Effects of Chronic Cholinergic Stimulation Associated With Aerobic Physical Training on Cardiac Morphofunctional and Autonomic Parameters in Spontaneously Hypertensive Rats

Camila B. Gardim Ribeirão Preto Medical School, University of São Paulo Ana Catarine V. Oliveira Ribeirão Preto Medical School, University of São Paulo Bruno Augusto Aguilar Ribeirão Preto Medical School, University of São Paulo Stella V. Philbois Ribeirão Preto Medical School, University of São Paulo Hugo C. D. Souza (  hugocds@fmrp.usp.br ) Ribeirão Preto Medical School, University of São Paulo


Introduction
Systemic arterial hypertension is accompanied by important cardiovascular autonomic impairments, characterized by a decrease in barore ex sensitivity (BRS) and changes in heart rate (HRV) and blood pressure variability (BPV) 1,2 . These autonomic impairments are associated with cardiovascular morphological and functional changes, and when they are not identi ed and treated promptly, might contribute to the development of heart failure, reducing quality and life expectancy.
Thus, the search for new pharmacological and non-pharmacological therapies with cardioprotective actions is of great relevance to the aim of preventing and/or reversing the development and evolution of hypertension [3][4][5] . In this case, clinical and experimental studies have demonstrated that pyridostigmine bromide, a drug regularly used to treat myasthenia gravis, promotes cardiac and cardiovascular autonomic bene ts by reducing the heart rate (HR) at rest 6,7 and enhancing HRV [8][9][10][11] . On the other hand, lifestyle modi cations, such as regular physical exercise, also promote important bene ts in cardiovascular autonomic control, in addition to positive cardiac remodulation 4,12 .
In fact, some cholinergic stimulation effects of pyridostigmine bromide on cardiac autonomic control are similar to those observed in aerobic physical training, both in human and experimental models 5,10,13,14 . An experimental study in spontaneously hypertensive rats (SHRs) showed that cholinergic stimulation with pyridostigmine bromide for two weeks reduced blood pressure (BP) and HR.
However, the ejection fraction, a parameter evaluated by radioisotopic ventriculography, was also reduced. Additionally, autonomic parameters showed an increase in vagal participation in autonomic tonic balance and a reduction in systolic BPV 9 .
The hemodynamic and autonomic improvements are important and suggest a bene cial effect; however, the reduction of ejection fraction is worrying and might be related to the pyridostigmine bromide dosage used in the study (25 mg/kg/day), or perhaps, to the methodology employed to investigate the ejection fraction. However, with only one functional parameter evaluated, it was not possible to precisely state the real effects of pyridostigmine bromide on cardiac function. In this case, our hypothesis was that treatment with low doses of pyridostigmine bromide at 5 mg/kg/day and 15 mg/kg/day in SHRs would promote positive hemodynamic and cardiovascular autonomic effects, in addition to changes in morphological and functional cardiac parameters but would not affect the ejection fraction. In turn, the association with aerobic physical training will promote a catalytic effect, possibly increasing the gains obtained with pyridostigmine bromide.

Animals and procedures
Fifty-four SHR 18-week-old subjects were divided into two treatment groups: untrained (n = 27) and trained (n = 27). Each group was subdivided into three smaller groups (n = 9): vehicle (H 2 O), pyridostigmine bromide (Sigma-Aldrich, Saint Louis, MO, USA) diluted in drinking water at a dose of ~5 mg/kg/day, and pyridostigmine bromide diluted in drinking water at a dose of ~15 mg/kg/day for 2 weeks, between the 11 th and 12 th weeks of aerobic physical training. The dose was determined based on the results of a previous study 9 . During the experiments, the animals were housed in the Animal Facility at the Ribeirão Preto Medical School, which was maintained at 23°C and 60%-70% humidity. The rats were kept on a 12/12-hour light/dark cycle and had free access to food and water. The experimental protocols used in the present study were in accordance with the ethical principles of animal experimentation adopted by the Brazilian College of Animal Experimentation and were evaluated and approved by the Animal Experimentation Ethics pulsatile BP and to administer drugs, respectively. The catheters were subcutaneously tunneled and exteriorized in the nape. Twentyfour hours after the surgical procedures, BP was measured in conscious rats kept in a quiet environment. The BP was recorded using a pressure transducer (MLT0380; AD Instruments, Bella Vista, Australia), and the ampli ed signal (ML110; AD Instruments, Bella Vista, Australia) was fed to a computer acquisition system (PowerLab 8/30; AD Instruments, Bella Vista, Australia). The mean BP (MBP) and HR were calculated from the pulsatile BP.

Physical Training
The rats in the training groups underwent a protocol of aerobic physical training that consisted of swimming sessions in a glass tank (100 cm long × 80 cm wide × 80 cm high), which allowed for the simultaneous training of six animals. The tank was lled with 50 cm of warm water (30 ± 2°C), which was changed after every group training session. The training program was conducted in two different stages over a total of 12 (from 18 to 30 weeks of age). The rst stage consisted of a two-week adaptation period, during which the session length was gradually increased from 5 to 30 min per day, ve times per week (in increments of 5 min per day). The second stage consisted of 10 weeks, with 30 min of physical training sessions conducted ve times per week 15 . To evaluate physical training intensity, blood was collected from the tail veins of the animals at the fourth, seventh, and tenth weeks immediately before and after the 30-minute exercise sessions, and the lactate concentration was measured (Accutrend® Plus, Roche Diagnostics, Mannheim, Germany). The expected lactate level ranged from 5.5 to 6 mmol/L, as previously determined. If the animals did not achieve the expected lactate concentration, the level of training exertion was increased by fastening a leaded, impermeable Velcro strap to the chest to increase body weight by 2%-6% 16 . Echocardiography At 30 weeks of age, all animals underwent an echocardiographic evaluation. We used a Vevo 2100® High-Resolution Imaging System ultrasound (VisualSonics, Toronto, ON, Canada) instrument with a high-resolution transducer (21 MHz). For the procedure, the anterior regions of the thorax were previously trichotomized (Veet®, Reckitt Benckiser, São Paulo, SP, Brazil), and all animals were anesthetized with 1.5% iso urane supplemented with 1% O 2 and placed on a heated (37°C) platform. Echocardiography and temperature measurements were monitored.
High-resolution B-mode and M-mode images were acquired. Wall thickness and left ventricle dimensions were obtained from a shortaxis view at the level of the papillary muscles. Diastolic measurements were performed at the point of maximum cavity dimension, and systolic measurements were performed at the point of minimal cavity dimension. All measurements were performed according to the standards of the American Society of Echocardiography and by an evaluator who was blinded to which group the rats were assigned at the time of measurement 17  Surgical procedure Forty-eight hours after echocardiography, the rats were anesthetized with ketamine (80 mg/kg) and xylazine (10 mg/kg), and polyethylene catheters made in our laboratory (PE-50 spliced by melting to PE-10; Intramedic, Clay Adams, Parsippany, New Jersey, USA) were implanted into the left femoral artery and vein. Catheters were tunneled subcutaneously and exteriorized at the nape. To prevent blood clotting, the catheters were lled with heparinized saline solution (500 IU/mL). The rats were then allowed to recover for 24 hours before the cardiac sympathovagal assessment protocol, which was performed without anesthesia.

Experimental protocols
Heart rate variability and systolic blood pressure variability Twenty-four hours after the surgical procedure, arterial pulse (AP) pressure was measured in conscious rats kept in a quiet environment. BP was recorded using a pressure transducer (MLT0380, ADInstruments). Additionally, the ampli ed signal (ADInstruments ML110) was fed to a computer acquisition system (LabChart 7 Pro). MBP and heart rate (HR) were calculated from the arterial pulse pressure.
HR (pulse interval) and systolic BPV analysis were performed using custom computer software (CardioSeries v2.4, Dias, DPM, University of São Paulo, Brazil http://sites.google.com/site/cardioseries) 20 . The software was designed to perform time-frequency analysis of cardiovascular variability, allowing precise adjustment of parameters related to frequency domain analysis (e.g., interpolation rate, segment length, and boundaries of frequency bands). Because the software does not perform data sampling, a beat-by-beat time series must be generated and loaded into the CardioSeries software. The BP baseline and pulse interval (PI) series, obtained from 60-minute recordings, were processed using computer software (LabChart v7.0, ADInstruments, Bella Vista, Australia) that applies an algorithm to detect cycle-to-cycle in ection points in the pulsatile BP signal, thus determining beat-by-beat values of systolic blood pressure. Beat-by-beat PI series were generated from the pulsatile BP signal by measuring the time interval between adjacent systolic peaks. Next, a beat-by-beat series of PI and SBP were converted to data points every 100 ms using cubic spline interpolation (10 Hz). The interpolated series were divided into half-overlapping sequential sets of 512 data points (51.2 seconds), which were tested for stationarity. It is important to mention that cardiovascular variability analysis requires at least a weakly stationary data series (i.e., mean and covariance stable over time; Berntson et al., 1997). Data series stationarity can be veri ed by means of stationarity tests (i.e., enhanced reproducibility of the results among users and laboratories 22 , as well as through visual inspection of the data series 23,24 . In our study, a well-experienced researcher visually inspected the segments of interpolated time series (i.e., PI or SBP values) searching for transients that could affect the calculation of the power spectral density (PSD). To con rm that the visual inspection of the time series was properly performed, a Hanning window was used to attenuate side effects, and the spectrums of all segments were calculated using a direct fast Fourier transform algorithm for discrete time series. All segments were visually inspected for abnormal spectra. Finally, considering the results from the time series and spectra inspections, non-stationary data were not taken into consideration for PSD calculation. The spectra were integrated in low-frequency (LF; 0.2-0.75 Hz) and high-frequency (HF; 0.75-3 Hz) bands, and the results were expressed in absolute (ms 2 or mmHg 2 ) and normalized units (nu). The normalized values were created by calculating the percentage of LF and HF power of the total spectrum power minus the very low frequency band (VLF; < 0.2 Hz) and power 24,25 . To assess the sympathovagal balance, the LF/HF ratio of PI variability was calculated 26 .
Spontaneous barore ex sensitivity Barore ex sensitivity (BRS) was assessed in the time domain using the sequence technique, as described by Di Rienzo et al. (1985). Custom computer software (CardioSeries v2.4, http://sites.google.com/site/cardioseries) scanned the beat-by-beat time series of SBP and PI, searching for sequences of at least four consecutive beats in which increases in SBP were followed by PI lengthening (up sequence) and decreases in SBP followed by PI shortening (down sequence) with a linear correlation higher than 0.8. The slope of the linear regression lines between SBP and PI was used as a measure of spontaneous BRS 28 .

Assessment of cardiac sympathovagal balance
The in uence of sympathetic and parasympathetic autonomic tone on HR was assessed by administering propranolol (5 mg/kg, intravenous [i.v.], Sigma-Aldrich, USA) and methylatropine (4 mg/kg, i.v.; Sigma-Aldrich, USA), respectively. For this purpose, the femoral artery catheter was attached to a pressure transducer (MLT844, AD Instruments, Bella Vista, Australia), which converts the AP uctuations into electrical signals. Next, signals were ampli ed using a bridge ampli er (FE117, AD Instruments, Bella Vista, Australia), and pulsatile AP was continuously sampled (2 kHz) using a computer equipped with an analog-digital interface (ML866, AD Instruments, Bella Vista, Australia). After 60 minutes of basal HR recording, methylatropine was injected into half of the rats in each group, and HR was recorded for the following 15 min to assess the effect of vagal blockade on HR. Propranolol was then injected into the same rats, and HR was recorded for another 15 min to determine the intrinsic HR (IHR). In the other half of the rats, the methylatropine-propranolol sequence was reversed to assess the effect of sympathetic blockade on HR, following the same recording procedure (15 min each) for each drug, as described previously, to determine the IHR. The data from methylatropinepropranolol and propranolol-methylatropine sequences were pooled to provide the basal HR (before any drugs) and the IHR (after drugs).

Statistical analysis
The results are presented as mean ± standard error of the mean (SEM). The effects of hypertension and pharmacological treatments were assessed using two-way analysis of variance (ANOVA). When appropriate, post-hoc comparisons were performed using the Student-Newman-Keuls test. For comparison between two groups, the Student's t-test for independent measures or the Mann-Whitney Rank Sum test was used as required. Differences were considered signi cant at P < 0.05. All statistical tests were performed using SigmaStat software (version 3.5; Systat Software Inc., San Jose, CA, USA).

Baseline parameters
Treatment with pyridostigmine bromide at a dosage of 15/mg/kg/day, associated with aerobic physical training, promoted an increase in body weight. Both aerobic physical training and pyridostigmine bromide reduced the BP and HR values. When associated, the reduction in SBP and diastolic BP (DBP) was more prominent.
Double pharmacological blockade with atropine and propranolol-sympathovagal balance Table 1 shows the HR values after administration of atropine and propranolol, as well as the pacemaker intrinsic HR after administration of both drugs. Aerobic physical training and pyridostigmine bromide increased the HR response after atropine administration and reduced the HR response after propranolol administration. This association enhanced the response. Figure 1 shows the cardiac autonomic tonic balance in percentage values for each group.
Aerobic physical training and treatment with pyridostigmine bromide reduced the predominance of sympathetic autonomic components in determining the baseline HR. However, only the association of physical training and pyridostigmine bromide (15 mg/kg/day) promoted an inversion in cardiac autonomic tonic balance, characterized by a vagal predominance in determining the baseline HR.
Heart rate variability (HRV) and systolic blood pressure variability (BPV) Table 2 presents the results of HRV. The aerobic physical training increased the total variance, oscillation of LF in absolute units, and LF/HF ratio. In turn, the pyridostigmine bromide treatment increased the oscillation of LF and HF in absolute units. On the other hand, the association of both attenuated the oscillation of HF in absolute units and increased the LF/HF ratio. Table 2 also shows the BPV results. The aerobic physical training decreased the LF oscillation of BPV, whereas the pyridostigmine bromide treatment enhanced it. However, the association increased the values of LF oscillation even more than the pyridostigmine bromide isolated treatment.
Barore ex sensitivity (BRS) Table 2 shows the BRS values. Both aerobic physical training and pyridostigmine bromide increased the BRS to tachycardic responses induced by a reduction in BP (gain down, ms/mmHg). In turn, only pyridostigmine bromide treatment enhanced the BRS for bradycardic responses induced by an increase in BP (gain up, ms/mmHg).

Cardiac morphological and functional parameters
The cardiac morphological and functional results obtained using two-dimensional echocardiography are shown in Table 3.
Physical training increased LVEDD and LVESD, whereas pyridostigmine bromide treatment reduced RWT and IVST and enhanced LV mass. The association between physical training and pyridostigmine bromide treatment, mainly at a dose of 15 mg/kg/day, increased IVST, LVEDD, and LV mass.
Physical training increased LVEDV, LVESV, and ejection fraction, while pyridostigmine bromide treatment increased LVEDV and cardiac output. The association between the two treatments did not promote any additional effects. The pyridostigmine bromide treatment at a dose of 15 mg/kg/day promoted an increase in cardiac output and cardiac index, compared to the untrained vehicle group. In turn, the aerobic physical training groups had higher values of LVEDV and LVESV than the untrained group. The association of physical training and pyridostigmine bromide, speci cally at a dose of 15 mg/kg/day, increased the ejection volume, ejection fraction, cardiac output, and cardiac index compared to the trained vehicle group and its respective untrained group.

Discussion
Aerobic physical training and pyridostigmine bromide treatment had similar effects on cardiac autonomic tonic balance, characterized by an increase in vagal in uence and/or a reduction in sympathetic in uence, in addition to a reduction in hemodynamic parameters, such as BP, baseline HR, and IHR. In turn, the association between both did not enhance the effects.
In SHRs the accentuated predominance of the sympathetic autonomic drive over the vagal drive is often observed and it contributes to a high baseline HR, accompanied by adverse cardiac morphological and functional adaptations 16,28,29 . The reduction in baseline HR, induced only by pyridostigmine bromide treatment, results from a greater acetylcholine level. In this case, there is a tendency to reduce the baseline HR-dependent dose; that is, a dose of 15 mg/kg/day promotes greater effects, including in trained animals. In turn, the baseline HR reduction induced by aerobic physical training seems to involve a more complex mechanism, characterized by adaptation in central sites of cardiovascular control [30][31][32] , downregulation of β-adrenergic receptors 33,34 , and intrinsic cardiac adaptations. According to the literature, autonomic adaptations in central sites involve hypothalamic nuclei, such as the paraventricular and supra-optic nuclei, and nuclei located in the brainstem, such as the nucleus of the solitary tract (NTS) and rostral-ventrolateral medulla (RVLM) 35 . These adaptations seem to involve a series of factors, such as neural remodeling and the in uence of endogenous factors, which result in a decrease in sympathetic autonomic drive and/or an increase in the participation of the vagal autonomic drive in cardiac tonic control 20,36-42 . On the other hand, intrinsic cardiac adaptations also seem to contribute to baseline HR reduction, together with IHR reduction, which might be due to cardiac morphological and functional changes.
In our study, we observed that trained SHR showed an increase in LVEDV and LVESF, resulting in increases in LVEDV and LVESV as well as an expressive increase in ejection volume. These adaptations probably arise from the greater venous return induced by aerobic physical training. However, we did not observe an increase in cardiac output and cardiac index at rest, since the baseline HR reduction offset the greater diastolic lling. In turn, the echocardiography analysis of the animals treated with pyridostigmine bromide did not show changes in left ventricle diameter but showed modi cations in IVST and RWT associated with an increase in LV mass, mainly in animals previously submitted to aerobic physical training. The causes of these morphological adaptations are uncertain, but they suggest that changes in autonomic dynamics induced by physical training associated with pyridostigmine bromide treatment result in morphological adaptations that favor better cardiac performance. In addition, animals treated with pyridostigmine bromide also showed an increase in cardiac output. This adaptation is also associated with the increase in LVEDV found in these animals, resulting in a tendency to increase the ejection volume (P < 0.074). In addition to the increased cardiac output, mainly in the 15 mg/kg/day treatment group, the cardiac index (mL/g) remained unchanged.
Some studies have related BP reduction to the effects of chronic acetylcholinesterase block on endothelial function and a reduction in oxidative stress, which is also associated with aerobic physical training 43 . On the other hand, the mechanism responsible for BP reduction is still much discussed. However, other physical training effects can be noted, such as the attenuation of vascular and cardiac sympathetic activity, decreased serum levels of vasoconstrictor factors, and increased endothelial vasodilator factor levels as a result of the enhanced shear stress induced by aerobic physical training, resulting in reduced peripheral vascular resistance 36, 44,45 . The causes of the reduction observed after pyridostigmine bromide treatment were more complex. One hypothesis is that pyridostigmine bromide might also block the non-neural acetylcholinesterase action that is produced in other locations, such as endothelium and/or lymphocytes, promoting hemodynamic changes, as previously mentioned 9 .
The results of HRV, BPV, and BRS were interesting. Both aerobic physical training and pyridostigmine bromide treatment increased HRV LF oscillations in absolute units. However, only the treatment with pyridostigmine bromide, more speci cally at a dose of 15 mg/kg/day, increased the HRV HF oscillation in absolute units. Until then, we observed that the effects of pyridostigmine bromide treatment on baseline HR, IHR, and cardiac autonomic tonic balance were somewhat similar, and that the association with physical training potentiated some responses such as increased vagal tone and reduced baseline HR, BP, and IHR. In contrast, the HRV results showed that the association of treatments did not potentiate the results obtained with the isolated treatments, but rather attenuated or even reversed the bene cial effects of aerobic physical training and pyridostigmine bromide treatment when applied alone.
Regarding HRV, there was a reduction in HF oscillation, which seems to indicate a saturation of vagal stimulation from the association of both treatments 46 . Nevertheless, the obtained values were still greater than those obtained for the untrained vehicle group. Although, when the autonomic modulation balance represented by the LF/HF ratio was observed, the association treatment effects were even more evident. This occurrence can be explained by the pyridostigmine bromide effect, which increases both oscillation bands, LF and HF, mainly at a dose of 15 mg/kg/day. The causes of this increase are still uncertain, but it might be associated with an increase in sympathetic modulation oscillation, as well as vagal oscillations, since the LF band seems to be mediated by both autonomic components.
In the case of BPV, the ndings were also surprising. While only the trained animals showed a reduction in LF oscillation modulation, the animals treated with pyridostigmine bromide, trained or not, showed a signi cant enhancement of the LF band. In fact, SHR animals are classically known to have a high vascular sympathetic in uence, resulting in increased LF oscillations of the BPV 47-51 .
The cause of this increase is unknown and cannot be attributed to barore ex malfunction, since the BRS was increased, even in animals treated only with pyridostigmine bromide. This statement is based on the concept that barore ex is one of the main mechanisms involved in LF oscillations of BPV, together with vascular sympathetic drive 52 and relaxing factors derived from the vascular endothelium, such as nitric oxide (NO). The barore ex and NO would act as a buffer system for BP uctuations induced by sympathetic activation 53 . Therefore, some authors suggest that the increase in LF oscillation of BP is related to reductions in BRS and decrease in NO production and release. However, a study has demonstrated that long-term administration of acetylcholinesterase inhibitors, pyridostigmine or donepezil, attenuates vascular reactivity dysfunction in SHRs by decreasing reactive oxygen species generation and increasing NO bioavailability, possibly via increased endothelial NO synthase activity and inhibition of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase activity 43 . In fact, as noted, the explanation of the effects of pyridostigmine bromide on vascular modulation is even more complex than that of the effects on cardiac control, although both require further investigation.
The treatment with pyridostigmine bromide, mainly at a dose of 15 mg/kg/day, appears to be bene cial to cardiac autonomic regulation; it increased vagal autonomic tonic in uence in determining baseline HR and increased HRV and BRS. In turn, the enhancement in BPV is worrying. The association with aerobic physical training potentiates the reduction in hemodynamic parameters and results in a greater vagal autonomic tonic in uence on the heart. However, this association reduces vagal modulation of the heart. The mechanism involved in these ndings is still uncertain and requires further investigation.  gain down, barore ex sequence with progressive decreases in blood pressure followed by progressive decreases in pulse interval; gain up, barore ex sequence with progressive increases in blood pressure followed by progressive increases in pulse interval; F, factor; df, degrees of freedom.