Low Bicarbonate Concentration Research Articles

Variation in canine exocrine pancreatic secretory activity during the interdigestive state.

Interdigestive changes in pancreatic volume flow, gastric motor activity, and intraduodenal pH were monitored automatically in eight dogs. As a result, two different periods of pancreatic secretion were found to occur alternately during the interdigestive state: one a secretory period (2.2 +/- 0.32 ml/10 min with high protein and low bicarbonate concentration) and the other a nonsecretory period. The secretory period corresponded to the interdigestive phase II and III contraction period in the stomach, and the nonsecretory period coincided with the motor quiescent period in the stomach. However, another secretory period was identified during the interdigestive state; it occurred when the intraduodenal pH went below 7.0, and no typical interdigestive contractions were observed in the stomach. The secretory rate was 3.7 +/- 0.33 ml/10 min and was low in protein and high in bicarbonate concentration. The present study demonstrates that basal pancreatic secretory activity is never uniform but fluctuates with motor events in the stomach. However, its control mechanism remains unknown at present.

The effect of the growth hormone of Lymnaea stagnalis on (bi)carbonate movements, especially with regard to shell formation

In the freshwater snail Lymnaea stagnalis the role of the growth hormone in the control of bicarbonate movements was studied. This hormone is produced by the neurosecretory Light Green Cells (LGC) in the cerebral ganglia. After LGC removal the bicarbonate concentration in the hemolymph was 20–30% lower than in the sham-operated controls. Fifteen and thirty minutes after a single injection of [ 14C]bicarbonate, labeling of the haemolymph, mantle edge, mantle, and the other weak body parts was similar in both groups. The labeling of the shell edge of -LGC snails, however, was only about 10% of that of the shams. The deposition of carbonate in the shell was also lower. In the mantle edge, carbonic anhydrase was still present. The lower bicarbonate concentration in the hemolymph of -LGC snails is thought to be the result of a diminished metabolic activity. The lower carbonate deposition in the shell edge of -LGC snails is considered to be a secondary effect of growth hormone absence, the primary effect being the decrease of calcium movements through the mantle edge.

Photosynthetic Formation of Glycerate in Isolated Bundle Sheath Cells and Its Metabolism in Mesophyll Cells of the C4 Plant Panicum capillare L

In isolated bundle sheath cells of P. capillare, glycerate is a major product of 14CO2 fixation at low bicarbonate concentrations. After 6 min of CO2 fixation at 0.34 mM NaHCO3, about 20% of the labelled products was in glycerate and about 45% in glycine plus serine. Inhibitors of the glycollate pathway prevented incorporation of label into glycerate. After 6 min of CO2 fixation at high bicarbonate concentration (20 mM NaH14CO3), a large percentage of the label was incorporated in phosphate esters, insolubles, and sucrose, and little label was observed in metabolites of the glycollate pathway. The results indicate glycerate is primarily formed through the glycollate pathway in bundle sheath cells rather than through 3-phosphoglycerate phosphatase. Glycerate-dependent oxygen evolution in the light and 14C incorporation into phosphate esters from [1-14C]glycerate was observed with mesophyll chloroplasts and mesophyll protoplasts, respectively, but not with bundle sheath cells. This is consistent with our previously published results that glycerate kinase is localized only in C4 mesophyll chloroplasts. Both isolated bundle sheath cells and mesophyll chloroplasts exhibited 3-phosphoglycerate dependent O2 evolution in the light indicating that the reductive phase of photosynthesis is functional in both cells. The results suggest that, during C4 photosynthesis, glycerate is synthesized by the glycollate pathway in bundle sheath cells and subsequently metabolized in mesophyll cells without direct linkage to the reductive pentose phosphate pathway.

Reply by R. C. Sage and J. W. Lloyd

The authors would like to thank Mr Worthington for his comments and agree that the implications are beyond local considerations. In producing the interpretations, full use of the existing borehole and geophysical data has naturally been made so that any geological maps or sections are based on such data. The objective of the paper, however, is not to emphasize the value of hydrophysics, which would appear in any event to be unnecessary, or to examine the pros and cons of primary versus secondary permeability, but primarily to show the significance of the Pleistocene geology in relation to the recharge of the Triassic Sandstones. While the hydrochemical changes in parts of the Triassic Sandstone aquifer may be attributable to permeability differences in the sandstones, as has been pointed out in the paper, the integrity of low bicarbonate concentration areas correlated with Pleistocene sand deposits over significant distances, may well indicate that the sand permeability is locally an overriding factor at shallow levels in the aquifer system.

PH and bicarbonate effects on mitochondrial anion accumulation. Proposed mechanism for changes in renal metabolite levels in acute acid-base disturbances.

Mitochondria from rabbit and dog renal cortex were incubated with 1 mM (14)C-weak acid anions in media containing low (10 mM) or high (40 mM) concentrations of bicarbonate and the steady-state accumulation of labeled anion in the matrix was measured. In the absence of an energy source, no concentration of (14)C-anion in the mitochondrial matrix space was present, but the anion concentration was significantly higher at low- than at high-bicarbonate concentration. Addition of an energy source, usually ascorbate plus tetramethyl-p-phenylenediamine, led to increases in matrix space anion levels and to accentuation of the difference in anion uptake between low- and high-bicarbonate media, so that two to four times as much anion was present at low- than at high-bicarbonate concentrations. The anions affected included substrates for which inner membrane carriers are present in mitochondria, such as citrate, alpha-ketoglutarate, malate, and glutamate, as well as substances which diffuse passively across the inner membrane such as acetate and formate. When a nonbicarbonate medium buffered with Hepes was used, pH change did not alter anion uptake although anion concentrations exceeding those in the medium still developed when an energy source was present. The difference in mitochondrial anion accumulation between low- and high-bicarbonate levels diminished with decreasing temperature or with increasing anion concentration in the medium. Estimation of intramitochondrial pH with [(14)C]5,5-dimethyl-oxazolidine-2,4-dione showed that the pH gradient across the inner mitochondrial membrane was significantly greater with 10 than with 40 mM bicarbonate in the medium.A hypothesis is described that relates this effect of pH and bicarbonate on mitochondrial anion accumulation to the very rapid changes in substrate levels in renal cortex, which develop when acute metabolic acidosis or alkalosis is produced in the intact animal. It is suggested that an abrupt fall in systemic pH and bicarbonate is associated with a shift in substrate in renal cortex out of the cytoplasm and into mitochondria, where some of the added substrate is metabolized. Reduction in the size of the cytoplasmic pool of substrate occurs with relatively little accompanying change in the size of the mitochondrial pool, thus causing a net reduction in the total tissue pool. This mechanism accounts for the reduction in tissue levels of many mitochondrial substrates observed acutely in metabolic acidosis. In metabolic alkalosis, reversal of these effects leads to expansion of the cytoplasmic pool, thereby resulting in the rise in tissue levels of substrates which occurs in this condition.

Properties of sweat, burro ( Equus asinus) and man

1. 1. Sweat of two burros ( Equus asinus) was collected after two desert walks, one at 38°C, one at 41°C. 2. 2. Sweat from one burro was about twice as concentrated as from the other. In that respect and in respect to concentrations of chloride, sodium and potassium their sweat was like that of man. 3. 3. Low concentrations of bicarbonate were present in burros' sweat contrasted with little if any in man's. 4. 4. Urea nitrogen plus ammonia nitrogen were found in higher concentrations in burros' sweat than in man's sweat.

Child abuse presenting as immunodeficiency disease

duct.' The effects of a thoracic duct fistula are proportional to the duration and volume of the lymph drainage and include fever, hypoalbuminemia, and hypoglobulinemia,' chylofibrosis, and malnutrition. a Additional potential problems include respiratory distress,' salt and water depletion, and lymphocytopenia. To this list of possible complications we add metabolic acidosis. The normal anion gap suggested that this infant's acidosis was due to the loss of bicarbonate rather than the addition of acid. Comparing the cations (Na,K) with the anions (HCO Cl) normally reveals a cation preponder­ ance of approximately 11 to 19 mSq/l.' This anion gap is comprised of unmeasured anions, largely negatively charged plasma proteins. The addition of an acid (e.g., lactate or ,B-hydroxybutyrate) lowers the bicarbonate concentration without raising the concentration of other measured anions, and increases the gap. With acidosis secondary to bicarbonate loss, there is renal retention of chloride. The resulting hyperchloremia off-sets the low serum bicarbonate concentration, and the normal cation­ anion difference (gap) is maintained. As the electrolyte composition of lymph approximates that of plasma, s and as this infant's lymph loss averaged 700 nil/day (184 nil/kg/day), his water and electrolyte losses were substantial. Moreover, the magnitude of the infant's bicarbonate loss is impressive, considering that his normal total bicarbonate stores were only about 42 mEq (assuming bicarbonate's space of distribution to approxi­ mate 50%of body weight and a normal serumbicarbonate concentration to be 22 mfiq/l). Although the daily volume and bicarbonate concentration of the thoracic drainage varied, on some days the bicarbonate loss approximated SO% of the infant's normal body stores. Controversy does exist regarding the management of postoperative chylothorax. Many lymphatic leaks will seal spontaneously and can be managed with thoracentesis or chest-tube drainage. Additional conservative measures include intravenous administration of albumin, and a medium-chain triglyceride diet.' There is general agree­ ment, however, that thoracic duct ligation is necessary when profuse lymphatic drainage persists for more than a week or two.':

Reduced renal acid excretion in malnutrition: a result of phosphate depletion

Eleven infants recovering from protein-calorie malnutrition secondary to acquired monosaccharide intolerance were found to have reduced plasma bicarbonate concentration associated with inadequate weight gain. Renal net acid excretion (NAE) was decreased to a mean of 34.2 micronEq/1.73m2/min. Titratable acidity (TA) was markedly reduced, accounting for only 16% of NAE. This marked reduction in TA was associated with reduced mean phosphate (PO4) excretion (.074 mg/min) and a reduced mean serum PO4 (3.9 mg/dl), suggesting PO4 depletion. Two patients received intravenous phosphate loads, resulting in an increase in mean NAE from 35.1 to 89.7 microgEq/1.73m2/min. A similar response was seen after oral PO4 supplementation. Three patients were studied after partial correction of their acidosis. At a relatively low plasma bicarbonate concentration (mean = 16.6 mmoles/liter) significant amounts of bicarbonate were detected in the urine (mean = 8.7 micronEq/1.73m2/min), suggesting a defect in bicarbonate reabsorption. Five patients studied after complete recovery from malnutrition had normal NAE in response to ammonium chloride load. The reduction in NAE appears to be secondary to unavailability of urinary buffers and a reduction in bicarbonate reabsorption; both of these defects can be explained by phosphate depletion.

The influence of carbon dioxide, bicarbonate and other buffers on the potential of antimony microelectrodes.

Antimony microelectrodes were calibrated at 37 degrees C in phosphate buffers, in different bicarbonate solutions at various CO2-partial pressures and in buffers like TRIS1, TES2, MES3 and malonic acid. By use of the latter buffers (with exception of malonic acid) the most reliable calibration curves were obtained ("normal values"). The usual calibration in 67 mmol/l standard phosphate buffers turned out to be unacceptable because the obtained mV-values were too high (negative) in comparison to all other buffers. Different calibration curves resulted from the use of pure bicarbonate solution whether the pH-values were changed by variation of pCO2 or of the bicarbonate concentration. Low bicarbonate concentrations in combination with low pCO2 gave mV-values which were too low relative to the other buffers. Both the increase of pCO2 as well as of the bicarbonate concentration caused a shift of the potential of the antimony electrodes toward "normal values". In solutions containing other buffers the influence of bicarbonate and pCO2 became negligible with increasing buffer concentration. Decreasing oxygen partial pressure was found to cause an increase of the potential of the antimony electrodes. The influence of liquid junction potentials at the reference electrode is discussed.

D-Ribulose 1,5-diphosphate carboxylase from the blue-green alga Aphanocapsa 6308

d-Ribulose 1,5-diphosphate carboxylase from extracts of the unicellular blue-green alga Aphanocapsa 6308 has been purified by ammonium sulphate precipitation and linear sucrose density gradient centrifugation. The molecular weight was estimated to be 525 000 and the enzyme consisted of two types of sub-unit of molecular weights 51 000 and 15 000. The small sub-units were not detected after purification involving acid precipitation but were observed if the acid precipitation step was omitted. The Michaelis constants for Mg2+ and CO2, when tested under air, were 0.35 mM and 0.071 mM respectively. Oxygen acted as a competitive inhibitor with respect to CO2, suggesting that the enzyme also acts as an oxygenase. This was confirmed by measuring ribulose diphosphate-dependent O2 uptake. A 1:1 stoichiometry between ribulose diphosphate utilization and O2 consumption was observed. 6-Phosphogluconate inhibited carboxylase activity both at high (20 mM) and low (1 mM) bicarbonate concentrations. The data are compared with the properties of ribulose diphosphate carboxylase from other autotrophic prokaryotes and from chloroplasts.

Intrinsic differences in various segments of the proximal convoluted tubule.

Until recently it has not been possible to compare directly the function of superficial and juxtamedullary nephrons. The present studies, using in vitro microperfusion, were designed to examine whether functional differences exist between proximal convoluted tubule segments of superficial and juxtamedullary nephrons. Electrophysiological studies showed that major differences exist between the relative chloride and sodium permeabilities of these segments. In the 1st mm of the superficial proximal convoluted tubule, the permeability to sodium was greater than that to chloride, whereas in the 2nd mm of the superficial proximal convoluted tubule and all later segments, the permeability to chloride was greater than that to sodium. The juxtamedullary proximal convoluted tubule was found to differ from the superficial proximal convoluted tubule in two respects: first, the relative permeabilities to chloride and sodium did not differ in the various segments of the juxtamedullary proximal convoluted tubule; second, the permeability to sodium was greater than to chloride throughout. When perfused with a solution lacking glucose and amino acids, the superficial and juxtamedullary convolutions exhibited the same transepithelial potential change, a reversible decrease to less than -- 1 mV. It thus appears that in both convolutions there exists electrogenic sodium transport coupled to the transport of these organic solutes. This differs from pars recta of both of these nephrons, which have been shown to exhibit electrogenic sodium transport independent of organic solutes. However, when perfused with a solution lacking glucose and amino acids but also containing high chloride and low bicarbonate concentrations, the superficial convolution developed a significantly more positive potential than the juxtamedullary. This difference reflects greater relative chloride permeability in the superficial proximal convolution. These studies show that intrinsic functional differences exist between proximal convoluted tubules obtained from the superficial and juxtamedullary nephron populations.

Oxygenase properties of crystallized fraction 1 protein from tobacco

Ribulose 1,5-diphosphate-dependent oxygenase activity was demonstrated for crystallized Fraction 1 protein (RuDP 2 2 Abbreviations used: RuDP, ribulose 1,5-diphosphate; FDP, fructose 1,6-diphosphate; KCN, potassium cyanide. carboxylase EC 4.1.1.39) from tobacco. The kinetic properties of this oxygenase function were examined polarographically in air-equilibrated medium. Optimum activity was obtained at pH 8.4–8.6, and required 4–8 m m MgCl 2. Higher Mg 2+ concentrations decreased activity and slightly shifted the pH optimum to 8.2–8.3. The apparent K m (RuDP) and K m (Mg 2+) were 22 μ m and 0.5 m m, respectively. Oxygenase activity was inhibited by bicarbonate and indirectly by KCN. Kinetic studies suggest that the active inhibitory substance is the cyanohydrin derivative formed from the reaction of KCN with RuDP. Changes in oxygenase kinetics were observed upon addition of RuDP, as previously reported for the carboxylase function of this enzyme. Oxygenase activity required preincubation of the enzyme with both Mg 2+ and low concentrations of bicarbonate. Activities were enhanced about 20 and 70% when FDP (0.1 m m) and NADPH (0.5 m m), respectively, were included during preincubation.

Carbonate dehydratase (carbonic anhydrase) and the fetal lung.

Carbonic anhydrase activity (carbonic dehydratase, EC 4.2.1.1) has been detected in the fetal lungs of stillborn human infants and rhesus monkeys, but a role for this enzyme in the fetal lung has not been elucidated. In utero the mammalian lung develops as a liquid-filled structure, the liquid being secreted by the lung. In the fetal lamb this liquid, when compared with plasma, has a high chloride and a low bicarbonate concentration, suggesting a possible role for carbonate dehydratase. Studies on 10 fetal lambs confirmed the presence of carbonate dehydratase in the lung. Levels at 60-66 days were negligible and rose to 0.30 Meldrum Roughton units/mg protein at about 140 days (term 147 days), with little change after birth. In another six fetal lambs at 135-136 days, inhibition of this enzyme with 100 mg acetazolamide suppressed the mean rate of secretion of lung liquid by 64.5% (P less than 0.005), which correlated with a significant drop in chloride concentration (P less than 0.001). This magnitude of changes in secretion after acetazolamide is of the same order as that occurring in the secretion of cerebrospinal fluid when carbonate dehydratase is inhibited. This observation supports the hypothesis that carbonate dehydratase in fetal lung affects the secretion of lung liquid, although its mechanism is as yet unknown.

The relationship between carbonic anhydrase activity and glycolate excretion in the blue-green alga Coccochloris peniocystis

The rate of glycolate excretion by Coccochloris peniocystis Kütz. cells incubated under conditions of low bicarbonate concentration and high light intensity was linear for only the initial 15 min of incubation and no additional glycolate accumulated in the medium after 20 min. Excretion was maximal in cells grown on 5% CO2 in air when transferred to an incubation medium containing no added bicarbonate. The inhibitor INH (isonicotinyl hydrazide) had no measurable effect on the amount of glycolate released whereas HPMS (α-hydroxy-2-pyridyl methanesulfonate) stimulated excretion 3-fold. Cells transferred to air from growth on 5% CO2 in air increased in carbonic anhydrase activity, while a decrease occurred in the cells' ability to excrete glycolate. Cells grown on air and switched to 5% CO2 in air showed an increase in their ability to excrete glycolate with a concomitant decrease in carbonic anhydrase activity. Diamox, a specific inhibitor of carbonic anhydrase, was found to stimulate excretion with both airgrown and 5% CO2-grown cells which had been off 5% CO2 for approximately 30 min. The rate of carbon fixation by 5% CO2-grown cells put on air was found to rise over a 110 min period, corresponding to both the induction period of carbonic anhydrase and the period of decline in the ability of the cells to excrete glycolic acid. These results suggest that the absence of carbonic anhydrase in 5% CO2-grown cells causes a stimulation of glycolate excretion when these cells are transferred to a low bicarbonate medium, because of an increased rate of glycolate formation due to the oxidation of ribulose diphosphate by molecular oxygen at low internal CO2 concentrations.

Purification and Some Properties of Chlorella fusca Ribulose 1,5-Diphosphate Carboxylase

Ribulose 1,5-diphosphate carboxylase has been purified from extracts of autotrophically grown Chlorella fusca by ammonium sulfate precipitation and centrifugation on a linear sucrose density gradient. The enzyme was homogeneous by the criterion of polyacrylamide gel electrophoresis. The molecular weight of the enzyme was 530,000, and it was composed of two types of subunit of molecular weight 53,000 and 14,000. Ribulose 1,5-diphosphate, CO(2), and Mg(2+) had Michaelis constant values of 15 mum, 0.3 mm, and 0.37 mm, respectively. At high bicarbonate concentration (17 mm and 50 mm), 6-phosphogluconate inhibited the enzyme, the inhibition being noncompetitive with respect to ribulose 1,5-diphosphate (Ki 0.065 mm), whereas at low bicarbonate concentration (1 mm), 6-phosphogluconate activated the enzyme. Oxygen was a competitive inhibitor with respect to CO(2), suggesting the enzyme also functions as an oxygenase. This was confirmed by direct assay, a 1: 1 stoichiometry between ribulose 1,5-diphosphate consumed and O(2) uptake being observed.

A comparison of stool fluid and stool dialysate obtained in vivo.

Stool fluid has been obtained for analysis by homogenization and high speed centrifugation of fresh stool. This fluid, from two healthy individuals with soft stools has been compared with stool dialysate obtained in vivo by retrieval of swallowed dialysis bags from their stools. Stool fluid was more acid, with a higher osmolality, an increased concentration of organic anions, sugar and ammonia, and a lower bicarbonate concentration than dialysate in vivo. It is suggested that in the individuals studied, dialysate in vivo may not be in equilibrium with the stool fluid surrounding it, and may not represent a true dialysate of faecal water.

The action of the C-terminal octapeptide of cholecystokinin and related peptides on pancreatic exocrine secretion.

The effect of the C-terminal octapeptide of cholecystokinin-pancreozymin (CCK-PZ), caerulein, and the C-terminal tetrapeptide of gastrin on pancreatic secretion of fluid, electrolyte, amylase, and protein was studied in anaesthetized dogs prepared with pancreatic fistulae. Against background stimulation of fluid secretion with submaximal doses of secretin, all the polypeptides produced a qualitatively similar pancreatic response, causing a highly significant increase in amylase, protein, calcium, and zinc concentrations. Magnesium concentration was significantly increased only when the concentration preceding the administration of the peptide was below 100 mu-equiv/l. Octa-CCK-PZ was 13-35 times and 20-56 times more potent than tetragastrin on weight and molar bases, respectively, as a stimulant of amylase secretion. The threshold doses were largest for amylase, lower for calcium, and lowest for zinc. A significant linear correlation was observed between amylase and calcium concentration, zinc and protein concentration, and magnesium and calcium concentration. The peptides produced some increase in secretin-induced volume flow, whereas bicarbonate, chloride, sodium, and potassium concentrations remained unchanged. The direct relation between bicarbonate concentration and flow rate was limited to rates below 1.5 ml/5 minutes. At higher rates bicarbonate and chloride concentration reached a high and low plateau, respectively, although the first five-min sample of pancreatic juice after secretin stimulation exhibited a relatively low bicarbonate and high chloride concentration compared with its voluminous flow. Chloride concentration varied inversely with bicarbonate concentration, the sum of the two anions being constant.

Severity of metabolic acidosis as a determinant of bicarbonate requirements.

Abstract In the treatment of metabolic acidosis, the dose of sodium bicarbonate is usually calculated on the assumption that administered bicarbonate distributes in a space equivalent to 50 per cent of body weight. In a patient with profound metabolic acidosis the apparent space of distribution of bicarbonate was over 200 per cent of body weight, although no excessive ongoing acid production or external bicarbonate losses were evident, suggesting that the pre-existing plasma bicarbonate concentration can influence the apparent space of distribution of bicarbonate. This hypothesis was tested with a standard dose of bicarbonate (2.5 mmoles per kilogram of body weight) in 16 dogs with low, normal or high plasma bicarbonate concentration. The concentration increased by 2.8, 5.0 and 5.6 mmoles per liter in the three groups respectively, indicating that the proportion of administered bicarbonate passing into the intracellular space increases with decreasing concentration. In conditions of profound metabolic aci...

The composition of human lip mucous gland secretions

Both unstimulated and sour lemon-drop-stimulated secretions from the lip mucous glands (LMG) of young adults were collected into micropipettes, by a suction technique, for analysis for total protein, urea and the common electrolytes. The protein, sodium, potassium, calcium, magnesium, chloride and urea concentrations were not very dissimilar from those reported for the major salivary gland secretions. Sodium and chloride concentrations increased markedly on stimulation of flow rate. The inorganic phosphate content of the LMG secretions was only about one-tenth of that in the major salivary gland secretions and bicarbonate was virtually absent, chloride being the predominant anion. Because of the low bicarbonate and phosphate concentrations, the LMG secretions are poorly buffered and the degree of saturation with respect to hydroxyapatite is uncertain. Amylase did not appear to be present in the LMG secretions, but blood-group substances were secreted in much higher concentrations than those present in submandibular saliva.

Regulation by bicarbonate ion of intramitochondrial citrate concentration in kidney mitochondria

1. 1. Steady-state concentratins of citrate in the matrix space ([Citrate] M) in mitochondira from rabbit renal cortex were measured with low (10 mM) or high (40 mM) concentrations of bicarbonate in the medium. [Citrate] M was much less at high than a low bicarbonate concentration either in the presence or absence of rotenon. With rotenon and 1 mM citate in the medium [Citrate] M was over twice as greate at 10 as at 40 mM bicarbonate. 2. 2. The greatest effects on [Citrate] M occurred between 5 and 20 mM bicarbonate. 3. 3. The largest differences in [Citrate] M were observed when both pH and bicarbonate in the medium varied concomitantly. However significant differences in [Citrate] M occurred when bicarbonate concentration changed without change in pH; pH change with constant bicarbonate concentation caused little or no change in [Citrate] M. 4. 4. These results suggest that the concentation of bicarbonate ion regulates citrate transport across the mitochondrial membrane. This mechanism accounts for the inhibitory effects of bicarbonate on citrate oxidatin by slices and mitochondria from renal cortex and provides an explanation for the physiologic increase in citrate clearance in metabolic alkalosis.