INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The carbonic anhydrase inhibitor acetazolamide is used in some patients with chronic obstructive pulmonary disease (COPD) and sleep-disordered breathing syndromes. The use of diuretics and steroids may lead to metabolic alkalosis in these patients, and acetazolamide may increase ventilatory drive in these subjects by lowering arterial base excess (1, 2). The ubiquitous presence, however, of carbonic anhydrase, particularly in tissues involved in (the control of) breathing, implies that the effect of acetazolamide on respiration may be complicated (1). In a high dose, acetazolamide not only will cause inhibition of carbonic anhydrase in red cells, but also in peripheral chemoreceptors, kidney, muscles, and in endothelial cells (4), so that it is difficult to estimate the contribution of each of these to the resulting respiratory effects.

An attractive means to study the role of carbonic anhydrase in the control of breathing is to apply selective enzyme inhibition at specific sites and to prevent the physiological effect of inhibition at others. For example, selective red cell inhibition appears to result in a large decrease in the slope of the ventilatory CO₂ response curve (4). Inhibition of the carotid body enzyme in the cat reduces carotid body output and steady state CO₂ sensitivity in vivo and abolishes the hypoxic ventilatory response (8, 9). Selective inhibition of the renal enzyme results in a (species-dependent) metabolic acidosis, increasing ventilatory drive (2, 10).

In previous studies, we were not able to exclude an effect of low-dose acetazolamide on respiratory muscles, as a possible influence of acetazolamide on the ventilatory pump was not investigated. Studying the effect of acetazolamide at the (neuro)muscular level would be important for at least two reasons. First, muscle cells contain several isoforms of carbonic anhydrase and inhibition of these may affect the efficiency of the excitation- contraction mechanism (11). Second, patients with COPD may suffer from severe peripheral muscle weakness, particularly when they are treated with steroids (16).

The aim of the present investigation in anesthetized rabbits was to study possible effects of low-dose acetazolamide (4.6 ± 0.2 mg · kg¹, mean ± SEM) on the performance of the respiratory pump at different CO₂ levels. This was achieved by simultaneously measuring spontaneous ventilation, phrenic nerve activity as well as transpulmonary pressure (via intraesophageal pressure), and by determining the response of all these parameters to changes in arterial PCO₂ both before and after intravenous administration of the agent. To minimize a physiological effect of carotid body carbonic anhydrase inhibition, the experiments were carried out at a background FI_O2 (fraction of inspired oxygen) of 0.37, at which level the peripheral chemoreceptors in the rabbit are virtually silent (17).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The experiments were performed in seven male rabbits with an average weight (± SEM) of 3.48 ± 0.14 kg. They were anesthetized with sodium pentobarbital. An initial injection of 55.4 ± 2.61 mg · kg¹ was given prior to tracheotomy and dissection of a branch of the external jugular vein for continuous infusion of the anesthetic. Infusion was started at an average rate of 7.26 ± 0.26 mg · kg¹ · h¹ to cover the ensuing 1- to 4-h operative period and was later maintained at 7.65 ± 0.20 mg · kg¹ · h¹.

Operation and Apparatus

The animals were placed in a supine position and, after midcollicular incision, the trachea was cannulated and connected to an open gas flow system. Oxygen-enriched air (FI_O2 = 0.37) was delivered by an excess flow of 100 L · h¹ O₂ and 400 L · h¹ room air. Both femoral arteries were cannulated for continuously recording the blood pressure and for arterial blood sampling. The left femoral vein was cannulated for intravenous injections.

Measurements

During spontaneous breathing, tidal volume (VT) and inspiratory and expiratory durations (TI, TE) were calculated from the respiratory flow signal, obtained by a Fleisch tube (size 0) connected with a Gould-Godart pneumotachograph (type 17212; SensorMedics, Bilthoven, The Netherlands). Respiratory intraesophageal pressure changes were measured by a fluid-filled catheter and a manometer/amplifier system (Validyne, Northridge, CA). The difference between end-inspiratory and end-expiratory intraesophageal pressure was defined as PTP, representing the tidal transpulmonary pressure change. Phrenic nerve activity was recorded from a distally cut C1-C5 root of the right phrenic nerve, desheathed and placed on a bipolar silver electrode in close connection with a preamplifier system for extracellular recordings (DAM 50; World Precision Instruments, Sarasota, FL). The amplified nerve signals of the compound potential were rectified and integrated by a self-constructed leakage integrator with a time constant of 200 ms. Airway CO₂ was continuously sampled and analyzed by infrared absorption (Binos 1; Leybold-Heraeus, Hanau, Germany). Blood pressure was measured by means of a fluid filled catheter in one femoral artery connected to a pressure transducer (Statham P 23Gb) in combination with an AWP4 DC bridge amplifier (Astro-Med, W. Warwick, RI). Respiratory and cardiovascular variables were continuously recorded by a chart recorder type MT95K2 (Astro-Med).

From arterial blood samples, the PO₂ was measured directly by a Clark-type electrode E5046 (Radiometer, Copenhagen, Denmark); the PCO₂ was determined indirectly from pH measurements by means of the two-gas equilibration method (18), using a BMS2 Mk2 blood microsystem and PHM 84 Research pH meter (Radiometer) together with a precision gas-mixing pump (Wösthoff, Bochum, Germany). The equilibration method was used to determine directly the relationship between Pa_CO2 and pH_a of the actual blood sample, from which the CO₂-buffering capacity (logPCO₂/pH_a), standard bicarbonate concentration (HCO₃st) and base excess (BE) can be calculated. Blood gas analysis was performed at 38° C, corresponding to the average rectal temperature of 38.5 ± 0.1° C, being controlled by a heating pad throughout the experiment. The concentration of hemoglobin (Hb) was measured spectrophotometrically (Hitachi-100-10; Hitachi, Tokyo, Japan) as cyanhemiglobin after treatment with Merckotest solution (E. Merck, Darmstadt, Germany). Lactate concentrations in the arterial blood were also determined spectrophotometrically by means of enzymatic analysis (LACT MPR 3; Boehringer, Mannheim, Germany).

Experimental Protocol

Throughout the experiment, the animals were given O₂-enriched air (FI_O2 = 0.37). After the animal had stabilized from the operative procedure, a first steady state measurement was carried out. This included five breaths to obtain means of different respiratory variables, blood pressure, and heart rate. Blood samples were taken immediately for blood gas and acid-base analysis. The CO₂ response of the ventilatory variables was then tested at three different levels of elevated arterial PCO₂ by adding 2.6 ± 0.1, 4.3 ± 0.1, and 6.0 ± 0.2% CO₂ to the inspiratory gas mixture. After removal of the CO₂ load, 0.5-1.0 ml of acetazolamide dissolved in saline (5 mg/ml) was injected at 20-min intervals to reach average (± SEM) cumulative doses of 0.73 ± 0.03, 1.45 ± 0.06, 2.78 ± 0.09, and 4.64 ± 0.22 mg · kg¹, respectively. Subsequently, taking about 40 min, ventilatory variables were measured again at the different levels of Pa_CO2.

Evaluation and Data Processing

Under steady state conditions, tidal volume (VT), and durations of inspiration and expiration (TI, TE), as well as the end-expiratory to end-inspiratory changes in transpulmonary pressure (PTP), the amplitude of the integrated phrenic nerve activity (IPNA), and end-tidal PCO₂ (PET_CO2) were evaluated as the mean of five breaths. Systolic and diastolic blood pressure and heart rate were read concomitantly during the same recording. Respiratory rate (fR) and pulmonary ventilation () were calculated as fR = 60/(TI + TE) and = VT · fR. The integrated phrenic nerve activity, yielding tidal phrenic amplitude (IPNA), was normalized by setting the maximum value that could be obtained in each animal to 100 U and multiplying all measured values by the factor B = 100/IPNA_max (19). In analogy to , phrenic minute activity was calculated as IPNA · fR. The CO₂ sensitivity of respiratory variables was calculated for three isocapnic changes in Pa_CO2 as the ratio /Pa_CO2.

Statistical Analysis

Mean steady state values and standard errors of the mean (SEM) were calculated for the whole population (n = 7). Variables were tested for normal distribution by the one-sample Kolmogorov-Smirnov test, to analyze differences of corresponding means under control and acetazolamide conditions by Student two-tailed paired t test. To explore the effect of acetazolamide on different relationships between VT, PTP, and IPNA, linear regression analysis was performed for each animal to compare differences of mean slopes and intercepts. At the level of p  0.05 differences were regarded as being significant. Statistical analysis was in part carried out by using SPSS 8.0 for Windows software (SPSS, Chicago, IL).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The effects of acetazolamide on resting acid-base status and ventilatory parameters are summarized in Table 1. The agent increased tidal volume and ventilation by about 20 and 13%, respectively, but did not change respiratory frequency. Mean arterial and end-tidal PCO₂ (the former was measured in vitro) decreased by 0.36 and 1.03 kPa, respectively. The mean P(a-ET)_CO2 difference, measured about 20 min after acetazolamide application, increased from 0.23 ± 0.16 to 0.89 ± 0.26 kPa (p < 0.05), indicating that the infused dose was on the margin of starting to inhibit erythrocytic carbonic anhydrase effectively. The lack of significant changes by acetazolamide in mean TI, TE, the ratios VT/TI and TI/TE, as well as in mean end-expiratory PTP (by 0.1 ± 0.1 kPa, p > 0.20) indicates an unchanged end-expiratory lung volume and end-expiratory position of the diaphragm. Mean base excess was reduced by 3.7 mM without any change in blood lactate concentration. Arterial blood pressure did not change significantly after acetazolamide infusion.

The effect of acetazolamide on the CO₂ response curves of ventilation and tidal volume, as well as their neuronal equivalents, is shown in Figure 1. A first observation of interest was the displacement of both the VT-Pa_CO2 and IPNA-Pa_CO2 response curves to lower Pa_CO2 values, particularly in the lower Pa_CO2 range. Second, the general picture emerging from Figure 1 is that in spontaneously breathing rabbits acetazolamide reduced the sensitivity of ventilation and tidal volume to changes in arterial PCO₂. However, at the level of minute phrenic activity and tidal phrenic amplitude, this could not be demonstrated, although in the lower PCO₂ range a slight tendency for a slope reduction may be present. The most illustrative difference in this respect is provided by a comparison between the VT-Pa_CO2 and IPNA-Pa_CO2 response curves: Figure 1 (bottom panels) shows that in contrast to the IPNA/Pa_CO2 slope, the VT/Pa_CO2 slope was clearly affected by acetazolamide, indicating that the agent must have caused a change in the relationship between tidal volume and IPNA.

View larger version (22K): [in this window] [in a new window]   Figure 1.   The effect of acetazolamide on the CO2 response of neuronal and mechanical respiratory variables. (Top panels) Pulmonary ventilation () and phrenic minute activity (IPNA · fR). (Bottom panels) Tidal volume (VT) and integrated phrenic nerve amplitude (IPNA) under control conditions (solid circles) and after low-dose application of acetazolamide (open circles). Values represent means ± SEM of seven rabbits. *CO2 sensitivities of responses during treatment with acetazolamide significantly different (p  0.05) from those under control conditions. Note significantly lower CO2 sensitivity of and V T, but not of IPNA · fR and IPNA, on carbonic anhydrase inhibition.

The effect of acetazolamide on this VT-IPNA relationship in one individual animal is shown in Figure 2A. In this rabbit, VT/IPNA was reduced by 50%. To investigate at which level (lung or [neuro]muscular) acetazolamide affected the VT-IPNA relationship, we also evaluated both the relationship between PTP and IPNA (Figure 2B) and that between VT and PTP (Figure 2C) in this rabbit. Mean data resulting from regression analysis for all seven animals are given in Table 2. These data show that acetazolamide does reduce both the mean VT/IPNA and the mean PTP/IPNA by about 40%. Obviously, acetazolamide did not alter the slope of the mean VT-PTP relationship, indicating an unchanged lung compliance. Consequently, the aforementioned decrease in VT/IPNA most likely resides at the (neuro)muscular level.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Effect of acetazolamide on the relationship between different respiratory variables in response to CO2 inhalation in a single rabbit. (A) Tidal volume (VT) as a function of integrated phrenic nerve amplitude (IPNA). (B) Tidal change in transpulmonary pressure (PTP) as a function of IPNA. (C ) VT as a function of PTP. Closed symbols refer to the control condition, open symbols to that after treatment with acetazolamide. Lines were obtained by linear regression analysis. Note the slope-reduction by acetazolamide in panels A and B, but not in C.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we report that a low intravenous dose of acetazolamide (4.6 ± 0.2 mg · kg¹) decreased the ventilatory and tidal volume responses to changes in Pa_CO2. The CO₂ sensitivity of phrenic nerve output was not clearly altered by the agent, although in the lower Pa_CO2 range a tendency for the response to flatten may be visible (see right panels of Figure 1). Lung compliance was not affected. Inspection of the relationships between phrenic nerve activity and the associated change in end-expiratory to end-inspiratory transpulmonary pressure (PTP) on the one hand and that between tidal volume and PTP on the other, revealed that the decreased ventilatory CO₂ sensitivity is most likely caused by an effect of acetazolamide at the (neuro)muscular level. A second result of interest was the upward shift of both the -Pa_CO2 and IPNA-Pa_CO2 response curves.

The increased arterial-to-end-tidal PCO₂ difference after acetazolamide (from 0.23 to 0.89 kPa; see Table 1) implies that the in vitro Pa_CO2 values (also shown in Table 1) do not represent the true in vivo values, which, because of the delayed equilibrium, must have been lower. It also means that acetazolamide must have shifted the actual in vivo CO₂ response curves somewhat more leftward than is shown in Figure 1 (in which the equilibrated in vitro Pa_CO2 is plotted along the x-axis).

A possible explanation for the leftward shift of the IPNA- Pa_CO2 response curves by acetazolamide (see Figure 1) could be an increased carotid body output due to the induced renal (nonlactic) metabolic acidosis. To minimize possible influences from the peripheral chemoreceptors, we performed the present experiments at an inspiratory O₂ fraction of 0.37 (Pa_O2 ~ 20 kPa). At this high oxygen level, both the normo- to hypocapnic steady state activity and the CO₂ sensitivity of rabbit carotid body fibers are low (17). Inhibition of local carbonic anhydrase as described for the in vivo (8) as well as the in vitro (20) cat carotid body may then have prevented an acidosis-induced increase in IPNA via carotid chemoreceptors in our rabbits. An alternative explanation for the leftward shift of the IPNA-Pa_CO2 response curve could be an inhibition of a membrane-bound carbonic anhydrase at the lumenal side of brain capillaries, resulting in a rise in brain tissue PCO₂ (21).

Because rabbits possess nonlinear relationships between ventilation or tidal volume and Pa_CO2 (Figure 1; see also Reference 22), we did not perform linear regression on data relating ventilatory parameters to Pa_CO2. Some (three) animals also showed nonlinear VT-IPNA or PTP-IPNA relationships, particularly before acetazolamide administration (e.g., see Figure 2). However, in order to be able to make a rough quantitative comparison between control and acetazolamide data, we performed linear regression on these data in all animals (results shown in Table 2). The clear general picture emerging is that the inhibitor by approximately 40% reduced the effectiveness of the phrenic nerve to mediate increases in tidal volume on a rise in PCO₂. Lung compliance was unaffected, as can be seen by identical VT-PTP relationships before and after acetazolamide (see Table 2 and Figure 2C). Thus, this effect most likely resides at the (neuro)muscular level. Indeed, Table 2 shows that also the mean change in PTP on a given increase in IPNA is diminished by about 40%. From the appearance of the VT- IPNA and PTP-IPNA relationships (Table 2; see also Figure 2), we tend to conclude that at resting IPNA levels existing at zero inspiratory PCO₂ the response lines merge, implying no visible effect of acetazolamide at and below normocapnic Pa_CO2 values, but a substantial effect at hypercapnic values.

Our data did not indicate changes by acetazolamide in mean inspiratory flow rate and end-expiratory position of the diaphragm, so that the above-described findings cannot be adequately explained by changes in force-velocity or length-tension relationships. Consequently, we believe, rather that the effect of acetazolamide is mediated by a direct action at the (neuro)muscular level.

As we infused acetazolamide in a low dose, an effect at the (neuro)muscular level is probably caused by inhibition of a sulfonamide-sensitive carbonic anhydrase. More specifically, the effect is likely to be mediated via inhibition of an easily accessible isoform of the enzyme, because acetazolamide diffuses relatively slowly through biological membranes (3). The most easily accessible sites are the lumenal side of the capillary endothelium and the extracellular surface of the sarcolemma, which contain the carbonic anhydrase (CA) isoform IV (11). Both the endothelial and sarcolemmal enzymes may be important for effective buffering of acid in the extracellular space during enhanced muscular activity. Under hypercapnic conditions, in particular, a decreased extracellular buffering may lead to intracellular acid accumulation and to less efficiency in the excitation-contraction mechanism, explaining our observation of a less effective phrenic activity.

We cannot exclude, however, an intracellular action of acetazolamide. Intracellularly, the sarcoplasmic (SR) CA IV would be a possible target, because this isoform is sulfonamide sensitive (23). Inhibition of SR CA IV may affect intracellular pH regulation and hence intracellular [Ca²⁺], and this may influence the contraction mechanism. A more likely intracellular target for acetazolamide would be cytosolic CA II, which occurs in both red and white muscle cells (14, 15). Like the red cell Type II isoform, this sulfonamide-sensitive isozyme has a high carbon dioxide hydratase activity, so that its inhibition may impair intracellular pH regulation. The main cytosolic carbonic anhydrase isozyme in red muscle cells, however, is CA III, which has a high resistance to sulfonamides (15, 24), requiring large concentrations of the permeable sulfonamide methazolamide to change it from the active to the inhibited state (27). It seems unlikely therefore that effective inhibition of cytosolic CA III is responsible for the effects of low-dose acetazolamide that we found in our rabbits.

Another possible explanation for the change in the PTP- IPNA relationship by acetazolamide would be an action on (neuro)muscular impulse transmission. In a frog gastrocnemius/ischiadic nerve preparation, Scheid and Siffert (30) showed that inhibition of local carbonic anhydrase reduced the isometric force on indirect (but not direct) stimulation. Ultrastructural studies of rat soleus muscle (12) provided no clue as to a specific location of CA on the sarcolemma, for example, associated with motor end-plates. In an isolated and indirectly stimulated rat phrenic-diaphragm preparation, Carmignani and coworkers (31) showed that acetazolamide, without altering resting contractile tension, diminished its increase induced by prostigmine. Note the similarity with our present observation that in our rabbits the effects of the agent were visible at elevated Pa_CO2 levels only, when increased levels of acetylcholine at the phrenic-diaphragm junction must be present.

Clinical Application of Acetazolamide

Carmignani and coworkers (31) demonstrated that an intravenous infusion of 500 mg of acetazolamide in myasthenic patients prevented an increase in action potential amplitude of the opponens pollicis muscle induced by the anticholinesterase agent edrophonium. This indicates that intravenous administration of acetazolamide, even in doses that do not completely block the red cell enzyme, may lead to adverse effects. Swenson and Hughes (32) found that the same intravenous dose resulted in inhibiting effects on the control of breathing in healthy volunteers. In many (but not all) clinical applications, however, the agent is used by way of oral intake. In healthy subjects, a usual clinical oral dosage of 250 mg each 8 h does not impair the CO₂ sensitivities of both the peripheral and central chemoreflex loops (2). It can not be excluded, however, that, despite this, inhibiting effects at the (neuro)muscular level may occur. This, for example, is illustrated by observations reported by Brechue and coworkers (33): in healthy subjects they showed that acetazolamide, after three oral doses of 250 mg each 6 h, inhibited the Achilles tendon-tap reflex and associated isometric force.

In patients with COPD, the situation may be even more complicated. In these patients with severe hypercapnia, acetazolamide may improve blood gases considerably, without significant increases in ventilation (2). In the present study of rabbits we found that in a hypercapnic situation, after low-dose acetazolamide an increased neuronal drive was necessary to maintain a given tidal volume. Apart from metabolic alkalosis (which may be offset by an acetazolamide-induced decrease in base excess), steroids are able to cause myopathy in patients with COPD, resulting in severe muscle weakness (16). Our data raise questions about whether clinical doses of acetazolamide may worsen this situation.