- Email: [email protected]
Ventilatory responses to hyperkalemia and exercise in normoxic and hypoxic goats
Respiration Physiology, 82 (1990) 239-250
239
Elsevier
RESP 01716
Ventilatory responses to hyperkalemia and exercise in normoxic and hypoxic goats Margaret M. Warner and Gordon S. Mitchell Department of Comparative Biosciences, University of Wisconsin, Madison, Wisconsin, U.S.A. (Accepted 24 July 1990) Abstract. The ventilatory response to moderate exercise is potentiated during hypoxia in goats, causing P a c o 2 to decrease more from rest to exercise than in normoxia. We investigated the hypothesis that this response is due to the ventilatory stimulus provided by an interaction between exercise induced hyperkalemia and hypoxia. Plasma potassium concentration ([K ÷ ]), arterial blood gases and ventilation were measured in normoxia and hypoxia (Pao2 = 34-38 Torr) at rest and during steady-state exercise (5.6 kph; 5~o grade) in seven goats. P a c o ~ decreased during normoxic exercise (2.9 _+ 0.7 Torr; P < 0.01), and decreased significantly more during hypoxic exercise (6.4 _+ 0.6 Torr; P < 0.01). [K ÷ ] increased in both normoxic (1.0 _+ 0.1 mEq/L; P < 0.01) and hypoxic (0.9 + 0.2 mEq/L; P < 0.01) exercise, but these changes were not significantly different from each other. On a different day, resting goats were infused intravenously with 200 mM KCI for 5 min at a rate sufficient to obtain [K ÷ ] similar to exercise (8.6-12 ml/min) in normoxia and hypoxia. Hyperkalemia at rest caused similar Paco 2 decreases in normoxia (1.7 + 0.7 Torr; P < 0.05) and hypoxia (1.7 + 0.5 Torr; P < 0.01), but had no statistically significant effect on ventilation in either condition. These data indicate that hyperkalemia, at levels approximating those during moderate exercise, has a mild stimulatory effect on alveolar ventilation; however, hypoxia does not affect this response. We conclude that hyperkalemia does not provide sufficient ventilatory stimulation to account for exercise hyperpnea, nor does hypoxia potentiate the ventilatory stimulation from hyperkalemia at rest.
Animal, goat; Control of breathing, exercise hyperpnea; Exercise, ventilatory response to - ; Hypoxia, exercise hyperpnea in - ; Potassium, and exercise hyperpnea
The typical ventilatory response during moderate exercise is an increase in ventilation proportional to the increase in carbon dioxide production, regulating arterial blood gases and pH (cf. Dempsey et aL, 1985). This response is isocapnic in humans, but is characterized by mild hyperventilation and hypocapnic alkalosis in nonhuman mammals (cf. Dempsey et al., 1985). During exercise in hypoxia, ventilation increases in excess of that required to maintain control of arterial Pco2 (Paco2) (Forster and Klausen, 1973; Mason and Lahiri, 1974; Bisgard etal., 1982). The hallmark of this
Correspondence to: G.S. Mitchell, Dept. Comparative Biosciences, University of Wisconsin, 2015 Linden Drive West, Madison, WI 53706, U.S.A. 0034-5687/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
240
M.M. WARNER AND G.S. MITCHELL
hypoxia/exercise interaction is a greater decrease in Paco 2 from rest to exercise than in normoxia. The hypoxia/exercise interaction is eliminated by carotid denervation in goats (Mitchell etal., 1981; Bisgard etal., 1982), indicating that intact peripheral chemoreceptors are necessary for this response. On the other hand, constant carotid body stimulation with the D2-dopamine receptor antagonist domperidone does not elicit a similar response, indicating that carotid chemoreceptor stimulation alone is not sufficient to elicit the underlying mechanism (Schaefer and Mitchell, 1989). Thus, potentiation of the ventilatory response during hypoxic exercise may be due to a carotid chemoreceptor stimulus that is unique to hypoxic exercise. One chemoreceptor stimulus that has been proposed to account for the hypoxia/exercise interaction is exercise induced hyperkalemia (Burger et al., 1988). Plasma potassium concentration ([K + ]) increases during exercise in direct proportion to work intensity and with a time course similar to the ventilatory response (Lira et al., 1981; Band et al., 1982; Kuhlmann et al., 1985). In addition, [K + ] increases as arterial Po2 decreases in anesthetized cats (Paterson et aL, 1988). Increased plasma potassium concentrations stimulate carotid chemoreceptor activity and ventilation (Linton and Band, 1985). The ventilatory effect was eliminated by carotid body denervation (Band et al., 1985). Since hypoxia and hyperkalemia have synergistic effects on chemoreceptor discharge (Band and Linton, 1988; Burger et al., 1988), an interaction between exercise induced hyperkalemia and hypoxia could provide additional chemoreceptor stimulation during hypoxic exercise. Alternately, greater carotid chemoreceptor stimulation could result from greater levels of hyperkalemia during hypoxic exercise (Paterson et al., 1988). The objective of this study was to test the hypotheses that the hypoxia/exercise interaction is caused by (1) a greater increase in [K + ] during moderate hypoxic exercise, and/or (2)synergistic effects of exercise induced hyperkalemia and hypoxia in stimulating ventilation. These hypotheses were explored by determining changes in [K + ], ventilation and Paco 2 from rest to moderate exercise in normoxic and hypoxic goats. The hyperkalemia of moderate exercise was then simulated in awake, resting goats in an attempt to determine if hyperkalemia and hypoxia elicit a.ventilatory response sufficient to account for the hypoxia/exercise interaction without other factors attendant to exercise.
Methods
Experiments were conducted on seven adult female goats (38 to 56 kg; mean age: 3 years, as determined by tooth eruption). The animals were housed indoors and fed a diet of hay with occasional grain. At least one month prior to experimentation, each goat was prepared with a translocated carotid artery placed directly under the skin of the neck. Each goat had been extensively familiarized with the treadmill, facemask and experimental setup. Prior to an experiment, the subject was fasted for 24 h.
VENTILATION AND HYPERKALEMIA
241
Measurements. Arterial blood samples were drawn in triplicate (0.8 ml) in each experimental condition from an arterial catheter. Samples were immediately capped and stored anaerobically on ice while awaiting analysis with an automated blood gas analyzer (Radiometer, ABL-330) for pH, Pco2 and Po2. Each blood sample was analyzed at 37 °C and then corrected to the recorded rectal temperature of the animal. These values were then corrected with reference to three blood samples equilibrated with gases of known composition in a tonometer (Dynex-M) on each experimental day. Additional blood samples were centrifuged and the plasma analyzed for [K + ] using a potassium selective electrode (Beckman Electrolyte Analyzer System, E4A). Occasional mild hemolysis was noted as indicated by a pink tint in the plasma (presumably due to sampling trauma to red blood cells); however, these samples were analyzed as usual, did not have unusual levels of [K +] and were included in the analysis. A tight-fitting facemask (dead space < 200 ml) was equipped with an Hans Rudolph valve (series 2600). Tidal volume and frequency were measured with a pneumotachograph connected to the inspiratory port of the Hans Rudolph valve. Pressure across the pneumotachograph was measured with a variable reluctance differential pressure transducer (+ 2 c m H 2 0 ; Validyne, MP-45) and a carrier preamplifier (Gould). This output was integrated with a Gould resetting integrator to produce a signal proportional to tidal volume. The integrator signal was calibrated on each experimental day with three known volumes delivered from a calibration syringe (0.5, 1.0 and 1.5 Liter). Calculated values of inspired minute volume ('~I) and VT are reported at BTPS conditions. Metabolic CO2 production ('(/'co~) was calculated from measurements of the mixed expired CO2 fraction and ~/I (assuming that VI and VE were equal). By assuming that ~/I and ~/E are equal, an error of 2~o or less would result in calculated values of "~co2 if the respiratory exchange ratio changed from 0.5 to 1.0 since "~co2 is approximately 2~o of X/l. The mixed expired CO2 fraction was determined with an infrared CO 2 analyzer (Beckman, LB-2) after the expired gas had passed through a mixing chamber and desiccant column. Values of Vco 2 are reported in STPD conditions. Inspired oxygen levels were measured with an oxygen analyzer (Applied Electrochemistry, S-3A). All ventilatory and metabolic data were collected with an on line computer analysis system. Values of amplitude and timing were saved for each breath in a computer ASCII file. A second program was used for off line editing and summarization of these data files. Experimentalprotocol. The study was conducted in two experimental series. In Series /, the animals were studied at rest and during treadmill exercise (5.6 kph; 5~o grade) in conditions of normoxia (room air) and hypoxia (FIo2 = 0.09-0.11). The fraction of inspired O2 was varied to obtain a value of Pao2 between 34-38 Torr. If Pao2 was not in this range, the data were not considered further and the condition was repeated after making appropriate FIo2 adjustments. Measurements were made and blood samples drawn between five and seven min of exercise, thus assuring steady-states of O2-con-
242
M.M. WARNER AND G.S. MITCHELL
sumption and CO2-production (Smith et al., 1983). Following exercise, the goats were allowed to recover for at least 30 min before the next resting measurement was made. All goats performed this level of exercise with little difficulty, even during hypoxia. The same goats were studied three months later in Series II, to determine if levels of hyperkalemia similar to those recorded in exercise stimulate ventilation and produce similar changes in P a c o ~ at rest. Each subject was studied at rest in normoxia and hypoxia (Pao2 = 34-38 Torr). In each condition, measurements were made with or without continuous infusion o f a KC1 solution (200 mM) for 5 min into a jugular venous catheter. Infusion rates were used that mimic levels of [K + ] measured during exercise in Series I (8.6-12 ml/min). During infusion, the goats were carefully monitored for any signs of discomfort (e.g. muscle tremors or cardiac arrhythmias); none were observed unless greater rates of KCI infusion were used.
WI,VT, Paco2, f, [ K + ] and Vco2 were averaged for the seven goats and expressed + one standard error. VA was calculated for each goat from measurements of '~co2 and Paco:, averaged, and expressed + one standard error. Absolute changes from rest to exercise (Series I) or from no infusion to infusion at rest (Series H ) were determined and meaned across goats ( + one standard error). Mean data groups were compared via paired comparisons to determine significant changes from rest to exercise (Series I) or from no infusion to infusion at rest (Series II) in normoxia and hypoxia. Changes in variables were also compared between normoxic and hypoxic conditions within each series, and from one series to another (i.e. exercise vs infusion at rest). Paired t-tests were used with the Bonferroni correction procedure for multiple comparisons (Wallenstein et al., 1980). The number of multiple comparisons was four. P values were compared to a probability value of 0.0125 (0.05/4) for an overall experimental error rate of 0.05. Statistical analysis.
Results
Series I (exercise).
Figure la diagrams resting (lower bars) and exercising levels of [K ÷ ] (upper bars) in normoxia and hypoxia; mean values are given in table 1. Plasma potassium concentration increased during exercise in both normoxia (1.0 _+ 0.1 mEq/L; P < 0.01) and hypoxia (0.9 _+ 0.2 mEq/L; P < 0.01). These increases were not significantly different from each other. The change in Pa¢o 2 from rest to exercise (fig. lb) was significant in both normoxia (2.9 + 0.7 Torr; P < 0.01) and hypoxia (6.4 _+ 0.6 Torr; P < 0.01). The decrease in P a c o 2 during hypoxic exercise was significantly greater than in normoxia (P < 0.01); furthermore, the ratio of P a c o ~ decrease to [K + ] increase was greater in hypoxic (10.5 + 2.9 Torr/mEq/L) vs normoxic (3.2 + 0.7 Torr/(mEq/L)) exercise (P < 0.05). These results confirm that the hypoxia-exercise interaction occurs in goats under these experimental conditions. V~ significantly increased during exercise in both hypoxia and normoxia (fig. lc;
243
VENTILATION A N D HYPERKALEMIA TABLE 1 Mean values of measured variables during Series I experiments (exercise) Normoxic rest
[K +] (mEq/L) Paco2(Torr) (LSTPD/min) VI(LBTPs/min) ~'A(LBTPS/min) VT (LBTPS) f ( m i n t) pH Pao2 (Torr)
3.9 40.8 0.18 7.4 3.8 0.46 15.6 7.41 99
"Vco2
Normoxic exercise
(0.2) (1.0) (0.01) (0.6) (0.3) (0.02) (1.4) (0.01) (3)
4.9 37.9 1.13 45.8 26.3 0.79 55.8 7.41 105
(0.1)* (1.4)* (0.14)* (5.0)* (3.5)* (0.07)* (5.6)* (0.02) (2)
Hypoxic rest
Hypoxic exercise
3.6 36.7 0.21 9.9 5.0 0.46 20.9 7.45 37
4.5 (0.2)* 30.3 (1.4) *'~ 1.00 (0.11)* 56.0 (6.9)* 29.6 (4.3)* 0.79 (0.10)* 67.1 (4.6)* 7.46 (0.02) 37 (1) #
(0.1) (1.3) # (0.02) (1.0) # (0.6) (0.02) (2.2) (0.01) (1) '~
Mean values from seven animals + one standard error of the mean; (*) denotes a significant change from resting conditions (P < 0.05); ( # ) designates significant difference from normoxia (P < 0.05).
2-[• ~ .
~,
xfB"
.
5o [
45
r i
*#
r
]
4.o
25 ' --
_,oi
"~--
15
J
il
ill f] rio i
,ok ~lo\lq \1 RI I , I
II'~ I't I\ l( I \ l k l I'F
\()l~l(l\l( I\tt ,Ill\
I1"1PI)\I( I \ l l ,ll~\
C.
--
4o
3o
20
IO
0
F\FII(
O~F
F\FIII
[
i
I~F
I\FL
~lll\
iNFI ~IIIN
*
[
1 ~ORM()XI( F~\I, R ( I~F~
I1"* I,( IX I( ~- ~F R ( I".F
\ o 1 ~ "~I(IX I( INFI ~ION
I1"1 P( i'~ I( I%1"1%11)\
Fig. 1. (A) Absolute levels of plasma potassium concentration ([K ÷ ]) at rest or with no KCI infusion (lower bars) and exercise or KC1 infusion (upper bars) in normoxia and hypoxia. (*)designates a significant increase from rest to exercise or no infusion to infusion (P < 0.05). (B) The change in arterial Pco2 (Paco2); and (C) the change in inspired minute ventilation (Vx) from rest to exercise and no infusion to infusion in normoxia and hypoxia. (*) designates a significant difference from zero (P < 0.05). ( # ) designates a significant difference between normoxia and hypoxia (P < 0.05). In all figures, each value is the mean of 7 animals + SEM.
244
M.M. WARNER AND G.S. MITCHELL
P < 0.01), but the increase during hypoxic exercise was not statistically different from that in normoxia. However, when expressed as the slope of the relationship between ~'I and Qco2 (i.e. AQI/AQco2) , the ventilatory response in hypoxic exercise (60 + 5 LBTPS/LSTPD) was significantly greater than in normoxic exercise (35 + 6 LBTPS/LSTPD) (P < 0.01). VT increased 71 + 15 ~o in normoxia and 71 + 213o in hypoxia (both P < 0.01 ; fig. 2). f increased 250 + 35 ~o during normoxic exercise and
6O
o.s] A .
B. 50
0.4 °-
.=.
0.3
40
T
30 0.2 21} <1 0.1
0.0
10
NORMOXIC EXERCISE
HYPOXIC EXERCISE
NORMOXIC INFUSION
HYPOXIC INFUSION
0 NORMOXIC EXERCISE
HYPOXIC EXERCISE
NORMOXIC INFUSION
HYPOXIC INFUSION
Fig. 2. (A) The change in tidal volume (VT); and (B) the change in frequency (f) from rest to exercise or no infusion to infusion in normoxia and hypoxia. Each change is the mean of 7 animals _+ SEM. (*) designates a significant difference from zero (P < 0.05). 210 + 22~o in hypoxia (both P < 0.01 ; fig. 2). Increases in VT and f were not significantly different between normoxic and hypoxic exercise (neither absolute value nor°~ change). Thus, we were unable to draw any conclusions concerning the contribution of Vt vs f to the hypoxia/exercise interaction. VA (table 1) increased significantly during exercise in normoxia (22.5 + 3.3 L/min) and hypoxia (24.6 + 4.0 L/min; P < 0.01). The increase from rest to exercise in hypoxia was not significantly different from the increase in normoxia.
Series H (KCI infusion). Figure 1a shows [K + ] levels before (lower barr) and during (upper bars) KC1 infusion at rest in normoxia and hypoxia; absolute values are summarized in table 2. Increases in [K +] from rest to exercise and no infusion to infusion were not significantly different in any condition. Paco2 decreased significantly, and by the same amount during KC1 infusion in both normoxia (1.7 + 0.5 Torr; P < 0.01; fig. lb) and hypoxia (1.7 + 0.7 Torr; P < 0.01; fig. lb). Similarly, the ratio of P a c o 2 decrease to [K + ] increase during KC1 infusion did not differ between normoxia (1.9 + 0 . 6 T o r r / m E q / L ) and hypoxia (2.8 + 0.9 Torr/mEq/L). There were no statistically significant changes in VI, Vt or f during infusion in either
245
VENTILATION AND HYPERKALEMIA TABLE 2 Mean values of measured variables during Series II experiments (infusion) Normoxia/no infusion [K+](mEq/L) Paco2(Torr) Vco2 (LsxPt~/min) '¢~(LRxPs/min) ~/A(LBTPs/min) Vx (LBTeS) f(min 1) pH Pao2
3.9 (0.1) 41.1 (1.0) 0.24 (0.01) 11.2 (0.9) 4.9 (0.8) 0.48 (0.02) 22.1 (2.1) 7.42 (0.01) 106 (3)
Normoxia/ infusion
Hypoxia/no infusion
Hypoxia/ infusion
4.8 (0.1)* 39.4 (0.4)* 0.25 (0.01) 11.7 (0.8) 5.6 (0.3)* 0.48 (0.02) 23.3 (1.3) 7.43 (0.01) 109 (3)
3.8 (0.1) 37.0 (0.9)# 0.26 (0.02) 13.2 (1.0) 6.2 (0.5)# 0.51 (0.05) 25.5 (1.9) 7.44 (0.01) 37 (1) #
4.5 (0.1)* 35.3 (0.6)*# 0.29 (0.02) 15.3 (1.0)# 7.1 (0.5)*# 0.52 (0.05) 28.6 (3.0) 7.45 (0.01) 36 (1) #
Mean values from seven goats + one standard error of the mean; (*) denotes a value that is significantly different from the non-infusion value (P< 0.05); (#)designates significant difference from normoxia (e < 0.05).
normoxia or hypoxia, nor were there differences between normoxia and hypoxia (figs. lc and 2). "V'A (table 2) increased significantly during KC1 infusion in normoxia (0.7 + 0.3 L / m i n ; P < 0.025) and hypoxia(0.9 + 0.4 L / m i n ; P < 0.05),but the response during hypoxia was not significantly different from normoxia. These small changes in ~QA are appropriate to account for the 1.7 Torr decrease in P a c o 2. S e r i e s I vs S e r i e s H ( e x e r c i s e vs infusion). Neither absolute levels nor changes in [K + ] during KC1 infusion were significantly different from values measured during exercise at either oxygen level (fig. la). Comparison of changes in P a c o 2 during exercise vs KC1 infusion (fig. lb) indicates that there was no significant difference between the Pa¢o2 decrease during normoxic exercise vs normoxic KCI infusion. However, the decrease in Paco2 during hypoxic exercise was significantly greater than during hypoxic infusion (P < 0.01). Similarly, the ratio of change in P a c o 2 to change in [K + ] during normoxic infusion was not significantly different from normoxic exercise, but there was a significant difference between these values in hypoxic exercise vs hypoxic infusion (P < 0.01). Changes in ~rI, ~'I, Vt and f during exercise were substantially greater than during infusion at rest in either normoxia or hypoxia (figs. lc, 2 and'tables 1 and 2). These differences presumably reflect the primary exercise stimulus, a stimulus not present during KC1 infusion at rest.
Discussion We determined the degree of hyperkalemia during moderate exercise in normoxia and hypoxia in goats, and then simulated the hyperkalemia of conditions shown to produce
246
M.M. WARNER AND G.S. MITCHELL
the hypoxia/exercise interaction at rest. [K +] obtained was within the range of those reported to stimulate carotid chemoreceptor activity and ventilation in cats (Linton and B and, 1985) and were unaffected by oxygen level. However, hyperkalemia at rest elicited only mild stimulation of ventilation and, in the absence of other factors attendant to exercise, was not sufficient to account for exercise hyperpnea. Furthermore, there was no interaction between hypoxia and hyperkalemia in stimulating ventilation making it unlikely that hyperkalemia accounts for the hypoxia/exercise interaction. This study does not allow us to eliminate the possibility of a triple interaction between K +, hypoxia and exercise, nor can we draw conclusions about the ventilatory effects of greater levels of plasma potassium ( > 6 mEq/L) as would be seen during heavy exercise. Ventilatory control in exercise. During normoxic exercise, ventilation increases in proportion to increase in carbon dioxide production. In goats, like many other species (~f Dempsey, 1985), this response results in mild hypocapnic alkalosis. It has been suggested that potassium released from exercising muscles provides the stimulus for this ventilatory response (cf Burger et aL, 1988). However, our finding that hyperkalemia at rest, at levels equal to those measured during moderate exercise, causes only small increases in ~QI (fig. 1) suggests that hyperkalemia plays only a minor role in the hyperpnea of moderate exercise. The small but significant decrease in Paco 2 during infusion at rest implies that hyperkalemia does contribute some to exercise hyperpnea in goats, perhaps accounting for the mild hyperventilation. The possibility remains that potassium plays a role in the ventilatory response to exercise in hypoxia. The hypoxic response in not only different from that seen in normoxic exercise but is also unlike the response in other conditions that alter resting ventilation. When resting ventilation and blood gases are altered in conditions such as metabolic acidosis (Forster and Klausen, 1973), hormonal alterations (Knuttgen and Emerson, 1974; Pernoll et al., 1975; Skatrud et al., 1978), certain neurotransmitter alterations (Mitchell et al., 1984; Schaefer and Mitchell, 1989) and increased respiratory dead space (Mitchell, 1990), the ventilatory response to exercise is modulated such that regulation of Pa¢o 2 about its new resting level is unchanged. Such Paco 2 regulation requires active modulation of the ventilation-CO 2 production relationship in accordance with the alveolar ventilation equation. The varied nature of experimental conditions resulting in this careful control of Paco 2 during exercise has led to the hypothesis that a nonspecific central integrative mechanism exists which links the exercise ventilatory response to resting ventilation (Mitchell et al., 1984; Schaefer and Mitchell, 1989). If such a mechanism exists, linking resting ventilation with the ventilatory response to exercise, then hyperventilation elicited by hypoxia at rest would be expected to augment the exercise ventilatory response only enough to maintain the same precision of Paco ~ regulation from rest to exercise. However, ventilation is potentiated further during hypoxic exercise, decreasing Paco: more from rest to exercise than in normoxia (fig. 1). Thus, it has been proposed that the ventilatory response during hypoxic exercise results from at least three components: (1)the primary exercise stimulus, (2) a general mechanism modulating the ventilatory response to exercise in relation to resting
VENTILATION AND HYPERKALEMIA
247
ventilation, and (3) a carotid chemoreceptor stimulus that is unique to hypoxic exercise (Schaefer and Mitchell, 1989). The purpose of the work reported here was to explore the possibility that plasma potassium was the unique carotid body stimulus eliciting the hypoxia/exercise interaction. Hyperkalemia could account for potentiation of the ventilatory response to hypoxic exercise in two ways. First, ventilatory stimulation from hyperkalemia could be exaggerated during hypoxic exercise through a greater increase in [K + ]. It has been reported that [K + ] increases hyperbolically from 3.0 mM to 4.5 mM as arterial Po2 is lowered at rest (Paterson et al., 1988). Second, hyperkalemia could act by a unique interaction of hypoxia and the same [K + ] level as in normoxic exercise in stimulating the peripheral chemoreceptors (Burger et al., 1988). Interactions between potassium and other chemical stimuli at the peripheral chemoreceptors have been examined by several workers. However, some investigators report that potassium infusion increases ventilation by the same amount, regardless of oxygenation (Sneyd et al., 1988) whereas others report that hyperkalemia and hypoxia are synergistic in stimulating carotid chemoreceptors (Band and Linton, 1988; Burger et al., 1988). Most work concerning the role of hyperkalemia in ventilatory control has been done in anesthetized cats. The responses observed during anesthesia may differ from those of awake animals where regional circulation and O2 delivery to tissues are expected to be superior, and chemoreceptor responsiveness is greater. Thus, the work presented here tests the role of potassium in stimulating ventilation in normoxia and hypoxia in an awake, spontaneously breathing animal model. This study determined that elevations in [K + ] from rest to exercise are similar in normoxia and hypoxia in goats, indicating that a greater increase in [K+] during hypoxic exercise cannot account for the hypoxia/exercise interaction. Furthermore, hyperkalemia at rest did not interact with hypoxia to result in greater ventilatory stimulation when compared to normoxia. Thus, it does not appear likely that hyperkalemia has the requisite characteristics to account for the hypoxia/exercise interaction. Hyperkalemia at rest in normoxia and hypoxia caused apparent, nonsignificant increases in gI of 0.5 L/min and 2.1 L/min, respectively. These data could be interpreted as indicating a small interaction between hyperkalemia and hypoxia in stimulating ventilation. However, if we examine other indices of ventilation, neither Paco2, VA, nor the ratio of VA to ~/CO2 showed any indication of an interaction. Paco 2 decreased 1.7 Torr at both levels of oxygenation. "~A increased 0.7 + 0.3 L/min in normoxia and 0.9 + 0.4 L/min in hypoxia, and "V'A/Vco2increased 0.5 LBTPS/LSTPD in both normoxia and hypoxia. Collectively, these data indicate that hypoxia and hyperkalemia do not have any major interaction in stimulating ventilation at rest; however, we cannot rule out the possibility of a triple interaction between hypoxia, hyperkalemia and exercise. Furthermore, this study was not designed to explore the ventilatory effects of more extreme levels of hyperkalemia during more intense exercise. Thus, the design of this study prevents us from concluding that hyperkalemia plays no role in the hypoxia/exercise interaction.
Hyperkalemia and the hypoxia/exercise interaction.
248 Summary.
M.M. WARNER AND G.S. MITCHELL H y p e r k a l e m i a at r e s t c a u s e d o n l y slight s t i m u l a t i o n o f v e n t i l a t i o n at levels
s i m u l a t i n g m o d e r a t e exercise. T h u s , it a p p e a r s t h a t e x e r c i s e i n d u c e d h y p e r k a l e m i a h a s a v e r y l i m i t e d role in t h e h y p e r p n e a o f m o d e r a t e , n o r m o x i c exercise. T h e m i l d effect of p o t a s s i u m o n v e n t i l a t i o n w a s n o t p o t e n t i a t e d b y m o d e r a t e h y p o x i a . T h u s , it s e e m s u n l i k e l y t h a t p o t a s s i u m is n e c e s s a r y for t h e h y p o x i a / e x e r c i s e i n t e r a c t i o n .
Acknowledgements.This work was supported by grants from the National Institutes of Health (HL36780 and HL01494). We would like to thank S. Moen, P. Martin, P. Kaarakka and K.K. Nichols for excellent technical assistance.
References Band, D.M., M. Lim, R. A. F. Linton and C.B. Wolff (1982). Changes in arterial plasma potassium during exercise. J. Physiol. (London) 328: 74-75P. Band, D. M., R. A. F., Linton, R. Kent and F.L. Kurer (1985). The effect of peripheral chemodenervation on the ventilatory response to potassium. Respir. Physiol. 60: 217-225. Band, D.M. and R. A. F. Linton (1988). The effect of hypoxia on the response of the carotid body chemoreceptors to potassium in the anaesthetized cat. Respir. Physiol. 72: 295-302. Bisgard, G.E., H.V. Forster, J. Mesina and R.G. S arazin (1982). Role of the carotid body in hyperpnea of moderate exercise in goats. J. Appl. Physiol. 52: 1216-1222. Burger, R. E., J. A. Estavillo, P. Kumar, P. C. G. Nye and D.J. Paterson (1988). Effects of potassium, oxygen and CO 2 on the steady-state discharge of cat carotid body chemoreceptors. J. Physiol. (London) 401: 519-531. Dempsey, J. A., E. H. Vidruk and G. S. Mitchell (1985). Pulmonary control systems in exercise: update. Fed. Proc. 44: 2260-2270. Forster, H. V. and K. Klausen (1973). The effect of chronic metabolic acidosis and alkalosis on ventilation during exercise in hypoxia. Respir. Physiol. 17: 336-346. Knuttgen, H.G. and K. Emerson (1974). Physiological response to pregnancy at rest and during exercise. J. Appl. Physiol. 36: 546-553. Kuhlmann, W.D., D.S. Hodgson and M.R. Fedde (1985). Respiratory, cardiovascular and metabolic adjustments to exercise in the Hereford calf. J. Appl. Physiol. 58: 1273-1280. Lira, M., R.A.F. Linton, C.B. Wolff and D.M. Band (1981). Propranolol, exercise and arterial plasma potassium. Lancet 2: 591. Linton, R.A.F. and D.M. Band (1985). The effect of potassium on carotid chemoreceptor activity and ventilation in the cat. Respir. Physiol. 59: 65-70. Mason, R.G. and S. Lahiri (1974). Chemical control of ventilation during hypoxic exercise. Respir. Physiol. 22: 241-262. Mitchell, G.S., C.A. Smith, L.C. Jameson and J.A. Dempsey (1981). 02 and CO 2 induced changes in ventilatory gain in intact and carotid body denervated goats. Physiologist 24: 101. Mitchell, G. S., C. A. Smith and J. A. Dempsey (1984). Changes in the '¢1-'¢¢o 2 relationship during exercise in goats: role of carotid bodies. J. Appl. Physiol. 57: 1894-1900. Mitchell, G. S. (1990). Ventilatory control during exercise with increased respiratory dead space in goats. J. Appl. Physiol., 69: 718-727. Paterson, D.J., J.A. Estavillo and P. C. G. Nye (1988). The effect of hypoxia on plasma potassium concentration and the excitation of arterial chemoreceptors in the cat. Quart. J. Exp. Physiol. 73: 623-625. Pernoll, M. L, J. Metcalfe, P.A. Kovaeh, R. Wachtel and M. Dunham (1975). Ventilation during rest and exercise in pregnancy and post partum. Respir. Physiol. 25:295-310.
VENTILATION AND HYPERKALEMIA
249
Schaefer, S. L. and G. S. Mitchell (1989). Ventilatory control during exercise with peripheral chemoreceptor stimulation: hypoxia vs domperidone. J. Appl. Physiol. 67: 2438-2446. Skatrud, J. B., J. A. Dempsey and D. G. Kaiser (1978). Ventilatory response to medroxyprogesterone acetate in normal subjects: time course and mechanism. J. Appl. PhysioL 44: 939-944. Smith, C.A., G. S. Mitchell, L.C. Jameson, T.I. Musch and J.A. Dempsey (1983). Ventilatory response of goats to treadmill exercise: grade effects. Respir. Physiol. 54: 331-341. Sneyd, J.R., R.A.F. Linton and D.M. Band (1988). Ventilatory effects of potassium during hyperoxia, normoxia and hypoxia in anaesthetized cats. Respir. Physiol. 72: 59-64. Wallenstein, S., C. L. Zucker and J. L. Fleiss (1980). Some statistical methods useful in circulation research. Circ. Res. 47: 1-9.
239
Elsevier
RESP 01716
Ventilatory responses to hyperkalemia and exercise in normoxic and hypoxic goats Margaret M. Warner and Gordon S. Mitchell Department of Comparative Biosciences, University of Wisconsin, Madison, Wisconsin, U.S.A. (Accepted 24 July 1990) Abstract. The ventilatory response to moderate exercise is potentiated during hypoxia in goats, causing P a c o 2 to decrease more from rest to exercise than in normoxia. We investigated the hypothesis that this response is due to the ventilatory stimulus provided by an interaction between exercise induced hyperkalemia and hypoxia. Plasma potassium concentration ([K ÷ ]), arterial blood gases and ventilation were measured in normoxia and hypoxia (Pao2 = 34-38 Torr) at rest and during steady-state exercise (5.6 kph; 5~o grade) in seven goats. P a c o ~ decreased during normoxic exercise (2.9 _+ 0.7 Torr; P < 0.01), and decreased significantly more during hypoxic exercise (6.4 _+ 0.6 Torr; P < 0.01). [K ÷ ] increased in both normoxic (1.0 _+ 0.1 mEq/L; P < 0.01) and hypoxic (0.9 + 0.2 mEq/L; P < 0.01) exercise, but these changes were not significantly different from each other. On a different day, resting goats were infused intravenously with 200 mM KCI for 5 min at a rate sufficient to obtain [K ÷ ] similar to exercise (8.6-12 ml/min) in normoxia and hypoxia. Hyperkalemia at rest caused similar Paco 2 decreases in normoxia (1.7 + 0.7 Torr; P < 0.05) and hypoxia (1.7 + 0.5 Torr; P < 0.01), but had no statistically significant effect on ventilation in either condition. These data indicate that hyperkalemia, at levels approximating those during moderate exercise, has a mild stimulatory effect on alveolar ventilation; however, hypoxia does not affect this response. We conclude that hyperkalemia does not provide sufficient ventilatory stimulation to account for exercise hyperpnea, nor does hypoxia potentiate the ventilatory stimulation from hyperkalemia at rest.
Animal, goat; Control of breathing, exercise hyperpnea; Exercise, ventilatory response to - ; Hypoxia, exercise hyperpnea in - ; Potassium, and exercise hyperpnea
The typical ventilatory response during moderate exercise is an increase in ventilation proportional to the increase in carbon dioxide production, regulating arterial blood gases and pH (cf. Dempsey et aL, 1985). This response is isocapnic in humans, but is characterized by mild hyperventilation and hypocapnic alkalosis in nonhuman mammals (cf. Dempsey et al., 1985). During exercise in hypoxia, ventilation increases in excess of that required to maintain control of arterial Pco2 (Paco2) (Forster and Klausen, 1973; Mason and Lahiri, 1974; Bisgard etal., 1982). The hallmark of this
Correspondence to: G.S. Mitchell, Dept. Comparative Biosciences, University of Wisconsin, 2015 Linden Drive West, Madison, WI 53706, U.S.A. 0034-5687/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
240
M.M. WARNER AND G.S. MITCHELL
hypoxia/exercise interaction is a greater decrease in Paco 2 from rest to exercise than in normoxia. The hypoxia/exercise interaction is eliminated by carotid denervation in goats (Mitchell etal., 1981; Bisgard etal., 1982), indicating that intact peripheral chemoreceptors are necessary for this response. On the other hand, constant carotid body stimulation with the D2-dopamine receptor antagonist domperidone does not elicit a similar response, indicating that carotid chemoreceptor stimulation alone is not sufficient to elicit the underlying mechanism (Schaefer and Mitchell, 1989). Thus, potentiation of the ventilatory response during hypoxic exercise may be due to a carotid chemoreceptor stimulus that is unique to hypoxic exercise. One chemoreceptor stimulus that has been proposed to account for the hypoxia/exercise interaction is exercise induced hyperkalemia (Burger et al., 1988). Plasma potassium concentration ([K + ]) increases during exercise in direct proportion to work intensity and with a time course similar to the ventilatory response (Lira et al., 1981; Band et al., 1982; Kuhlmann et al., 1985). In addition, [K + ] increases as arterial Po2 decreases in anesthetized cats (Paterson et aL, 1988). Increased plasma potassium concentrations stimulate carotid chemoreceptor activity and ventilation (Linton and Band, 1985). The ventilatory effect was eliminated by carotid body denervation (Band et al., 1985). Since hypoxia and hyperkalemia have synergistic effects on chemoreceptor discharge (Band and Linton, 1988; Burger et al., 1988), an interaction between exercise induced hyperkalemia and hypoxia could provide additional chemoreceptor stimulation during hypoxic exercise. Alternately, greater carotid chemoreceptor stimulation could result from greater levels of hyperkalemia during hypoxic exercise (Paterson et al., 1988). The objective of this study was to test the hypotheses that the hypoxia/exercise interaction is caused by (1) a greater increase in [K + ] during moderate hypoxic exercise, and/or (2)synergistic effects of exercise induced hyperkalemia and hypoxia in stimulating ventilation. These hypotheses were explored by determining changes in [K + ], ventilation and Paco 2 from rest to moderate exercise in normoxic and hypoxic goats. The hyperkalemia of moderate exercise was then simulated in awake, resting goats in an attempt to determine if hyperkalemia and hypoxia elicit a.ventilatory response sufficient to account for the hypoxia/exercise interaction without other factors attendant to exercise.
Methods
Experiments were conducted on seven adult female goats (38 to 56 kg; mean age: 3 years, as determined by tooth eruption). The animals were housed indoors and fed a diet of hay with occasional grain. At least one month prior to experimentation, each goat was prepared with a translocated carotid artery placed directly under the skin of the neck. Each goat had been extensively familiarized with the treadmill, facemask and experimental setup. Prior to an experiment, the subject was fasted for 24 h.
VENTILATION AND HYPERKALEMIA
241
Measurements. Arterial blood samples were drawn in triplicate (0.8 ml) in each experimental condition from an arterial catheter. Samples were immediately capped and stored anaerobically on ice while awaiting analysis with an automated blood gas analyzer (Radiometer, ABL-330) for pH, Pco2 and Po2. Each blood sample was analyzed at 37 °C and then corrected to the recorded rectal temperature of the animal. These values were then corrected with reference to three blood samples equilibrated with gases of known composition in a tonometer (Dynex-M) on each experimental day. Additional blood samples were centrifuged and the plasma analyzed for [K + ] using a potassium selective electrode (Beckman Electrolyte Analyzer System, E4A). Occasional mild hemolysis was noted as indicated by a pink tint in the plasma (presumably due to sampling trauma to red blood cells); however, these samples were analyzed as usual, did not have unusual levels of [K +] and were included in the analysis. A tight-fitting facemask (dead space < 200 ml) was equipped with an Hans Rudolph valve (series 2600). Tidal volume and frequency were measured with a pneumotachograph connected to the inspiratory port of the Hans Rudolph valve. Pressure across the pneumotachograph was measured with a variable reluctance differential pressure transducer (+ 2 c m H 2 0 ; Validyne, MP-45) and a carrier preamplifier (Gould). This output was integrated with a Gould resetting integrator to produce a signal proportional to tidal volume. The integrator signal was calibrated on each experimental day with three known volumes delivered from a calibration syringe (0.5, 1.0 and 1.5 Liter). Calculated values of inspired minute volume ('~I) and VT are reported at BTPS conditions. Metabolic CO2 production ('(/'co~) was calculated from measurements of the mixed expired CO2 fraction and ~/I (assuming that VI and VE were equal). By assuming that ~/I and ~/E are equal, an error of 2~o or less would result in calculated values of "~co2 if the respiratory exchange ratio changed from 0.5 to 1.0 since "~co2 is approximately 2~o of X/l. The mixed expired CO2 fraction was determined with an infrared CO 2 analyzer (Beckman, LB-2) after the expired gas had passed through a mixing chamber and desiccant column. Values of Vco 2 are reported in STPD conditions. Inspired oxygen levels were measured with an oxygen analyzer (Applied Electrochemistry, S-3A). All ventilatory and metabolic data were collected with an on line computer analysis system. Values of amplitude and timing were saved for each breath in a computer ASCII file. A second program was used for off line editing and summarization of these data files. Experimentalprotocol. The study was conducted in two experimental series. In Series /, the animals were studied at rest and during treadmill exercise (5.6 kph; 5~o grade) in conditions of normoxia (room air) and hypoxia (FIo2 = 0.09-0.11). The fraction of inspired O2 was varied to obtain a value of Pao2 between 34-38 Torr. If Pao2 was not in this range, the data were not considered further and the condition was repeated after making appropriate FIo2 adjustments. Measurements were made and blood samples drawn between five and seven min of exercise, thus assuring steady-states of O2-con-
242
M.M. WARNER AND G.S. MITCHELL
sumption and CO2-production (Smith et al., 1983). Following exercise, the goats were allowed to recover for at least 30 min before the next resting measurement was made. All goats performed this level of exercise with little difficulty, even during hypoxia. The same goats were studied three months later in Series II, to determine if levels of hyperkalemia similar to those recorded in exercise stimulate ventilation and produce similar changes in P a c o ~ at rest. Each subject was studied at rest in normoxia and hypoxia (Pao2 = 34-38 Torr). In each condition, measurements were made with or without continuous infusion o f a KC1 solution (200 mM) for 5 min into a jugular venous catheter. Infusion rates were used that mimic levels of [K + ] measured during exercise in Series I (8.6-12 ml/min). During infusion, the goats were carefully monitored for any signs of discomfort (e.g. muscle tremors or cardiac arrhythmias); none were observed unless greater rates of KCI infusion were used.
WI,VT, Paco2, f, [ K + ] and Vco2 were averaged for the seven goats and expressed + one standard error. VA was calculated for each goat from measurements of '~co2 and Paco:, averaged, and expressed + one standard error. Absolute changes from rest to exercise (Series I) or from no infusion to infusion at rest (Series H ) were determined and meaned across goats ( + one standard error). Mean data groups were compared via paired comparisons to determine significant changes from rest to exercise (Series I) or from no infusion to infusion at rest (Series II) in normoxia and hypoxia. Changes in variables were also compared between normoxic and hypoxic conditions within each series, and from one series to another (i.e. exercise vs infusion at rest). Paired t-tests were used with the Bonferroni correction procedure for multiple comparisons (Wallenstein et al., 1980). The number of multiple comparisons was four. P values were compared to a probability value of 0.0125 (0.05/4) for an overall experimental error rate of 0.05. Statistical analysis.
Results
Series I (exercise).
Figure la diagrams resting (lower bars) and exercising levels of [K ÷ ] (upper bars) in normoxia and hypoxia; mean values are given in table 1. Plasma potassium concentration increased during exercise in both normoxia (1.0 _+ 0.1 mEq/L; P < 0.01) and hypoxia (0.9 _+ 0.2 mEq/L; P < 0.01). These increases were not significantly different from each other. The change in Pa¢o 2 from rest to exercise (fig. lb) was significant in both normoxia (2.9 + 0.7 Torr; P < 0.01) and hypoxia (6.4 _+ 0.6 Torr; P < 0.01). The decrease in P a c o 2 during hypoxic exercise was significantly greater than in normoxia (P < 0.01); furthermore, the ratio of P a c o ~ decrease to [K + ] increase was greater in hypoxic (10.5 + 2.9 Torr/mEq/L) vs normoxic (3.2 + 0.7 Torr/(mEq/L)) exercise (P < 0.05). These results confirm that the hypoxia-exercise interaction occurs in goats under these experimental conditions. V~ significantly increased during exercise in both hypoxia and normoxia (fig. lc;
243
VENTILATION A N D HYPERKALEMIA TABLE 1 Mean values of measured variables during Series I experiments (exercise) Normoxic rest
[K +] (mEq/L) Paco2(Torr) (LSTPD/min) VI(LBTPs/min) ~'A(LBTPS/min) VT (LBTPS) f ( m i n t) pH Pao2 (Torr)
3.9 40.8 0.18 7.4 3.8 0.46 15.6 7.41 99
"Vco2
Normoxic exercise
(0.2) (1.0) (0.01) (0.6) (0.3) (0.02) (1.4) (0.01) (3)
4.9 37.9 1.13 45.8 26.3 0.79 55.8 7.41 105
(0.1)* (1.4)* (0.14)* (5.0)* (3.5)* (0.07)* (5.6)* (0.02) (2)
Hypoxic rest
Hypoxic exercise
3.6 36.7 0.21 9.9 5.0 0.46 20.9 7.45 37
4.5 (0.2)* 30.3 (1.4) *'~ 1.00 (0.11)* 56.0 (6.9)* 29.6 (4.3)* 0.79 (0.10)* 67.1 (4.6)* 7.46 (0.02) 37 (1) #
(0.1) (1.3) # (0.02) (1.0) # (0.6) (0.02) (2.2) (0.01) (1) '~
Mean values from seven animals + one standard error of the mean; (*) denotes a significant change from resting conditions (P < 0.05); ( # ) designates significant difference from normoxia (P < 0.05).
2-[• ~ .
~,
xfB"
.
5o [
45
r i
*#
r
]
4.o
25 ' --
_,oi
"~--
15
J
il
ill f] rio i
,ok ~lo\lq \1 RI I , I
II'~ I't I\ l( I \ l k l I'F
\()l~l(l\l( I\tt ,Ill\
I1"1PI)\I( I \ l l ,ll~\
C.
--
4o
3o
20
IO
0
F\FII(
O~F
F\FIII
[
i
I~F
I\FL
~lll\
iNFI ~IIIN
*
[
1 ~ORM()XI( F~\I, R ( I~F~
I1"* I,( IX I( ~- ~F R ( I".F
\ o 1 ~ "~I(IX I( INFI ~ION
I1"1 P( i'~ I( I%1"1%11)\
Fig. 1. (A) Absolute levels of plasma potassium concentration ([K ÷ ]) at rest or with no KCI infusion (lower bars) and exercise or KC1 infusion (upper bars) in normoxia and hypoxia. (*)designates a significant increase from rest to exercise or no infusion to infusion (P < 0.05). (B) The change in arterial Pco2 (Paco2); and (C) the change in inspired minute ventilation (Vx) from rest to exercise and no infusion to infusion in normoxia and hypoxia. (*) designates a significant difference from zero (P < 0.05). ( # ) designates a significant difference between normoxia and hypoxia (P < 0.05). In all figures, each value is the mean of 7 animals + SEM.
244
M.M. WARNER AND G.S. MITCHELL
P < 0.01), but the increase during hypoxic exercise was not statistically different from that in normoxia. However, when expressed as the slope of the relationship between ~'I and Qco2 (i.e. AQI/AQco2) , the ventilatory response in hypoxic exercise (60 + 5 LBTPS/LSTPD) was significantly greater than in normoxic exercise (35 + 6 LBTPS/LSTPD) (P < 0.01). VT increased 71 + 15 ~o in normoxia and 71 + 213o in hypoxia (both P < 0.01 ; fig. 2). f increased 250 + 35 ~o during normoxic exercise and
6O
o.s] A .
B. 50
0.4 °-
.=.
0.3
40
T
30 0.2 21} <1 0.1
0.0
10
NORMOXIC EXERCISE
HYPOXIC EXERCISE
NORMOXIC INFUSION
HYPOXIC INFUSION
0 NORMOXIC EXERCISE
HYPOXIC EXERCISE
NORMOXIC INFUSION
HYPOXIC INFUSION
Fig. 2. (A) The change in tidal volume (VT); and (B) the change in frequency (f) from rest to exercise or no infusion to infusion in normoxia and hypoxia. Each change is the mean of 7 animals _+ SEM. (*) designates a significant difference from zero (P < 0.05). 210 + 22~o in hypoxia (both P < 0.01 ; fig. 2). Increases in VT and f were not significantly different between normoxic and hypoxic exercise (neither absolute value nor°~ change). Thus, we were unable to draw any conclusions concerning the contribution of Vt vs f to the hypoxia/exercise interaction. VA (table 1) increased significantly during exercise in normoxia (22.5 + 3.3 L/min) and hypoxia (24.6 + 4.0 L/min; P < 0.01). The increase from rest to exercise in hypoxia was not significantly different from the increase in normoxia.
Series H (KCI infusion). Figure 1a shows [K + ] levels before (lower barr) and during (upper bars) KC1 infusion at rest in normoxia and hypoxia; absolute values are summarized in table 2. Increases in [K +] from rest to exercise and no infusion to infusion were not significantly different in any condition. Paco2 decreased significantly, and by the same amount during KC1 infusion in both normoxia (1.7 + 0.5 Torr; P < 0.01; fig. lb) and hypoxia (1.7 + 0.7 Torr; P < 0.01; fig. lb). Similarly, the ratio of P a c o 2 decrease to [K + ] increase during KC1 infusion did not differ between normoxia (1.9 + 0 . 6 T o r r / m E q / L ) and hypoxia (2.8 + 0.9 Torr/mEq/L). There were no statistically significant changes in VI, Vt or f during infusion in either
245
VENTILATION AND HYPERKALEMIA TABLE 2 Mean values of measured variables during Series II experiments (infusion) Normoxia/no infusion [K+](mEq/L) Paco2(Torr) Vco2 (LsxPt~/min) '¢~(LRxPs/min) ~/A(LBTPs/min) Vx (LBTeS) f(min 1) pH Pao2
3.9 (0.1) 41.1 (1.0) 0.24 (0.01) 11.2 (0.9) 4.9 (0.8) 0.48 (0.02) 22.1 (2.1) 7.42 (0.01) 106 (3)
Normoxia/ infusion
Hypoxia/no infusion
Hypoxia/ infusion
4.8 (0.1)* 39.4 (0.4)* 0.25 (0.01) 11.7 (0.8) 5.6 (0.3)* 0.48 (0.02) 23.3 (1.3) 7.43 (0.01) 109 (3)
3.8 (0.1) 37.0 (0.9)# 0.26 (0.02) 13.2 (1.0) 6.2 (0.5)# 0.51 (0.05) 25.5 (1.9) 7.44 (0.01) 37 (1) #
4.5 (0.1)* 35.3 (0.6)*# 0.29 (0.02) 15.3 (1.0)# 7.1 (0.5)*# 0.52 (0.05) 28.6 (3.0) 7.45 (0.01) 36 (1) #
Mean values from seven goats + one standard error of the mean; (*) denotes a value that is significantly different from the non-infusion value (P< 0.05); (#)designates significant difference from normoxia (e < 0.05).
normoxia or hypoxia, nor were there differences between normoxia and hypoxia (figs. lc and 2). "V'A (table 2) increased significantly during KC1 infusion in normoxia (0.7 + 0.3 L / m i n ; P < 0.025) and hypoxia(0.9 + 0.4 L / m i n ; P < 0.05),but the response during hypoxia was not significantly different from normoxia. These small changes in ~QA are appropriate to account for the 1.7 Torr decrease in P a c o 2. S e r i e s I vs S e r i e s H ( e x e r c i s e vs infusion). Neither absolute levels nor changes in [K + ] during KC1 infusion were significantly different from values measured during exercise at either oxygen level (fig. la). Comparison of changes in P a c o 2 during exercise vs KC1 infusion (fig. lb) indicates that there was no significant difference between the Pa¢o2 decrease during normoxic exercise vs normoxic KCI infusion. However, the decrease in Paco2 during hypoxic exercise was significantly greater than during hypoxic infusion (P < 0.01). Similarly, the ratio of change in P a c o 2 to change in [K + ] during normoxic infusion was not significantly different from normoxic exercise, but there was a significant difference between these values in hypoxic exercise vs hypoxic infusion (P < 0.01). Changes in ~rI, ~'I, Vt and f during exercise were substantially greater than during infusion at rest in either normoxia or hypoxia (figs. lc, 2 and'tables 1 and 2). These differences presumably reflect the primary exercise stimulus, a stimulus not present during KC1 infusion at rest.
Discussion We determined the degree of hyperkalemia during moderate exercise in normoxia and hypoxia in goats, and then simulated the hyperkalemia of conditions shown to produce
246
M.M. WARNER AND G.S. MITCHELL
the hypoxia/exercise interaction at rest. [K +] obtained was within the range of those reported to stimulate carotid chemoreceptor activity and ventilation in cats (Linton and B and, 1985) and were unaffected by oxygen level. However, hyperkalemia at rest elicited only mild stimulation of ventilation and, in the absence of other factors attendant to exercise, was not sufficient to account for exercise hyperpnea. Furthermore, there was no interaction between hypoxia and hyperkalemia in stimulating ventilation making it unlikely that hyperkalemia accounts for the hypoxia/exercise interaction. This study does not allow us to eliminate the possibility of a triple interaction between K +, hypoxia and exercise, nor can we draw conclusions about the ventilatory effects of greater levels of plasma potassium ( > 6 mEq/L) as would be seen during heavy exercise. Ventilatory control in exercise. During normoxic exercise, ventilation increases in proportion to increase in carbon dioxide production. In goats, like many other species (~f Dempsey, 1985), this response results in mild hypocapnic alkalosis. It has been suggested that potassium released from exercising muscles provides the stimulus for this ventilatory response (cf Burger et aL, 1988). However, our finding that hyperkalemia at rest, at levels equal to those measured during moderate exercise, causes only small increases in ~QI (fig. 1) suggests that hyperkalemia plays only a minor role in the hyperpnea of moderate exercise. The small but significant decrease in Paco 2 during infusion at rest implies that hyperkalemia does contribute some to exercise hyperpnea in goats, perhaps accounting for the mild hyperventilation. The possibility remains that potassium plays a role in the ventilatory response to exercise in hypoxia. The hypoxic response in not only different from that seen in normoxic exercise but is also unlike the response in other conditions that alter resting ventilation. When resting ventilation and blood gases are altered in conditions such as metabolic acidosis (Forster and Klausen, 1973), hormonal alterations (Knuttgen and Emerson, 1974; Pernoll et al., 1975; Skatrud et al., 1978), certain neurotransmitter alterations (Mitchell et al., 1984; Schaefer and Mitchell, 1989) and increased respiratory dead space (Mitchell, 1990), the ventilatory response to exercise is modulated such that regulation of Pa¢o 2 about its new resting level is unchanged. Such Paco 2 regulation requires active modulation of the ventilation-CO 2 production relationship in accordance with the alveolar ventilation equation. The varied nature of experimental conditions resulting in this careful control of Paco 2 during exercise has led to the hypothesis that a nonspecific central integrative mechanism exists which links the exercise ventilatory response to resting ventilation (Mitchell et al., 1984; Schaefer and Mitchell, 1989). If such a mechanism exists, linking resting ventilation with the ventilatory response to exercise, then hyperventilation elicited by hypoxia at rest would be expected to augment the exercise ventilatory response only enough to maintain the same precision of Paco ~ regulation from rest to exercise. However, ventilation is potentiated further during hypoxic exercise, decreasing Paco: more from rest to exercise than in normoxia (fig. 1). Thus, it has been proposed that the ventilatory response during hypoxic exercise results from at least three components: (1)the primary exercise stimulus, (2) a general mechanism modulating the ventilatory response to exercise in relation to resting
VENTILATION AND HYPERKALEMIA
247
ventilation, and (3) a carotid chemoreceptor stimulus that is unique to hypoxic exercise (Schaefer and Mitchell, 1989). The purpose of the work reported here was to explore the possibility that plasma potassium was the unique carotid body stimulus eliciting the hypoxia/exercise interaction. Hyperkalemia could account for potentiation of the ventilatory response to hypoxic exercise in two ways. First, ventilatory stimulation from hyperkalemia could be exaggerated during hypoxic exercise through a greater increase in [K + ]. It has been reported that [K + ] increases hyperbolically from 3.0 mM to 4.5 mM as arterial Po2 is lowered at rest (Paterson et al., 1988). Second, hyperkalemia could act by a unique interaction of hypoxia and the same [K + ] level as in normoxic exercise in stimulating the peripheral chemoreceptors (Burger et al., 1988). Interactions between potassium and other chemical stimuli at the peripheral chemoreceptors have been examined by several workers. However, some investigators report that potassium infusion increases ventilation by the same amount, regardless of oxygenation (Sneyd et al., 1988) whereas others report that hyperkalemia and hypoxia are synergistic in stimulating carotid chemoreceptors (Band and Linton, 1988; Burger et al., 1988). Most work concerning the role of hyperkalemia in ventilatory control has been done in anesthetized cats. The responses observed during anesthesia may differ from those of awake animals where regional circulation and O2 delivery to tissues are expected to be superior, and chemoreceptor responsiveness is greater. Thus, the work presented here tests the role of potassium in stimulating ventilation in normoxia and hypoxia in an awake, spontaneously breathing animal model. This study determined that elevations in [K + ] from rest to exercise are similar in normoxia and hypoxia in goats, indicating that a greater increase in [K+] during hypoxic exercise cannot account for the hypoxia/exercise interaction. Furthermore, hyperkalemia at rest did not interact with hypoxia to result in greater ventilatory stimulation when compared to normoxia. Thus, it does not appear likely that hyperkalemia has the requisite characteristics to account for the hypoxia/exercise interaction. Hyperkalemia at rest in normoxia and hypoxia caused apparent, nonsignificant increases in gI of 0.5 L/min and 2.1 L/min, respectively. These data could be interpreted as indicating a small interaction between hyperkalemia and hypoxia in stimulating ventilation. However, if we examine other indices of ventilation, neither Paco2, VA, nor the ratio of VA to ~/CO2 showed any indication of an interaction. Paco 2 decreased 1.7 Torr at both levels of oxygenation. "~A increased 0.7 + 0.3 L/min in normoxia and 0.9 + 0.4 L/min in hypoxia, and "V'A/Vco2increased 0.5 LBTPS/LSTPD in both normoxia and hypoxia. Collectively, these data indicate that hypoxia and hyperkalemia do not have any major interaction in stimulating ventilation at rest; however, we cannot rule out the possibility of a triple interaction between hypoxia, hyperkalemia and exercise. Furthermore, this study was not designed to explore the ventilatory effects of more extreme levels of hyperkalemia during more intense exercise. Thus, the design of this study prevents us from concluding that hyperkalemia plays no role in the hypoxia/exercise interaction.
Hyperkalemia and the hypoxia/exercise interaction.
248 Summary.
M.M. WARNER AND G.S. MITCHELL H y p e r k a l e m i a at r e s t c a u s e d o n l y slight s t i m u l a t i o n o f v e n t i l a t i o n at levels
s i m u l a t i n g m o d e r a t e exercise. T h u s , it a p p e a r s t h a t e x e r c i s e i n d u c e d h y p e r k a l e m i a h a s a v e r y l i m i t e d role in t h e h y p e r p n e a o f m o d e r a t e , n o r m o x i c exercise. T h e m i l d effect of p o t a s s i u m o n v e n t i l a t i o n w a s n o t p o t e n t i a t e d b y m o d e r a t e h y p o x i a . T h u s , it s e e m s u n l i k e l y t h a t p o t a s s i u m is n e c e s s a r y for t h e h y p o x i a / e x e r c i s e i n t e r a c t i o n .
Acknowledgements.This work was supported by grants from the National Institutes of Health (HL36780 and HL01494). We would like to thank S. Moen, P. Martin, P. Kaarakka and K.K. Nichols for excellent technical assistance.
References Band, D.M., M. Lim, R. A. F. Linton and C.B. Wolff (1982). Changes in arterial plasma potassium during exercise. J. Physiol. (London) 328: 74-75P. Band, D. M., R. A. F., Linton, R. Kent and F.L. Kurer (1985). The effect of peripheral chemodenervation on the ventilatory response to potassium. Respir. Physiol. 60: 217-225. Band, D.M. and R. A. F. Linton (1988). The effect of hypoxia on the response of the carotid body chemoreceptors to potassium in the anaesthetized cat. Respir. Physiol. 72: 295-302. Bisgard, G.E., H.V. Forster, J. Mesina and R.G. S arazin (1982). Role of the carotid body in hyperpnea of moderate exercise in goats. J. Appl. Physiol. 52: 1216-1222. Burger, R. E., J. A. Estavillo, P. Kumar, P. C. G. Nye and D.J. Paterson (1988). Effects of potassium, oxygen and CO 2 on the steady-state discharge of cat carotid body chemoreceptors. J. Physiol. (London) 401: 519-531. Dempsey, J. A., E. H. Vidruk and G. S. Mitchell (1985). Pulmonary control systems in exercise: update. Fed. Proc. 44: 2260-2270. Forster, H. V. and K. Klausen (1973). The effect of chronic metabolic acidosis and alkalosis on ventilation during exercise in hypoxia. Respir. Physiol. 17: 336-346. Knuttgen, H.G. and K. Emerson (1974). Physiological response to pregnancy at rest and during exercise. J. Appl. Physiol. 36: 546-553. Kuhlmann, W.D., D.S. Hodgson and M.R. Fedde (1985). Respiratory, cardiovascular and metabolic adjustments to exercise in the Hereford calf. J. Appl. Physiol. 58: 1273-1280. Lira, M., R.A.F. Linton, C.B. Wolff and D.M. Band (1981). Propranolol, exercise and arterial plasma potassium. Lancet 2: 591. Linton, R.A.F. and D.M. Band (1985). The effect of potassium on carotid chemoreceptor activity and ventilation in the cat. Respir. Physiol. 59: 65-70. Mason, R.G. and S. Lahiri (1974). Chemical control of ventilation during hypoxic exercise. Respir. Physiol. 22: 241-262. Mitchell, G.S., C.A. Smith, L.C. Jameson and J.A. Dempsey (1981). 02 and CO 2 induced changes in ventilatory gain in intact and carotid body denervated goats. Physiologist 24: 101. Mitchell, G. S., C. A. Smith and J. A. Dempsey (1984). Changes in the '¢1-'¢¢o 2 relationship during exercise in goats: role of carotid bodies. J. Appl. Physiol. 57: 1894-1900. Mitchell, G. S. (1990). Ventilatory control during exercise with increased respiratory dead space in goats. J. Appl. Physiol., 69: 718-727. Paterson, D.J., J.A. Estavillo and P. C. G. Nye (1988). The effect of hypoxia on plasma potassium concentration and the excitation of arterial chemoreceptors in the cat. Quart. J. Exp. Physiol. 73: 623-625. Pernoll, M. L, J. Metcalfe, P.A. Kovaeh, R. Wachtel and M. Dunham (1975). Ventilation during rest and exercise in pregnancy and post partum. Respir. Physiol. 25:295-310.
VENTILATION AND HYPERKALEMIA
249
Schaefer, S. L. and G. S. Mitchell (1989). Ventilatory control during exercise with peripheral chemoreceptor stimulation: hypoxia vs domperidone. J. Appl. Physiol. 67: 2438-2446. Skatrud, J. B., J. A. Dempsey and D. G. Kaiser (1978). Ventilatory response to medroxyprogesterone acetate in normal subjects: time course and mechanism. J. Appl. PhysioL 44: 939-944. Smith, C.A., G. S. Mitchell, L.C. Jameson, T.I. Musch and J.A. Dempsey (1983). Ventilatory response of goats to treadmill exercise: grade effects. Respir. Physiol. 54: 331-341. Sneyd, J.R., R.A.F. Linton and D.M. Band (1988). Ventilatory effects of potassium during hyperoxia, normoxia and hypoxia in anaesthetized cats. Respir. Physiol. 72: 59-64. Wallenstein, S., C. L. Zucker and J. L. Fleiss (1980). Some statistical methods useful in circulation research. Circ. Res. 47: 1-9.