them at a greater cardiovascular risk. Clinical evidence demonstrates that a large proportion of hypertensive patients, including the ones with refractory hypertension, exhibit excessive sympathetic activity to the cardiovascular system [4], indicating a neurogenic component in the development of arterial hypertension. This possibility is confirmed by clinical and experimental data showing that changes in the tonus and rhythmicity of the sympathetic activity are commonly associated with abnormal and variable levels of arterial pressure (which represent a bad prognosis in patients with CVD [5–9]).

The maintenance of arterial pressure at adequate levels requires a tonic sympathetic drive to the vascular tree, controlling the contraction of smooth muscle cells and the caliber of blood vessels. These sympathetic nerves, formed by postganglionic neurons, are controlled by preganglionic neurons whose cells bodies are in the intermediate lateral column (IML) of the spinal cord at the thoracic and upper lumbar segments [10]. The activity of the IML preganglionic neurons is determined by synaptic inputs from sympathetic premotor regions located in the medulla oblongata, pons, and hypothalamus [9, 10]. Among these regions, a cluster of bulbospinal glutamatergic neurons located in the rostral ventrolateral aspect of the medulla oblongata (RVLM) is a major source of excitatory inputs to the preganglionic sympathetic neurons [11, 12] and necessary for the tonic regulation of sympathetic outflow to the cardiovascular system [9]. The RVLM activity is mainly determined by a balance of excitatory and inhibitory synaptic drives arising from multiple brain areas [13, 14], although RVLM neurons may also be capable of intrinsic firing based on pacemaker properties [15]. Experimental data indicate that augmented sympathetic outflow and arterial pressure levels in models of neurogenic hypertension are associated with elevated activity of the RVLM presympathetic neurons [16, 17]. Several mechanisms have been described to contribute to the RVLM hyperactivity in hypertensive animals, such as enhanced excitatory drive arising from the carotid body chemoreceptors [18] or the respiratory network [19], impaired inhibitory baroreflex control [20], changes in neuromodulatory mechanisms [21, 22], or altered expression of transcription factors that modulate cellular activity [23]. These multiple observations highlight the complexity of the mechanisms underlying the development of sympathetic overactivity in neurogenic hypertension and indicate the necessity of additional investigation to identify their etiology.

Less attention has been given to how the quality of postnatal life impacts sympathetic control in adulthood. Brain circuitries can be shaped by environmental inputs experienced during critical periods of development [24]. The postnatal window represents a sensitive period where complex neural networks undergo a final differentiation and refinement through neurogenesis, formation or pruning of synaptic inputs, alterations in the balance of excitatory and inhibitory inputs, myelination, and other processes [25]. Evidence indicates that early life stressors can promote long-lasting changes through genetic and epigenetic mechanisms that affect system functioning and increase the propensity for developing diseases in adulthood [26, 27]. Premature babies and infants with sleep-disordered breathing (SDB) can experience recurrent apnea/hypopnea episodes due to breathing instability or airway obstruction [28–30], leading to exposure to intermittent hypoxia. This condition can negatively impact brain functions, especially during early postnatal life, causing motor impairments, cognitive delays, and respiratory irregularities [30]. Concerning the cardiovascular system, clinical evidence shows that children with SDB may develop high blood pressure levels associated with increased urinary catecholamines [31, 32]. Moreover, a longitudinal study reported that children with obstructive sleep apnea accompanied by intermittent hypoxia showed a higher risk of hypertension in adult life [33]. These clinical findings parallel experimental evidence showing that rodents exposed to intermittent hypoxia during postnatal age exhibited, in juvenile and adult life, elevated arterial pressure levels, augmented plasma catecholamines, and baroreflex dysfunctions [34, 35]. Therefore, intermittent hypoxia exposure during the postnatal period can promote long-term effects on the autonomic control of the cardiovascular system that persist until adult life. However, additional experiments with direct measurements of the sympathetic nerve activity are needed to assess this association. The lack of research and understanding of autonomic outcomes associated with intermittent hypoxic episodes during early life prevents the development of an evidence-based approach to their management.

In the present study, we employed a rat model of postnatal intermittent hypoxia to explore the changes in sympathetic activity accompanying elevated arterial pressure levels. We also evaluated the firing activity of presympathetic neurons in the RVLM and sought to identify potential mechanisms that may sustain the autonomic changes after exposure to intermittent hypoxia.

Methods

Animals and ethical approval

The experimental procedures comply with the Guide for the Care and Use of Laboratory Animals published by the Brazilian National Council for Animal Experimentation Control (CONCEA) and were approved by the Ethics Committee Care and Use of the School of Dentistry of Araraquara, São Paulo State University (protocol number CEUA 18/2014 and 30/2016). Neonate Holtzman rats, along with their mothers (300–320 g), were obtained from the Animal Care Unit of the São Paulo State University (UNESP) and maintained in polypropylene cages housed in a room with controlled temperature (23 ± 2°C), humidity (55 ± 10%), and light–dark cycle (12 hours, lights on 7:00 am), and free access to chow and water. The dams and offspring (8 males/mother) were divided into two experimental groups: a control group (total of 37 rats) and a postnatal chronic intermittent hypoxia (pCIH) group (total of 46 rats). The dam was kept with the pups until weaning (P21). As detailed below, the experiments were performed on juvenile (P21–P45; 60–100 g) and adult male rats (P90–P99; 300–340 g).

Postnatal chronic intermittent hypoxia

Animals were exposed to an intermittent hypoxia paradigm as previously described [7, 36]. After birth (P0–1), rat pups with their dams were housed in collective cages and placed inside plexiglass chambers that allowed the control of inspired oxygen (O₂) levels through a system of solenoid valves and O₂ sensors (Oxycycler, Biospherix, Lacona, NY, USA). This system allowed regulated injections of pure O₂ and nitrogen (N₂) into the chambers using appropriate software (Anawin 2, version 2.4.17). The conditions of temperature, humidity, light–dark cycle, and food/water access inside the chambers were kept as aforementioned. Rats from the control group were maintained under normoxia (20.8% O₂), while the pCIH group was exposed to cyclic periods of hypoxia (6% O₂ for 30–40 seconds) induced by injections of N₂ inside the chamber for 4 minutes. After this hypoxic period, O₂ was flushed inside the chamber to return and maintain at 20.8% for 5 minutes [37]. These 9-minute cycles occurred uninterruptedly for 8 hours daily (09:30–17:30 hours). During the remaining 16 hours, the animals of the control and pCIH groups were maintained under normoxic conditions. Gas injections were performed in the upper portion of the chambers to avoid additional stress caused by direct air jets. Animals were subjected to pCIH for 10 days (until P10). A separate group of animals (n = 7) were exposed to pCIH for 15 days (until P15); however, the cardiovascular and electrophysiological parameters evaluated were similar to animals exposed to pCIH for 10 days. Therefore, the animals treated with pCIH for 10 and 15 days were grouped into a single experimental group. After the pCIH exposure, the animals were maintained under normoxia until the experimental day.

Recordings of cardiovascular parameters

Under ketamine (80 mg/kg, i.p.) and xylazine (7 mg/kg, i.p.) anesthesia and in aseptic conditions, juvenile (P35–40) and adult (P90–99) pCIH and control animals had their femoral artery isolated for the implant of a polyethylene tube (PE-10 connected to a PE-50) filled with saline. In the adult groups, a catheter was also inserted into the femoral vein for systemic administration of the ganglionic blocker hexamethonium bromide (25 mg/kg, i.v.)—used to assess sympathetic vasoconstrictor tone [37]. Afterward, the catheters were exteriorized between the scapulae and fixed on the back of the animal. At the end of the surgery, the animals received the veterinary antibiotic (penicillin–streptomycin, 1 200 000 IU, 1 mg/kg, i.p.) and the analgesic/anti-inflammatory ketoprofen (1 mg/kg, s.c.). After recovering from anesthesia and regaining consciousness, the animals returned to the animal facility and were monitored for postsurgical pain or discomfort. On the next day, the animals were transferred to the experimental room and acclimated. The arterial catheter was then connected to a pressure transducer (Statham Gould, El Segundo, CA, USA) coupled to an amplifier (model ETH-200 Bridge Bio Amplifier, Chicago, IL, USA) to record the pulsatile arterial pressure (PAP). The pressure signals were acquired in a computer (2 kHz sampling rate) using a data acquisition system (Powerlab 16SP, ADInstruments, Colorado Springs, CO, USA) and appropriate software (LabChart, ADInstruments). Systolic arterial pressure (SAP), mean arterial pressure (MAP), heart rate (HR), and pulse interval (PI) values were derived from PAP. The cardiovascular parameters were recorded under unrestrained conditions for at least 30 minutes. After the experiments, the animals were euthanized with an overdose of anesthesia (urethane, i.p.).

Spectral analyses of arterial pressure and PI

Power spectral analyses of SAP and PI were performed as described previously [37, 38]. Briefly, SAP and IP beat-by-beat time series were assessed in the frequency domain using fast Fourier transform spectral analysis (CardioSeries v2.4—http://www.danielpenteado.com) and grouped into low- (LF: 0.20–0.75 Hz) and high-frequency bands (HF: 0.75–3.0 Hz) [39, 40]. LF oscillations in the SAP and PI are linked with the modulatory effects of sympathetic activity controlling vascular tonus and heart activity, while fluctuations at the HF range are associated with a respiratory or parasympathetic modulation of blood vessels and the heart, respectively [40, 41]. The power of the oscillatory components was expressed in absolute values (mm Hg² or ms²). Oscillations lower than 0.20 Hz were not analyzed.

Decerebrated arterially perfused rat in situ preparations

In situ preparations were obtained from juvenile (P21–25) pCIH and control rats as previously described [7, 36, 42]. In brief, the rats were deeply anesthetized with isoflurane (Cristália, São Paulo, Brazil) until the paw and tail pinch reflexes were abolished, transected sub-diaphragmatically and submerged in ice-chilled Ringer solution (in mM: NaCl 120; NaHCO₃ 24; KCl 3.75; CaCl₂ 2.50; MgSO₄ 1.25; KH₂PO₄ 1.25; and glucose 10). The animals were decerebrated precollicularly, and the skin, lungs, and diaphragm were carefully removed. In some preparations, the ventral medullary surface was exposed from the vertebral arteries to the pontine nuclei by removing the trachea, esophagus, muscles, connective tissues, the atlanto-occipital membrane, and the basilar portion of the occipital bone. Preparations were then transferred to a recording chamber, and a double-lumen catheter (DLR-4, Braintree Scientific, MA, USA) was inserted into the descending aorta for retrograde perfusion and pressure monitoring. Perfusion was supplied via a peristaltic roller pump (Watson Marlow 520S, Falmouth, UK) and consisted of a Ringer solution containing an oncotic agent (polyethylene glycol: 1.25%; Sigma, USA) and a neuromuscular blocker (vecuronium bromide, 40 mg/mL, Cristália, São Paulo, Brazil), warmed at 32°C, gassed with carbogen (95% O₂ and 5% CO₂—pH 7.4) and filtered using a nylon screen (pore size: 25 µm, Millipore, Billirica, MA, USA). The perfusion pressure was maintained in the range of 60–80 mm Hg by adding vasopressin (0.6–1.0 nM, Sigma, MO, USA) to the perfusate and adjusting the flow at 21–25 mL/min according to the size of the animal. A syringe attached to the perfusion system allowed intra-arterial injections of potassium cyanide (KCN, 0.05%, 50 µL) for activating peripheral chemoreceptors [43].

Nerve recordings and analyses

The left phrenic nerve (PN) and the right side of the thoracic sympathetic chain (tSN), at the level of T10–12, were isolated and cut distally for recordings of their electrical activity using glass suction bipolar electrodes held in a 3D micromanipulator (Narishige, Tokyo, Japan). In some preparations with the ventral medullary surface exposed, the hypoglossal nerve (HN) was recorded in replacement of PN to monitor the inspiratory activity. All signals were amplified (model P511, Grass Technologies, Middleton, USA), band-pass filtered (100 Hz to 3 kHz), and acquired in an A/D converter (CED 1401, Cambridge Electronic Design, Cambridge, UK) to a computer running Spike2 software (6 kHz, CED, Cambridge, UK). The rhythmic ramping pattern of PN and HN activities was used as a continuous physiological index of preparation viability [44]. We also determined the frequency and amplitude of PN bursts (expressed as bursts per minute, bpm, and µV, respectively). From the tSN, we analyzed the average levels of total activity and the activity during the inspiratory (coinciding with the PN bursts) and expiratory phases (between PN bursts). Values were expressed in absolute units (µV) after noise subtraction (obtained after the death of the preparation at the end of the experiments after turning the perfusion pump off). Variations of PN and tSN activities during peripheral chemoreceptor activation were expressed in absolute units (µV) or analyzed as percentage changes relative to baseline activity. All analyses were performed on rectified and smoothed (50 ms) signals and performed offline using Spike2 software with custom-written scripts.

Neuronal recordings and analyses

In preparations of control and pCIH rats with exposure of the ventral medullary surface, glass microelectrodes (10–30 MΩ) filled with KCl (1 M) were placed in the RVLM to perform extracellular single-unit recordings of presympathetic neurons. The microelectrodes were mounted in a 3D manipulator (Scientifica PatchStar Micromanipulator, UK) and positioned in the RVLM under visual control (binocular microscope MC 12MF; DFVasconcellos, Brazil) and using anatomical landmarks as references: 800–1100 µm caudal to the caudal end of the trapezoid body, 1500–1700 µm lateral to the midline (aligned with the rootlets of the HN), and 350–500 µm beneath the ventral surface. All the signals recorded were low-pass filtered (2 kHz, Neuroprobe Amplifier 1600 A-M System, EUA) and digitalized (10 kHz, CED 1401, CED) to a computer using the Spike2 software (CED). The presympathetic neurons were identified based on their excitatory response to peripheral chemoreceptor stimulation and the presence of collisions between spontaneous and antidromic potentials triggered by the electrical stimulation (100–200 mA, 1 ms pulses) of the IML cells in the thoracic spinal cord (T8–10) using a concentric electrode [45, 46]. Baseline and KCN-induced firing frequencies were analyzed offline (Spike2, CED) and compared between groups.

Fluorescent in situ hybridization and immunohistochemistry

The hypoxia-inducible factor 1 alpha (HIF-1α) mRNA expression in the RVLM C1 neurons was investigated in separate groups of animals not subjected to any surgical procedure. Rats from pCIH and control groups were initially anesthetized with isoflurane and then perfused transcardially with cold phosphate-buffered saline (PBS, 10 mM, pH 7.4, 100 mL/100 g BW) followed by paraformaldehyde (PFA, 4%, 100 mL/100 g BW). The brains were extracted and fixed in PFA for 12 hours at 4°C. Next, the brains were immersed and kept in 10% sucrose at 4°C until the tissue sank. This process was repeated with 20% and 30% sucrose. Then, the brains were frozen in Tissue Freezing Medium (Triangle Biomedical Sciences, Durham, NC, USA) using dry ice and maintained in airtight containers at −80°C until sectioning. The brains were sectioned at 20 µm in a cryostat, and sections containing the RVLM region (between −12.48 and −11.88 mm caudal to bregma) were mounted on microscope slides (Superfrost Plus, Fisher Scientific, Pittsburgh, PA, USA). Fluorescent in situ hybridization (RNAscope, Advanced Cell Diagnostics, Newark, CA, USA) was performed according to the manufacturer instructions (document #323100-USM, available at https://acdbio.com/documents/product-documents) using the kit RNAscope Multiplex Fluorescent Detection Reagents v2 (product #323110), the kit RNAscope H₂O₂ and Protease Reagents (product #322381), the RNAscope probe for HIF-1α (product #432281), and the TSA Cyanine 3 Plus Evaluation kit from PerkinElmer (product #NEL744E001KT, PerkinElmer, Boston, MA, USA). Immediately after completion of the RNAscope protocol, the immunofluorescence for tyrosine hydroxylase (TH) was performed to identify catecholaminergic neurons within the RVLM. Briefly, slides were incubated in a blocking solution (0.1 M PBS, 10% normal horse serum, and 0.3% Triton X-100) for 15 minutes and rinsed 3 × 10 minutes in 0.1 M PBS at room temperature. After, the slides were incubated with a primary antibody (Mouse anti-TH antibody, 1:1000, product #MAB5280, Millipore, Billerica, MA, USA) for 1 hour at room temperature plus 36 hours at 4°C. Next, the slides were rinsed with 0.1 M PBS and incubated with a secondary antibody (Alexa Fluor 488 donkey anti-mouse antibody, 1:200, product #R37114, Molecular Probes-Life Technologies, Eugene, OR, USA) for 4 hours at room temperature. Finally, slides were rinsed with 0.1 M PBS, the excess liquid was drained, a mounting medium was added (Fluoromount, product # F4680, Sigma, St. Louis, MO, USA), and slides were covered with glass coverslips (Fisherfinest, product #125485M, Fisher Scientific).

Images were acquired using a laser scanning confocal microscope (LSM800, Zeiss, Jena, Germany), and figures were prepared using the Zen 2 software (Blue edition, Zeiss). Analyses were performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Three sections per animal corresponding to three different rostrocaudal levels containing the RVLM region (−12.48 to −12.36 mm, −12.24 to −12.12 mm, and −12.00 to −11.88 mm caudal to bregma) were considered for analyses. RVLM level identification was based on anatomical references (facial nucleus, inferior olives, nucleus ambiguus, and the fourth ventricle) according to the rat brain atlas [47]. Quantifications were performed on ×20 magnification images within RVLM boundaries containing the highest number of TH-positive cell bodies (C1 neurons). The number of C1 neurons was manually counted, and the expression of HIF-1α mRNA (number of dots) was automatically determined following RNAscope manufacturer instructions (Advanced Cell Diagnostics, Tech note #TS 46-003, available at https://acdbio.com/documents/product-documents). For each section, equal size areas were considered for analyses.

Data analyses

The results are described and graphically represented as mean ± SD. The normal distribution of the data was initially verified with the Shapiro–Wilk normality test. The results about baseline cardiovascular parameters and spectral analyses, baseline tSN, PN frequency and neuron firing frequency, raw changes in tSN and RVLM activities during peripheral chemoreceptor activation, and TH and HIF-1α quantification were compared using Student’s unpaired t-test (parametric data). The results about baseline PN burst amplitude, PN burst frequency and amplitude, and percentage tSN change in response to peripheral chemoreceptor stimulation were compared using the Mann–Whitney test (nonparametric data). Significant differences were considered when p < 0.05. Statistical and graphic operations were performed using GraphPad Prism software (version 8, GraphPad, La Jolla, USA).

Results

pCIH elevates baseline blood pressure and increases sympathetic modulation in juvenile rats

Under baseline conditions, juvenile pCIH rats (n = 13) presented higher MAP levels (84 ± 7 vs 95 ± 5 mm Hg; p = 0.0002; Figure 1, A) than the control group (n = 11). Spectral analysis of the SAP demonstrated elevated powers of LF (2.06 ± 0.99 vs 3.22 ± 1.36 mm Hg; p = 0.0280; Figure 1, B) and HF oscillatory components (0.93 ± 0.32 vs 2.33 ± 1.30 mm Hg; p = 0.0023; Figure 1, C) in the pCIH group than in controls, indicating an excessive modulation of the blood vessels by the sympathetic activity and breathing in the pCIH group. Regarding heart activity, we observed a significant increase in the baseline HR of pCIH rats (402 ± 24 vs 459 ± 26 bpm; p < 0.0001; Figure 1, D) compared to controls. This tachycardic condition in the pCIH-treated animals was associated with no changes in the power of LF variability (0.54 ± 0.44 vs 0.51 ± 0.34 ms; p = 0.7645; Figure 1, E) and reduced power of HF oscillations of the PI (2.02 ± 1.14 vs 1.21 ± 0.41 mm Hg; p = 0.9072; Figure 1, F) relative to the control rats, suggesting a decrease in the parasympathetic-related modulation of resting HR.

Figure 1.

Postnatal intermittent hypoxia increases arterial pressure and heart rate and causes an autonomic imbalance in juvenile rats. Average levels of baseline mean arterial pressure (MAP, panel A), low- (LF, panel B) and high-frequency variabilities (HF, panel C) of systolic arterial pressure, baseline heart rate (HR, panel D), and LF (panel E) and HF variability (panel F) of pulse interval of juvenile animals maintained under normoxia (control, n = 11) or exposed to intermittent hypoxia during the postanal period (pCIH, n = 13). * different from the control group, P < 0.05.

pCIH induces excessive sympathetic discharge and presympathetic neuronal activity

Recordings of the electrical activity of the thoracic sympathetic chain revealed an augmented baseline tSN outflow in in situ preparations from the pCIH rats (n = 14) compared to controls (n = 7) (5.2 ± 1.5 vs 11.5 ± 6.3 µV; p = 0.0200; Figure 2, A and D). The sympathetic overactivity in pCIH animals was accompanied by a larger PN burst amplitude relative to controls (22.3 ± 12.1 vs 64.2 ± 45.9 µV; p = 0.0292; Figure 2, B), while PN burst frequency was similar between groups (18 ± 4 vs 22 ± 7 bpm; p = 0.2142, Figure 2, C). In both pCIH and control groups, the tSN exhibited a rhythmic activity pattern with increased discharge during the inspiratory phase (coincident with the PN bursts, Figure 2, D). The magnitude of these respiratory-related sympathetic bursts (ΔtSN: 6.7 ± 3.6 vs 14.2 ± 8.25 µV; p = 0.0346) and the average tSN levels during the inspiratory phase (6.2 ± 2.7 vs 12.5 ± 6.8 µV; p = 0.0308; Figure 2, E) were higher in pCIH than in control preparations. However, the mean tSN during the expiratory phase (between PN bursts) was also increased in the pCIH animals compared to controls (5.4 ± 2.1 vs 12.2 ± 5.8 µV; p = 0.0084; Figure 2, F), indicating that pCIH induced a global amplification of baseline sympathetic activity in in situ preparations of juvenile rats.

Figure 2.

Excessive sympathetic nerve activity in juvenile rats exposed to postnatal intermittent hypoxia. Panel A: traces of raw and integrated (∫) phrenic (PN) and thoracic sympathetic nerves (tSN) from representative in situ preparations of a rat maintained under normoxia (control) or exposed to intermittent hypoxia during the postanal period (pCIH). Panels B-F: average values of PN burst frequency and amplitude, mean tSN levels, and mean tSN activity during the inspiratory and expiratory phases in in situ preparations of control (n = 7) and pCIH rats (n = 14). * different from the control group, P < 0.05.

Spinally projecting presympathetic neurons were identified in the RVLM region of the control (n = 6 neurons from 6 preparations) and pCIH (n = 7 neurons from 7 preparations) in situ preparations. The recorded neurons showed a tonic (nonphasic) firing pattern (Figure 3, A and B). The average firing frequency of RVLM neurons during baseline conditions was significantly higher in preparations of the pCIH group compared to controls (10.3 ± 5.1 vs 24.8 ± 11.2 Hz; p = 0.0140; Figure 3, C). Moreover, the activity of the RVLM neurons was more irregular in the pCIH than in the control group (frequency SD: 4.3 ± 1.8 vs 13.8 ± 5.8 Hz; p = 0.0029; Figure 3, D).

Figure 3.

Increased firing frequency of presympathetic neurons of juvenile rats exposed to postnatal intermittent hypoxia. Panel A: traces from representative in situ preparations of a rat maintained under normoxia (control) or exposed to intermittent hypoxia during the postanal period (pCIH) illustrating the pattern of firing frequency of spinally projecting presympathetic neurons of the RVLM. Panel B: example of a collision test (marked by the asterisk) evoked by the electrical stimulation of the thoracic spinal cord (triangle). Panels C and D: average values of firing frequency and frequency variability of RVLM neurons recorded in animals from the control (n = 6 neurons from 6 preparations) and pCIH (n = 6–7 neurons from 7 preparations). * different from the control group, p < 0.05.

Sympathetic responses to peripheral chemoreceptor stimulation

The activation of peripheral chemoreceptors with intra-arterial injections of KCN caused transient increases in the PN, tSN, and RVLM activities in both groups (Figure 4, A). The average values of tSN during peripheral chemoreceptor stimulation were larger than baseline in control (5.3 ± 2.9 vs 8.4 ± 3.7 µV; p = 0.0026, n = 7; Figure 4, B) and pCIH groups (12.3 ± 5.8 vs 19.02 ± 9.5 µV; p = 0.0005, n = 14; Figure 4, C). The maximal tSN evoked by the activation of the peripheral chemoreceptors was higher in pCIH than in controls (p = 0.0084). However, the magnitude of the sympatho-excitatory response relative to baseline was similar between groups (ΔtSN: 85 ± 42% vs 81 ± 53%; p = 0.4880; Figure 4, D). Likewise, the firing frequency of presympathetic neurons of the RVLM during the peripheral chemoreceptor activation was higher in the pCIH preparations (n = 7 neurons from 7 preparations) compared to controls (n = 5 neurons of 5 preparations) (25.2 ± 6.7 vs 42.9 ± 15.0 Hz; p = 0.0348; Figure 4, E), but the magnitude of the response was similar between groups (Δfrequency: 14.4 ± 3.9 vs 18.1 ± 8.7 Hz; p = 0.3973; Figure 4, F). Regarding the PN activity, the increases in burst frequency (ΔPN freq: 14 ± 7 vs 18 ± 12 bpm; p = 0.7017, Figure 4, G) and amplitude (ΔPN amp: 22 ± 18% vs 39 ± 39%; p = 0.2177) were similar between control and pCIH groups.

Figure 4.

Sympatho-excitatory response evoked by peripheral chemoreceptor stimulation in juvenile animals exposed to postnatal intermittent hypoxia. Panel A: traces from representative in situ preparations of a rat maintained under normoxia (control) or exposed to intermittent hypoxia during the postanal period (pCIH), illustrating the increase in the raw and integrated (∫) thoracic sympathetic discharge (tSN, top traces) and presympathetic neuronal activity (bottom traces) in response to peripheral chemoreceptor stimulation with KCN (arrows). Panels B–D: average values of tSN activity before and during peripheral chemoreceptor activation with KCN, and variation (Δ) of tSN activity to KCN in in situ preparations of the control (n = 7) and pCIH (n = 14) groups. Panels E and F: average values and variation (Δ) of firing frequency of spinally projecting RVLM neurons during peripheral chemoreceptor activation with KCN in control (n = 5 neurons) and pCIH (n = 7 neurons) in situ preparations. Panel G: KCN-induced variation of PN burst frequency in control and pCIH in situ preparations. * different from the control group, p < 0.05.

pCIH increases HIF-1α mRNA expression in the C1 neurons of juvenile rats

HIF-1α overexpression in target cells is commonly associated with functional adaptations in response to hypoxic conditions [48, 49]. A previous study revealed that the presympathetic neurons of the RVLM exhibiting a tonic (nonphasic) discharge pattern (as recorded in the present study) express TH [14]. Therefore, we combined immunofluorescence with fluorescent in situ hybridization to evaluate the expression of HIF-1α mRNA in C1 (TH-positive) neurons of the RVLM of juvenile control and pCIH-treated animals (Figure 5). The average number of TH-positive neurons per section in the RVLM region was not different between control (n = 5) and pCIH rats (n = 5) (13 ± 4 vs 11 ± 4 neurons/section; p = 0.1465). The HIF-1α expression in the caudal regions of the RVLM of pCIH rats (−12.48 to −12.36 mm: 2430 ± 683 vs 3688 ± 1007 dots; p = 0.0497; Figure 5, A–C; −12.24 to −12.12 mm: 3050 ± 480 vs 3972 ± 693 dots; p = 0.0403; Figure 5, D–F), but not in the most rostral area (−12.00 to −11.88 mm: 3211 ± 1600 vs 3928 ± 1374 dots; p = 0.4688; Figure 5, I), was significantly greater compared to the expression in control animals. The HIF-1α overexpression in the RVLM of the pCIH rats markedly overlapped with the TH staining (Figure 5, A₃–H₃), suggesting an accumulation of HIF-1α mRNA in C1 neurons.

Figure 5.

Postanal intermittent hypoxia increases HIF-1α expression in presympathetic neurons of the rostral ventrolateral medulla. Panels A–I: representative images showing the immunofluorescence for TH (green), the fluorescent in situ hybridization for HIF-1α mRNA (red) and the merge of C1 and HIF-1α (yellow) in three different levels of the rostrocaudal levels of the rostral ventrolateral medulla of animals maintained under normoxia (control) or exposed to intermittent hypoxia during the postanal period (pCIH). Panels A₃–G₃ represent the magnification of the area outlined in panels A₂–G₂ (white square). The graphs on the right indicate the average HIF-1α mRNA expression (dots) in the RVLM region of the control (n = 5) and pCIH rats (n = 5). * different from the control group, p < 0.05. Panel J: low-magnification photomicrographs with their corresponding schematic images showing the C1 neurons (green) in the three rostrocaudal levels of the RVLM considered for analyses. Scale bars—panels A_1–2–H_1–2: 100 µm; panels A₃–H₃: 100 µm; panel J: 250 µm. Magnifications—panels A_1–2–H_1–2: ×20; panels A₃–H₃: ×40; panel J: ×5.

Increased arterial pressure and elevated sympathetic vasoconstrictor tone in adult animals exposed to pCIH

At adult age (P90–99), pCIH-treated rats (n = 6) exhibited higher MAP (121 ± 9 vs 103 ± 10 mm Hg; p = 0.0033; Figure 6, A) and similar HR (395 ± 47 vs 354 ± 45 bpm; p = 0.1236; Figure 6, B) when compared to the age-matched control group (n = 8). The systemic injection of hexamethonium markedly lowered the MAP and increased HR levels in both groups. The depressor response observed after the ganglionic blockade was larger in the pCIH rats (n = 5) than in controls (n = 8) (ΔMAP: −58 ± 4 vs −45 ± 7 mm Hg; p = 0.0019; Figure 6, C), indicating a greater dependence on sympathetic activity to maintain arterial pressure levels in the pCIH group. In contrast, the increase in HR was similar between groups (ΔHR: 34 ± 32 vs 58 ± 43 bpm; p = 0.2979; Figure 6, D).

Figure 6.

Adult animals exposed to pCIH exhibit increased arterial pressure levels and elevated vascular sympathetic tone. Panels A and B: average values of baseline MAP and HR of adult animals maintained under normoxia (control, n = 8) or exposed to intermittent hypoxia during the postanal period (pCIH, n = 6). Panels C and D: magnitude of the depressor (ΔMAP) and tachycardic (∆HR) responses to ganglionic blockade with hexamethonium (25 mg/kg, i.v.) in control (n = 8) and pCIH groups (n = 5). * different from the control group, p < 0.05.

Discussion

During postnatal life, several physiological functions undergo a final development and maturation at the cellular and systems levels, ultimately shaping their sensitivity and gain in adulthood. Because of that, the postnatal period also represents a critical window of vulnerability where exposure to insults can introduce plasticity and promote sustained functional changes affecting adult life. Our study provides direct evidence showing that postnatal intermittent hypoxia dysregulates the control of sympathetic activity to the cardiovascular system and increases the expression of a hypoxia-induced transcription factor in medullary presympathetic neurons. These new findings advance our understanding of the negative impacts of hypoxia during the postnatal age on the control of autonomic nervous system activity, contributing to the development of life-threatening CVD in adulthood.

Our first relevant result was that neonatal rats exposed to intermittent hypoxia exhibited higher MAP and HR levels in early adult life. This finding confirms a previous observation [34] and supports the notion that pCIH is a risk factor for developing chronic high blood pressure. We found that either 10 or 15 days of exposure to pCIH produced comparable cardiovascular alterations—an observation also reported for the development of respiratory irregularities and respiratory motor dysfunction [36]. Therefore, the initial stages of postnatal life may represent a period of higher vulnerability of the cardiorespiratory systems. The increased blood pressure levels induced by pCIH were associated with an augmented vasoconstrictor sympathetic outflow. The direct measurements of sympathetic nerve activity were performed in decerebrated in situ preparations, in conditions free of the depressant effects of anesthetics, where the cardiorespiratory peripheral afferent feedbacks were absent or silent, and the perfusion pressure was controlled at constant levels [50]. Although these features can be interpreted as technical limitations, they indicate that the sympathetic changes observed in the pCIH rats likely result from alterations in the functioning of brainstem circuitries. Notably, the sympathetic overactivity in situ correlated with the increased power of the sympathetic-mediated oscillations of arterial pressure of freely behaving juvenile pCIH rats, supporting the in situ data and confirming the impact of pCIH-induced high sympathetic traffic on the control of vascular resistance in vivo. Like the juvenile rats, the adult pCIH animals showed higher arterial pressure levels than the controls. These pCIH rats also displayed a larger depressor response to hexamethonium, a ganglionic blocker that interrupts the sympathetic transmission to blood vessels, showing that the sympathetic overactivity state was still present in adulthood. Therefore, our results indicate that the centrally mediated, excessive sympathetic drive critically contributed to increasing arterial pressure levels and introducing variability in the systemic blood flow of pCIH rats. Both conditions have deleterious effects as they favor the development of cardiovascular dysfunctions and end-organ damage [51, 52].

The ventral surface of the medulla oblongata houses neurons that are necessary for the generation of sympathetic outflow [11]. We identified an enhanced resting activity and firing variability in presympathetic neurons in the RVLM of pCIH rats, which parallel the findings of baseline sympathetic overactivity and augmented sympathetic-related oscillations of arterial pressure. The RVLM activity is markedly modulated by afferent signals, including from the carotid bodies [9]. A previous study demonstrated that juvenile pCIH exhibited higher ventilation and carotid body sensory response during hypoxia [34], suggesting that enhanced carotid body activity contributes to maintaining high blood pressure in pCIH. In our experiments, we identified that the sympathetic and RVLM activities remained elevated during the stimulation of the carotid bodies with KCN. However, the magnitude of these responses was comparable between pCIH and control groups, suggesting a similar peripheral chemoreflex gain. Based on our current data and the fact that the in situ preparations are under hyperoxia (a condition that silences the carotid body chemoreceptors), we speculated that the excessive sympathetic activity to the cardiovascular system in pCIH rats is primarily driven by RVLM hyperactivity leading to sympathetic nervous system dysfunction. However, our data do not rule out that pCIH-induced changes in the carotid body sensitivity, as reported previously [34], may also contribute to activating the sympathetic nervous system in pCIH rats and promoting a further increase in arterial pressure level.

The respiratory network is a significant source of central modulatory inputs to the RVLM neurons. Previous studies reported distinct firing patterns of RVLM neuronal activity according to the respiratory cycle, such as inspiratory or expiratory modulated, as well as neurons that are not modulated by respiration [14]. Collectively, the activity of these neurons contributes to generating bursts of sympathetic activity during the inspiratory/early expiratory phases, as reported in our experiments. We identified an increased respiratory modulation of arterial pressure (HF oscillations) in pCIH rats in vivo and larger inspiratory-related sympathetic bursts in pCIH rats in situ, suggesting a strengthening of respiratory synaptic drive to the RVLM. However, the sympathetic activity levels during expiration were also heightened in the pCIH group, indicating potentiation of the sympathetic drive during all phases of the respiratory cycle. Therefore, we speculate that the overall RVLM activity is amplified by pCIH. In agreement with these observations, we found that the activity of RVLM neurons showing a tonic firing pattern was increased in the pCIH rats. We cannot exclude that the RVLM neurons with respiratory-modulated firing patterns in pCIH rats might also exhibit changes in their activity since pCIH promotes relevant respiratory disturbances [36]. However, our data indicate that the heightened sympathetic outflow in pCIH rats was associated with a sustained higher firing frequency of tonic (nonphasic) RVLM neurons, suggesting that this population is a target of plasticity during pCIH exposure.

It is well described that HIF-1α is a key regulator of several transcriptional cascades that leads to cellular and systemic adaptations in conditions of hypoxia [48]. In contrast, dysregulated expression of HIF-1α can be associated with maladaptive modifications [49]. Increased HIF-1α expression in glomus (carotid bodies) and adrenal chromaffin cells of adult animals exposed to chronic intermittent hypoxia is associated with enhanced hypoxia sensitivity and potentiated catecholamine release [53, 54]. We found a higher HIF-1α mRNA expression in the RVLM of pCIH rats, especially in the C1 (catecholaminergic) population. The fact that neurons in the RVLM with a tonic firing activity may contain TH (hence are part of the C1 group) [14], and that this population showed a higher firing frequency in pCIH animals, suggests a connection between the enhanced RVLM activity and HIF-1α overexpression in pCIH-treated rats. Although the HIF-1α expression is tightly linked with the O₂ levels, maintaining its high levels in the RVLM of pCIH rats under normoxic conditions may rely on epigenetic mechanisms, such as DNA hypermethylation of genes encoding HIF-1α-regulating proteins [34]—a possibility that awaits future experiments to be elucidated. Moreover, HIF-1α can modify the transcription of a wide range of genes [48]. Based on the evidence in the literature, we speculate that HIF-1α may activate genes producing enzymes associated with oxidative stress control (e.g. Sod2 encoding manganese superoxide dismutase), leading to increased reactive oxygen species (ROS) formation and augmented neuronal activity—a mechanism mediating the increased excitability of carotid body glomus cells after CIH exposure [34, 54] and the potentiated blood pressure response of pCIH-treated spontaneously hypertensive rats [55]. However, the contribution of ROS-related and other HIF-1α target genes to increasing the activity and synaptic response of RVLM C1 neurons of pCIH rats requires additional experiments to be confirmed. We also observed the presence of HIF-1α mRNA in cells at the RVLM level of pCIH rats that were not positive for TH, which could represent other presympathetic neurons (non-C1) or respiratory neurons [14, 56]. These different cell types expressing HIF-1α could also contribute to increasing sympathetic drive or promoting respiratory irregularities and motor dysfunction [36]. Therefore, we propose that HIF-1α overexpression in the RVLM represents a potential mechanism underlying the occurrence of sympathetic overactivity in juvenile rats exposed to intermittent hypoxia during the postnatal period.

Our analysis of cardiovascular variability in vivo also demonstrated that the high HR in juvenile rats exposed to pCIH was associated with suppressed parasympathetic-related modulation of cardiac activity. In our previous study, we found that central vagal activity, measured at the cervical level, is reduced in animals exposed to pCIH [36]. Although parasympathetic and respiratory motor-related activities can be recorded in the vagus nerve at this level, the finding of diminished vagal efferent activity in pCIH rats agrees with our HR variability data, indicating that pCIH also affects the regulation of cardiac parasympathetic activity. Considering evidence showing that pCIH during the first 30 days of life does not cause anatomical disruptions of the cholinergic neurons of the nucleus ambiguus—a major vagal motor nucleus [57], we speculate that changes in the excitability or synaptic drive to preganglionic parasympathetic neurons mediated the reduced parasympathetic outflow to the heart. It is also possible that pCIH can increase inhibitory neurotransmission or reduce excitatory neurotransmission to heart-innervating neurons in neurons of the nucleus ambiguus or dorsal motor nucleus of the vagus—another important brainstem source of preganglionic vagal neurons—as documented in adult animals exposed to chronic intermittent hypoxia [58]. Nonetheless, further experiments are required to determine the mechanisms underlying the pCIH-induced HR changes in juvenile rats.

In summary, our study demonstrates the impact of intermittent hypoxia exposure during the postnatal period on the autonomic control of arterial pressure levels. Apneas/hypopneas episodes during postnatal age, as seen in children with obstructive or central sleep apnea, conditions affecting 1%–6% of children between 2 and 8 years old [59, 60], are risk factors for the development of life-threatening CVD in adulthood [33]. Our findings indicate that pCIH promotes a long-lasting sympathetic nervous system dysfunction, leading to heightened sympathetic outflow to blood vessels in adult life. The link between augmented presympathetic neuronal activity and increased expression of HIF-1α in the C1 neurons of the RVLM suggests that genomic-associated mechanisms underlie the maintenance of sympathetic overactivity and high blood pressure after pCIH exposure. Additional investigation exploring the target genes and regulatory mechanisms modulated by HIF-1α in the RVLM neurons will be required to expand our comprehension and consider this pathway a potential therapeutic target.

Funding

This work was funded by the São Paulo Research Foundation (FAPESP; grants 2019/11196-0; 2020/05045-6; 2013/17251-6) and the National Council for Scientific and Technological Development (CNPq, grant, 303481/2021-8).

Author Contribution

All authors have contributed to the experimental design, data collection, and interpretation. The authors have read and approved the submitted version of the manuscript.

Disclosure Statement

None declared.

Author notes

Marlusa Karlen-Amarante. Present address: Center for Integrative Brain Research, Seattle Children’s Research Institute, Seattle, USA.