Gastric HC03ANDREW Department
GARNER
secretion in the guinea pig AND
of Physiology
GUNNAR
FLEMSTROM
and Medical Biophysics,
GARNER, ANDREW, AND GUNNAR FLEMSTR~M. Gastric HCO,secretion in the guinea pig. Am. J. Physiol. 234(6): E535-E541, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 3(6): E535-E541, 1978. - Measurement of gastric intraluminal Pcoz and pH in the anesthetized guinea pig enabled simultaneous determination of total H+ and HCO,- gastric secretions. There was quantitative agreement between the release of CO, and decrease in HCO,after intragastric instillation of HCl. The basal rate of HCO,secretion (-40 peq=h+) was, in most cases, smaller than spontaneous H+ secretion, but gastric net secretory output was alkaline (HCO,- > H+) after inhibition of acid secretion with histamine Hz-receptor antagonists (cimetidine 20 mgekg-l or metiamide 35 mg=kg+). Carbachol (l-2 pg*kg-l) stimulated secretion of both HCO,- and H+; only the latter response was sensitive to the histamine antagonists. Atropin (100 ,ug.kg+) blocked stimulation of HCO,- secretion but did not affect the basal output of HCO,-. An increase in HC03secretion was associated with an equivalent increase in net Na+ influx and an increase in the net influx of Cl- with H+ plus K+. Intragastric neutralization of H+ by HCO,- is likely to occur at the mucosal surface and may protect the mucosa from the damaging effects of intraluminal acid. atropin, carbachol; H+ secretion; HCO,- secretion; Hz-receptor antagonists; intragastric Pco2
histamine
ACTIVE SECRETION of HCO,- by the gastric surface epithelial cells was recently demonstrated in antral (12, 13) and fundic (8-10) mucosa in vitro. This secretion was sensitive to acetazolamide and stimulated by carbachol, DB-cGMP and Ca2+ ions. Secretion of HCO,probably hreduces H+ ion concentration at the luminal surface of the gastric epithelium and thus may act as a protective mechanism. The aim of the present study was to determine whether a similar secretion occurs under in vivo conditions and to examine some of the properties of HCO,transport in the mammalian stomach. The occurrence of free HCO,- in gastric nonacid secretions in vivo has been reported previously (17, 20); however, at the low values of pH prevailing in the stomach, HC03- would rapidly decompose to CO, and water. Measurements of intragastric Pco2 and pH were used here for simultaneous determination of total gastric HCO,- and H+ secretions. MATERIALS
AND
METHODS
Male albino guinea pigs (300-420 g) were starved for 20 h before use in cages with mesh bottoms to minimize coprophagy, but allowed free access to drinking water. 0363-6100/78/0000-0000$01.25
Copyright
0 1978 the American
University
of Uppsala,
Uppsala,
Sweden
Animals were anesthetized with urethane (1.5 g-kg-‘), administered intraperitoneally. Body temperature was maintained at 37°C with a heater controlled by an intrarectal thermistor. Surgical procedures. The abdomen was opened by a median epigastric incision and a multiorifice soft rubber tube (inset, Fig. 1) passed into the stomach through a cut in the duodenum proximal to the insertion of the common bile duct. The tube was secured in place by means of a ligature around the pylorus, with care taken to exclude blood vessels from the tie, and exteriorized via a stab wound in the right flank before the abdomen was closed. After tracheostomy, a soft catheter (umbilical artery catheter size 8 Fr., Sherwood Medical Industries, Inc., St. Louis, MO.) was passed down the esophagus into the stomach and positioned so that the tip terminated just beyond the cardia. This catheter was held in place by a tie around the esophagus in the neck region, a procedure which also prevented the contamination of gastric samples with salivary secretions. The left external jugular vein was cannulated for drug administration. Collection of gastric samples. Before beginning an experiment, the stomach was rinsed clear of food debris with lo-ml volumes of isotonic NaCl or mannitol at 37OC. Fluid was introduced via the esophageal catheter and allowed to escape through the pyloric outlet while the abdominal surface of the animal was gently massaged. Complete removal of the fasting gastric contents was obtained after 6-10 rinses. A diagram of the equipment used for filling and emptying the stomach is shown in Fig. 1. The animal was secured to a hinged table and the esophageal catheter attached to a 3-way tap (A). The pyloric tube was connected to a second 3-way tap (B) by means of glass tubing. The stomach was then filled with 10 ml of fluid at 37°C from a heated reservoir, with the animal tilted so that the pyloric region was uppermost (Fig. 1, position 1) allowing air in the stomach to escape via the 3-way tap (B). Fluid was recovered by tilting the animal so that the pyloric region was the lowest point of the stomach (Fig. 1, position 2) and turning on the flow inducer (WatsonMarlow, Falmouth, England) set at 100 mlamin-l with the 3-way taps adjusted to form a closed circuit. Intragastric fluid was thus replaced with an equal volume of air. The method allowed rapid filling and emptying of the stomach (approximately 15 s for each), and stretching of the gastric wall during these procedures was minimized. Between 96 and 103% of the instilled volume was recovered after 10 min.
Physiological
Society
E535
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E536
A. GARNER
combine to form COZ, representing ization
AN’D G. FLEMSTROM
intragastric
neutral-
H+ + HC03- = CO, + H,O
FIG. 1. Diagram of apparatus used to collect gastric samples from guinea pig. Animal tilted through an angle of 60” betweenposition 1 (filling stomach) and position 2 (emptying stomach); otherwise it was placed horizontally. Inset shows a diagram of pyloric outlet tube and position of cannula used to remove intragastric samples for measurements of Pco, levels.
In most experiments, lo-ml volumes of either isotonic NaCl or mannitol solutions were instilled and removed for analysis after 10 min. Mannitol was used in experiments in which gastric outputs of Na+ and Cl- were determined. There was no difference between NaCl and mannitol instillations with respect to gastric outputs of HCO,- and H+ and some of these data were therefore pooled. Measurement of intragastric Pco, and PH. Intragastric Pco, was measured immediately before removal of the instilled solutions with a Pco, electrode (E5036, Radiometer, Copenhagen) connected to an expanded scale millivolt meter. Samples were withdrawn into the electrode by means of gentle suction from a syringe attached to the electrode outlet. The exposed section of this tube was made from glass capillary to minimize escape of CO, by diffusion. Before and after each experiment, the electrode was calibrated with gas mixtures containing known concentrations of CO, in the range of 1.5-12.8%. Calibration and all measurements were carried out at 37°C. The rest of the instilled fluid (-9 ml) was collected in a glass test tube and the pH measured with a glass calomel electrode (GK-2321, Radiometer, Copenhagen) connected to an expanded scale pH meter. The final volume of each gastric sample was recorded after addition of fluid used for PCS measurement. The sample was then gassed with 100% N1, which had been prewashed in Ba(OH),. Immediately after gassing for 5-15 min, a 5-ml aliquot was titrated to pH 7.00 with 5 mM NaOH. In the case of samples, alkaline after gassing (HCO,- secretion > H+ secretion), remaining HCO,was determined by back titration after acidification with 50 ~1 of 100 mM HCl standard solution. Analysis of gastric HCO,- and H+ secretions. The following principles were used. a) Some HCO,- and H+
The equal amounts of HCO,- and H+ thus removed intragastrically were designated HCO,- (CO,) and H+ (CO,) and calculated from the intragastric PcoZ and the solubility coefficient of CO,. b) Substitution of measured pH and Pco, in the Henderson-Hasselbach equation enabled calculation of free HCOI- in the instillate pH = pK, + log (HC03-/~ Pco,) Values used for PK,, the dissociation constant, and (Y, the solubility coefficient, were 6.09 and 0.03, respectively. These values were determined for measurement of HCO,- in kidney tabular fluid (22) which has a composition similar to the gastric samples examined here. c) Gassing of the recovered instillate with Nz promoted formation of COZ from HCO,- and H+ (by removal of CO, formed intragastrically and during gassing). Acid or base remaining after gassing was determined by titration. Total concentrations of HCO,- and H+ in the instillates were then calculated total HCOI-: HC03- (CO,) plus HC03- (free) total H+: i) in samples acid after gassing with N2 H+ (CO& plus HCOS- (free) plus H+ (titrated) ii) in samples alkaline after gassing with N, H+ (CO,) plus HC03- (free) minus HCO1- (titrated) At pH values below 4.0, amounts of HC03- (free) were negligible. In samples alkaline aRer gassing, HC03(free) minus HCOI- (titrated) represents H+ reacting with HCO$- during the gassing and thus removed. In the above analysis of gastric HC03- and H+, it was assumed that gastric luminal CO, arises from reaction between secreted HCO,- and H+ and that diffusion of CO, out from or into the gastric lumen is slow. Experimental evidence for these assumptions will be presented (vide infra). Other analyses. Concentrations of Na+ and K+ were determined by flame photometry (Eppendorf, Netheler and Hinz, FGR) and Cl- was determined by titration (chloride titrator CMTlO, Radiometer, Copenhagen). The recovered solutions were always tested for hemoglobin and glucose (Labstix, Ames Co., Slough, England). These methods were checked with standard solutions and gave positive reactions for hemoglobin at a concentration of 3 pg+ml-l and for glucose at 500 pgernl-‘. Drugs and chemicals. All drug solutions were adjusted to pH 7.4 and isotonicity before use and injected intravenously in a volume of 1 mlkg-’ body wt. The histamine HZ-receptor antagonists, cimetidine and metiamide, were kind gifts from Dr. M. E. Parsons, Smith, Kline & French Laboratories, Ltd., Welwyn Garden, England. Carbachol (carbamylcholine chloride) and
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
HC03-
E537
SECRETION
atropin (as sulphate) were obtained ical Co., St. Louis, MO.
from Sigma Chem-
RESULTS
Intragastric C02. A prerequisite for the present method is that intraluminal CO, arises mainly from reaction between secreted HCO,- and H+ that CO, thus formed remains in the instilled solution. Thus diffusion of CO, out from and into the intragastric solution must be considered. Intragastric Pco, levels were determined over increasing time intervals in animals with spontaneous H+ secretion (Fig. 2). The pH of the gastric samples was low (~3.5) and no free HCO,- was present. The CO, concentration increased with time, and this relationship was linear with samples collected after 5- and lo-min intervals. A deviation from linearity occurred after time intervals of 15 min (Pco, - 30 mmHg) and longer, but values up to 70 mmHg were recorded in the 60-min instillates. Arterial blood PcoZ in samples obtained by cardiac puncture was measured in some experiments; these values did not exceed 42 mmHg. The deviation from linearity of Pco, with time at values below those in blood could indicate that some intraluminal CO, was removed, due to utilization by the parietal cells in the H+ secretory process (29, 30). The effects of inhibition of H+ secretion were therefore tested. This was done by intravenous injection of the histamine HZ-receptor antagonists, cimetidine or metiamide. Over a 3-h period in untreated stomachs (Fig. 3), the mean rate of H+ secretion fell from 102 to 47 whereas total HC03-, which was present peq*h+, throughout in the form of COZ, remained constant at approximately 35 ,ueq* h-l. Cimetidine (20 mgmkg-l) produced a more rapid decline in H+ rate, and a net
alkaline (total HC03- > total H+) secretion appeared (Fig. 4). Total HC03- increased somewhat after cimetidine from 36.0 to 41.2 .peq*h-l (P < O.Ol), which may indicate that some CO, was removed during H+ secretion. Cimetidine was therefore used to inhibit the H+ secretion in most experiments. Metiamide had a similar effect but the mean times required for the appearance of a net alkaline secretion were 145 t 13 min after cimetidine (20 mgekg-l, n = 13) and 195 t 11 min after metiamide (35 mg*kg+, n = 5). Cimetidine (mol wt 252) was thus a more potent inhibi1
I I
’
1
1
1
I
I
HCO;
(total
)
’
peq, 6’
20 0
1 20
0
I 40
I 60
1 80
1 100
I 120
1,’ 140
time
3. Rates of spontaneous H+ and Means + SE are given. n = 6.
HCO,-
FIG.
pig.
I’ I vwl
’ Clmetldtne
’
’
’
’
’
1 180
160
(min)
secretion
’
’
in guinea
’
I
(20 mg, kg-‘)
I
1 80
60
pC0;
(mm
Hg) 40
70 c
60 20 c
4
1
0
PC02 (mm 30
Hg)
PH 6.00
r
5 .oo
4 .oo
0
I 10
1 20
I 30
I 40
I 50 time
2. Means + SE of Pcoz in lo-ml after different times of instillation. There between intragastric Pcoz and time for duration. Acid secreting (noncimetidine-treated) n = 8 for all points. FIG.
I 60
J
(min)
gastric samples recovered was a linear relationship samples of 5- and lo-min animals were used.
PC02
10
I1 0
20
I 40
11 60
80
1’ 100
120
“1 140
2 .oo 160 time
4. Intravenous administration of cimetidine resulted in appearance of a net alkaline secretion total H+). Means + SE are given. n = 13. FIG.
3.00
180
(mm)
(20 mgekg-l) (total HCO,>
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E538
A.
tor of spontaneous H+ secretion in the guinea pig than metiamide (mol wt 242). Diffusion of CO,. The more rapid rise in COZ in samples collected after short time intervals of instillation (Fig. 2) might also reflect diffision of tissue or blood CO, into the gastric lumen, occurring when Pcoz in the latter was low. This possibility was tested by inhibition of H+ secretion, with a higher dose of cimetidine (40 mgekg-l) than that initially employed (20 mgakg-l). A reduction of the H+ secretory rate to values below 1 peq.10 min-l and reduction of Pco2 values to below 5 mmHg in 10 min were recorded without affecting the total output of HCO,- (free HCO,- + CO,). Ditision of CO, thus seems to be slow in the H+ inhibited mucosa. An additional test of the source of intragastric CO, was made by instilling exogenous HCl solutions (NaCl to isotonicity) into stomachs with steady rates of net alkaline secretion after cimetidine treatment (Fig. 5). Exogenous HCl caused a fall in pH together with an increase in CO,. Free HCO,- was reduced to zero but there was no significant change in total HCO,-. There was quantitative agreement between loss of exogenous H+ (i.e., neutralization), decrease in gastric free HC03-, and increase in CO, formation (Table 1). This effect very probably excludes rapid escape of CO, from the gastric lumen. Effects of carbachol on H+ and HC03- transport. Cholinergic stimuli increase H+ secretion in several
I
exogenous HCl (4.3 peq 1
GARNER
AND
G.
FLEMSTROM
species in vivo (18). In vitro, stimulation of both H+ secretion and alkalinization has been demonstrated (9). In a first series of experiments, the effect of carbachol (1 or 2 pgmkg-I) was investigated in animals with spontaneous net acid secretion (noncimetidine-treated). As shown in Fig. 6, the drug produced a marked increase in the rates of secretion of both H+ and HCO,-. The greater rise in total H+ was reflected by a fall in pH. The smaller increase in total HCO,- output all appeared in the form of CO,. Values returned to predrug levels within 20-30 min. In a second series of experiments, carbachol (2 pg=kg-l) was administered to animals pretreated with 20 mgmkg-l cimetidine to induce an alkaline net secretion (Fig. 7). Carbachol increased the mean output of total HCO,- more than that of total H+, and there was TABLE 1. Effects of exogenous HCL on gastric CO, and free HC03HCl Instilled
H+ Recovered
4.3 9.3
1.5 + 0.4 7.1 2 0.3
H+ Neutralized
Increase in Gastric CO,
Decrease in Gastric HCO,-
n
2.8 it@; 3 2.2 + 0:3
2.5 2 0.4 2.3 + 0.3
2.5 + 0.5 1.9 t 0.2
(5)
(6)
Values are means + SE. Consecutive lo-min gastric instillations of 10 ml of isotonic mannitol solution were performed in animals pretreated with cimetidine (20 mg kg-‘) until steady outputs of free HC03were obtained (see Fig. 5). A mannitol solution containing HCl was then instilled for one lo-min period followed by further lomin instillations of mannitol alone. The amount of exogenous HCl neutralized by the stomach was calculated as HCl instilled minus HCl recovered. No significant change in the gastric outputs of total HC03or Na+, K+, and Cl- occurred during these experiments. l
wq In each
sample
r
/
Carbachol (1 pg, kg-‘)
HCO; (total)
6 i
01
’
pC02 (mm 50
4.00l----J
0
20
GO
60
I
1
1
1
I
1
Hg)
PH
r
1
5.00
80 time
FIG. 5. Effect of instilling exogenous net alkaline secretion (total HCO,pretreated with 20 mgekg-’ cimetidine HCO,secretion was obtained. Note HCO,and increase in CO,.
I
(min)
HCl into a stomach with a > total H+). Animal was until a steady rate of free equivalent decrease in free
0
20
40
60
80
100
120
140 time
FIG. 6. Effect of carbachol (1 pg*kg-l) in net acid secreting (noncimetidine-treated) both secretions are stimulated and that than total HC09-.
160
(mtn)
on H+ and HCO,outputs stomach. Note that total H+ increases by more
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
peq,
HCO,-
Carbachol (2 pg, kg-‘)
6'
E539
SECRETION
T
(
1
1
pC0,
(mm
I
Hg)
pg
e kg-')
PH - 700 I
T
- 600
- 5.50
1 10
I 20
I 30
I 40 time
I 50
1
1
0
- 650
' 0
2
I
25 -
101
Carbachol (2 pgt kg-‘)
Carbachol
250
1
I
I
I
40
20
60
80
I
I
100 tbme (mln)
FIG. 8. This experiment illustrates that repeated injection of carbachol produced repeated stimulation of gastric secretions without affecting magnitude of response. Gastric net alkaline secretion (total HCO,- > total H+) was induced by pretreatment with cimetidine (20 fig-kg-l).
J 5.00
(mln)
I”
7. Effect of carbachol (2 pg*kg-l) on ion outputs in stomachs with a net alkaline secretion induced by cimetidine (20 mg/kg-l) or metiamide (35 mgekg-l). As with net acid secreting preparations (Fig. 6), carbachol stimulated both secretions, but in this case total HCO,- increased by more than total H+ and there was an increase in pH. Means + SE are given. n = 17. FIG.
an increase in pH. The rise in total HCO,- was mainly in the form of free HCOS-, the rise in CO, being smaller than that observed in the net acid secreting stomach (Fig. 6). The net influx of Na+, K+, and Cl- ions was also measured in animals with an alkaline net secretion (Fig. 7). Only mannitol solutions were instilled in the stomach. Under steady-state conditions, the relationship between the outputs of the ions studied was: Na+ > Cl- > HCO,- > H+ > K+. All these differences were significant (P C 0.01, n = 17). The output of Cl- was also greater than that of H+ plus K+ (P c 0.01). On stimulation by carbachol, however, there was no significant difference between the increase in Na+ (61 t 9 eq*h-l) and that in HCO,- (60 t 8) or between that in Cl- (76 t 13) and that in H+ plus K+ (69 t 10). Neither blood nor glucose were detected in the samples. Any major secretion of ions other than those measured here seems unlikely because no consistent excess of anions or cations was observed. Effect of atropin. In animals with a net alkaline secretion after administration of 20 mgekg-l cimetidine,
’
’
’
”
”
At ropine (100 pg. kg-‘)
Carbachol (2 pg, kg-‘)
’
’
Carbachol kg-‘)
(2 yg,
100 -
50 -
oi
’ 0
’
’
20
’
’
40
’
’
60
’
’
’
80 time
’
100 (min)
9. Injections of atropin (100 pg*kg-l) completely inhibitea stimulation of gastric HC03- and H+ secretions by carbachol (2 pg=kg-‘) but had no effect on basal ion outputs. FIG.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E540 repeated injections of carbachol (2 ,ug*kg+) stimulated ion outputs without affecting the nature or magnitude of the response (Fig. 8). This provided a convenient means of testing the effect of possible inhibitors. Atropin (100 pgekg-l), injected 30 min prior to the stimulant completely inhibited the response to carbachol (n = 5) An example of one such experiment is shown in Fig. 9 This dose of atropin, however, did not affect the basal output of any of the ions measured. DISCUSSION
The aim of the present work was to develop a method to investigate HCO,- transport into the gastric lumen in vivo and to study some properties of this transport. It was found that the stomach secreted HCO,- at a steady basal rate and that this transport could be stimulated by low doses of carbachol. Atropin inhibited this response but did not affect the basal rate of HCO,secretion. Evaluation of method. In most animals there was some spontaneous secretion of HCl that converted HCO,- into CO,. It was necessary, therefore, to consider gastric intraluminal CO, in the analysis of gastric output of HCO,- (and H+) and to establish that this CO, arose mainly from reaction between secreted HCO,and H+. Diffusion of tissue or blood CO, into the gastric lumen was unlikely because the alkaline secretion that occurred during powerful inhibition of H+ secretion was mainly in the form of free HCO,- and very little CO, appeared. The quantitative agreement between release of CO, and decrease in free HCO,- when exogenous HCl was instilled seems to exclude also rapid escape of CO,. These observations are in agreement with previous studies in dogs (23) and amphibian mucosa (21), which have indicated that gastric mucosal permeability to CO, is low. In the net acid secreting (noninhibited) stomach, all HCO,- appeared in the form of CO,. Total HCO,- (CO, plus free HCOJ increased lo-15% when H+ secretion was inhibited. This result may be due to removal of some luminal CO, by the H+-secreting mucosa, resulting in a slight underestimate of the HCO,- (and H+) secretion. Removal of CO, during H+ secretion may reflect the utilization of CO, by the parietal cells in the H+ secretory process (29, 30) or be due to the greater mucosal blood flow during H+ secretion (5). Origin of-gastric HCO,-. Fundic (9, 10) and most of the antral (12, 13) transport of HCO,- in vitro is an active process and very probably originates from gastric surface epithelial cells. It is stimulated by carbachol, DB-cGMP and Ca2+ ions, whereas histamine, gastrin, and DB-CAMP have no effect. Carbachol (Fig. 7) and Ca2+ ions (unpublished results) stimulated HCO,transport also in the guinea pig, and basal secretion of HCO,- amounted to 5-10% of histamine-stimulated (maximal) H+ secretion (4) both in vitro and in vivo. These similarities may suggest that HCO,- transport in vivo is an active process. Some of the basal (atropin insensitive) HCO,- transport in the guinea pig may, however, reflect passive
A.
GARNER
AND
G.
FLEMSTROM
leakage of this ion through the mucosa. Passive diffusion of HCO,- contributes to gastric luminal alkalinization in vitro in antral but not in fundic epithelium and both these parts of the stomach were included in the present preparation. Gastric arterial injection of acetylcholine (5-50 pg*kg-l) or an artificially applied gastric arterial pressure of 200 mmHg, increases mucosal interstitial pressure and causes ultrafiltration of HCO,- into the dog gastric fundus (1, 2). However, this ultrafiltrate also contains glucose and lo-20 gl-l of protein. It bears a marked resemblance to the fluid appearing in the gastric lumen when the paracellular pathways of the mucosa are increased by exposure to high concentrations of unionized acetylsalicylic acid (6, 11, 16) or dithiothreitol (25). The technique for collecting samples in the present study minimized gastric mucosal trauma (15) and only low doses of carbachol (l-2 pgekg-l) injected into a peripheral vein, were employed. Neither blood nor glucose were detected during the 4- to 5-h duration of the experiments, and stable rates of ion transport were recorded. This finding very probably indicates that the permeability of the mucosa was normal. Previous work suggests some correlation between gastric secretion of H+ plus K+ and Cl- tra-nsport (27, 28). Furthermore, HCO,- and Na- transport were stimulated simultaneously in the present study, and it is possible that Na+ is- the passive coion during active HCO,- transport. Basal (unstimulated) transport of H+ plus K+ was, however, smaller than that of Cl-, and HCO,- was smaller than Na+. It is therefore likely that, in addition to parietal cell transport of (H+ + K+) Cland surface epithelial cell transport of HCO,- (+Na+), there was some passive diffusion of Na+ and Cl- into the gastric lumen. Passive components in gastric transepithelial transport of Cl- and Na+ have been reported before in the isolated amphibian mucosa (19). Role of gastric HCO,- transport. The present work provides direct evidence that transport of HCO,- into the gastric lumen occurs simultaneously with H+ secretion. Basal secretion of HCO,- (5-10% of maximal H+ secretion) is quantitatively sufficient to account for the well-known (24, 26) continuous loss of H+ ions from the lumen of the normal stomach. The results are in accordance with a previous hypothesis of Hollander (20) based, however, on indirect evidence from experiments in which only the net acid output was measured. Reaction between H+ and HCO,- should reduce the osmolarity of the mixed secretion. It has been reported (3) that instillation of isotonic HCl solution into the dog fundic pouch causes a reduction in osmolarity together with a decrease in H+ and an increase in Na+ concentrations. Some of the excess water may, however, diffuse along its chemical gradient into the mucosa (7). Due to the higher mobility and usual excess of the H+ ions, it is likely that neutralization of H+ by HCO,occurs in the immediate vicinity of the luminal surface (10). The net result of HCO,- (+Na+) transport might then appear as an interdiffusion between H+ and Na+ ions (31). Surface neutralization would reduce the con-
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
HCO,-
E541
SECRETION
centration of H+ ions at the luminal surface and may protect the epithelium from damage. Acetazolamide, which inhibits gastric HC03- transport (9, 12) decreases the ability of the mucosa to resist luminal HCl (32). Adrenergic alpha receptor agonists (10) and acetylsalicylate (14, 16) inhibits gastric HC03- transport, and this may contribute to ulcerogenic properties of these agents.
The
authors
wish
to thank
Miss
Carin
Hulth
for skillful
technical
assistance-
The study was supported by grants from The Royal Society (London), the Swedish Medical Research Council (04X-3515), and the Medical Faculty of the University of Uppsala. A. Garner was a European Postdoctoral Fellow of The Royal Society (London). His present address: Dept. of Pharmacology, I. C.I. Ltd., Macclesfield SK10 4TG, England. Received
for
publication
27 June
1977.
REFERENCES 1. ALTAMIRANO, M., M. REQUENA, AND R. P. DURBIN. Effects of gastric arterial and venous pressures on gastric secretion in the dog. Am. J. Physiol. 227: 152-160, 1974. 2. ALTAMIRANO, M., M. REQUENA, AND T. C. PEREZ. Interstitial fluid pressure and alkaline gastric secretion. Am. J. Physiol. 229: 1421-1426, 1975. 3. BERKOWITZ, J. M., AND H. D. JANOWITZ. Alteration in composition of hydrochloric acid solutions by gastric mucosa. Am. J. Physiol. 210: 216-220, 1966. 4. BREIDENBACH, A. W., AND F. E. RAY. Effects of ascorbic acid on gastric secretion in guinea-pigs. Am. J. Physiol. 180: 637-640, 1955. 5. CHEUNG, L. Y. AND S. F. LOWRY. Canine gastric blood flow and oxygen consumption during cimetidine inhibition of acid secretion. Surg. Forum 27: 390-392, 1976. 6. DAVENPORT, H. W. Fluid produced by the gastric mucosa during damage by acetic and salicylic acids. Gastroenterology 50: 487499, 1966. 7. DURBIN, R. P. Coupling of water to solute movement in isolated gastric mucosa. Ciba Found. Symp. 38: Lung Liquids (new series). New York: Elsevier, 1976, p. 161-177. 8. FLEMSTR~M, G. Properties of isolated gastric antral and SCNinhibited fundic mucosa. In: Gastric Hydrogen Ion Secretion, edited by D. K. Kasbekar, G. Sachs, and W. S. Rehm. New York: Dekker, 1976, p. 102-116. 9. FLEMSTR~M, G. Active alkalinization by amphibian gastric fundic mucosa in vitro. Am. J. Physiol. 233: El-E12, 1977 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. PhysioZ. 2: ElE12, 1977. 10. FLEMSTR~M, G. Effect of catecholamines, Ca++ and gastrin on gastric HCO,secretion. Acta Physiol. Stand., Suppl.: Gastric Ion Transport, 1978, 81-90. 11. FLEMSTR~M, G., N. V. B. MARSDEN, AND W. RICHTER. Passive cutaneous anaphylaxis in guinea pigs elicited by gastric absorption of dextran induced by acetylsalicylic acid. Intern. Arch. Allergy AppZ. ImmunoZ. 51: 627-636, 1976. 12. FLEMSTR~M, G., AND G. SACHS. Ion transport by amphibian antrum in vitro. I. General characteristics. Am. J. PhysioZ. 228: 1188-1198, 1975. 13. FROMM, D., J. H. SCHWARTZ, R. ROBERSTON, AND R. FUHRO. Ion transport across isolated antral mucosa of the rabbit. Am. J. Physiol. 231: 1783-1789, 1976. 14. GARNER, A. Effects of acetylsalicylate on alkalinization, acid secretion and electrogenic properties in the isolated gastric mucosa. Acta PhysioZ. Stand. 99: 281-291, 1977. 15. GARNER, A. Influence of salicylates on the rate of accumulation of deoxyribonucleic acid in gastric washings from the guinea-
pig. ToxicoZ. AppZ. PharmacoZ. 42: 119-128, 1977. 16. GARNER, A. Mechanisms of action of aspirin on the gastric mucosa of the guinea pig. Acta PhysioZ. Stand., Suppl.: Gastric Ion Transport, 1978, 101-110. 17. GROSSMAN, M. I. The secretion of the pyloric glands of the dog. Intern. Congr . Physiol. Sci . , 21st, Buenos Aires, 1959, p. 226228. 18. HIRSCHOWITZ, B. I., AND G. SACHS. Lessons from the intact stomach. In: Gastric Secretion, edited by G. Sachs, E. Heinz, and K. J. Ullrich. New York: Academic, 1972, p. 7-34. 19. HOGBEN, C. A. M. Active transport of chloride by isolated frog gastric epithelium. Origin of the gastric potential. Am. J. Physiol. 180: 641-649, 1955. 20. HOLLANDER, F. The two-components mucous barrier. Arch. InternaL Med. 94: 107-120, 1954. 21. IMAMURA, A. The effects of carbon dioxide and bicarbonate on chloride fluxes across frog gastric mucosa. Biochim. Biophys. Acta 196: 245-253, 1970. 22. KARLMARK, B., AND B. G. DANIELSON. Titrable acid, PcoZ, bicarbonate and ammonium ions along the rat proximal tubule. Acta Physiol. &and. 91: 242-258, 1974. 23. KURTZ, L. D., AND B. B. CLARK. The inverse relationship of the secretion of hydrochloric acid to the tension of carbon dioxide in the stomach. GastroenteroZogy 9: 594-602, 1947. 24. MAKHLOUF, G. M., J. P. A. MCMANUS, AND W. I. CARD. A quantitative statement of the two-component hypothesis of gastric secretion. Gastroenterology 51: 149-171, 1966. 25. MUNRO, D. R. Route of protein loss during a model proteinlosing gastropathy in dogs. GastroenteroZogy 66: 960-972, 1974. 26. OBRINK, K. J. Studies on the kinetics of the parietal secretion of the stomach. Acta PhysioZ. &and. 15: Suppl. 51, 1948. 27. OBRINK, K. J. Sodium secretion as part of the formation of gastric juice. Uppsala J. Med. Sci. 80: 65-70, 1975. 28. RAY, T. K., AND L. L. TAGUE. Role of K+-stimulated ATPase in H+ and K+ transport by bullfrog gastric mucosa in vitro. Acta Physiol. Stand., Suppl.: Gastric Ion Transport, 1978, 283-292. 29. RUSSEL, J. C., AND K. KOWALEWSKI. The acid-base balance in gastric H+ secretion. Physiol. Chem. Physics 6: 269-274, 1974. 30. SANDERS, S. S., V. B. HAYNE, AND W. S. REHM. Normal H+ rates in frog stomach in absence of exogenous CO, and a note on pH stat method. Am. J. PhysioZ. 225: 1311-1321, 1973. 31. TEORELL, T. Electrolyte diffusion in relation to the acidity regulation of the gastric juice. GastroenteroZogy 9:. 425-443, 1947. 32. WERTHER, J. L., F. HOLLANDER, AND M. ALTAMIRANO. Effect of acetazolamide on gastric mucosa in canine vivo-vitro preparations. Am. J. Physiol. 209: 127-133, 1965.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GARNER
secretion in the guinea pig AND
of Physiology
GUNNAR
FLEMSTROM
and Medical Biophysics,
GARNER, ANDREW, AND GUNNAR FLEMSTR~M. Gastric HCO,secretion in the guinea pig. Am. J. Physiol. 234(6): E535-E541, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 3(6): E535-E541, 1978. - Measurement of gastric intraluminal Pcoz and pH in the anesthetized guinea pig enabled simultaneous determination of total H+ and HCO,- gastric secretions. There was quantitative agreement between the release of CO, and decrease in HCO,after intragastric instillation of HCl. The basal rate of HCO,secretion (-40 peq=h+) was, in most cases, smaller than spontaneous H+ secretion, but gastric net secretory output was alkaline (HCO,- > H+) after inhibition of acid secretion with histamine Hz-receptor antagonists (cimetidine 20 mgekg-l or metiamide 35 mg=kg+). Carbachol (l-2 pg*kg-l) stimulated secretion of both HCO,- and H+; only the latter response was sensitive to the histamine antagonists. Atropin (100 ,ug.kg+) blocked stimulation of HCO,- secretion but did not affect the basal output of HCO,-. An increase in HC03secretion was associated with an equivalent increase in net Na+ influx and an increase in the net influx of Cl- with H+ plus K+. Intragastric neutralization of H+ by HCO,- is likely to occur at the mucosal surface and may protect the mucosa from the damaging effects of intraluminal acid. atropin, carbachol; H+ secretion; HCO,- secretion; Hz-receptor antagonists; intragastric Pco2
histamine
ACTIVE SECRETION of HCO,- by the gastric surface epithelial cells was recently demonstrated in antral (12, 13) and fundic (8-10) mucosa in vitro. This secretion was sensitive to acetazolamide and stimulated by carbachol, DB-cGMP and Ca2+ ions. Secretion of HCO,probably hreduces H+ ion concentration at the luminal surface of the gastric epithelium and thus may act as a protective mechanism. The aim of the present study was to determine whether a similar secretion occurs under in vivo conditions and to examine some of the properties of HCO,transport in the mammalian stomach. The occurrence of free HCO,- in gastric nonacid secretions in vivo has been reported previously (17, 20); however, at the low values of pH prevailing in the stomach, HC03- would rapidly decompose to CO, and water. Measurements of intragastric Pco2 and pH were used here for simultaneous determination of total gastric HCO,- and H+ secretions. MATERIALS
AND
METHODS
Male albino guinea pigs (300-420 g) were starved for 20 h before use in cages with mesh bottoms to minimize coprophagy, but allowed free access to drinking water. 0363-6100/78/0000-0000$01.25
Copyright
0 1978 the American
University
of Uppsala,
Uppsala,
Sweden
Animals were anesthetized with urethane (1.5 g-kg-‘), administered intraperitoneally. Body temperature was maintained at 37°C with a heater controlled by an intrarectal thermistor. Surgical procedures. The abdomen was opened by a median epigastric incision and a multiorifice soft rubber tube (inset, Fig. 1) passed into the stomach through a cut in the duodenum proximal to the insertion of the common bile duct. The tube was secured in place by means of a ligature around the pylorus, with care taken to exclude blood vessels from the tie, and exteriorized via a stab wound in the right flank before the abdomen was closed. After tracheostomy, a soft catheter (umbilical artery catheter size 8 Fr., Sherwood Medical Industries, Inc., St. Louis, MO.) was passed down the esophagus into the stomach and positioned so that the tip terminated just beyond the cardia. This catheter was held in place by a tie around the esophagus in the neck region, a procedure which also prevented the contamination of gastric samples with salivary secretions. The left external jugular vein was cannulated for drug administration. Collection of gastric samples. Before beginning an experiment, the stomach was rinsed clear of food debris with lo-ml volumes of isotonic NaCl or mannitol at 37OC. Fluid was introduced via the esophageal catheter and allowed to escape through the pyloric outlet while the abdominal surface of the animal was gently massaged. Complete removal of the fasting gastric contents was obtained after 6-10 rinses. A diagram of the equipment used for filling and emptying the stomach is shown in Fig. 1. The animal was secured to a hinged table and the esophageal catheter attached to a 3-way tap (A). The pyloric tube was connected to a second 3-way tap (B) by means of glass tubing. The stomach was then filled with 10 ml of fluid at 37°C from a heated reservoir, with the animal tilted so that the pyloric region was uppermost (Fig. 1, position 1) allowing air in the stomach to escape via the 3-way tap (B). Fluid was recovered by tilting the animal so that the pyloric region was the lowest point of the stomach (Fig. 1, position 2) and turning on the flow inducer (WatsonMarlow, Falmouth, England) set at 100 mlamin-l with the 3-way taps adjusted to form a closed circuit. Intragastric fluid was thus replaced with an equal volume of air. The method allowed rapid filling and emptying of the stomach (approximately 15 s for each), and stretching of the gastric wall during these procedures was minimized. Between 96 and 103% of the instilled volume was recovered after 10 min.
Physiological
Society
E535
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E536
A. GARNER
combine to form COZ, representing ization
AN’D G. FLEMSTROM
intragastric
neutral-
H+ + HC03- = CO, + H,O
FIG. 1. Diagram of apparatus used to collect gastric samples from guinea pig. Animal tilted through an angle of 60” betweenposition 1 (filling stomach) and position 2 (emptying stomach); otherwise it was placed horizontally. Inset shows a diagram of pyloric outlet tube and position of cannula used to remove intragastric samples for measurements of Pco, levels.
In most experiments, lo-ml volumes of either isotonic NaCl or mannitol solutions were instilled and removed for analysis after 10 min. Mannitol was used in experiments in which gastric outputs of Na+ and Cl- were determined. There was no difference between NaCl and mannitol instillations with respect to gastric outputs of HCO,- and H+ and some of these data were therefore pooled. Measurement of intragastric Pco, and PH. Intragastric Pco, was measured immediately before removal of the instilled solutions with a Pco, electrode (E5036, Radiometer, Copenhagen) connected to an expanded scale millivolt meter. Samples were withdrawn into the electrode by means of gentle suction from a syringe attached to the electrode outlet. The exposed section of this tube was made from glass capillary to minimize escape of CO, by diffusion. Before and after each experiment, the electrode was calibrated with gas mixtures containing known concentrations of CO, in the range of 1.5-12.8%. Calibration and all measurements were carried out at 37°C. The rest of the instilled fluid (-9 ml) was collected in a glass test tube and the pH measured with a glass calomel electrode (GK-2321, Radiometer, Copenhagen) connected to an expanded scale pH meter. The final volume of each gastric sample was recorded after addition of fluid used for PCS measurement. The sample was then gassed with 100% N1, which had been prewashed in Ba(OH),. Immediately after gassing for 5-15 min, a 5-ml aliquot was titrated to pH 7.00 with 5 mM NaOH. In the case of samples, alkaline after gassing (HCO,- secretion > H+ secretion), remaining HCO,was determined by back titration after acidification with 50 ~1 of 100 mM HCl standard solution. Analysis of gastric HCO,- and H+ secretions. The following principles were used. a) Some HCO,- and H+
The equal amounts of HCO,- and H+ thus removed intragastrically were designated HCO,- (CO,) and H+ (CO,) and calculated from the intragastric PcoZ and the solubility coefficient of CO,. b) Substitution of measured pH and Pco, in the Henderson-Hasselbach equation enabled calculation of free HCOI- in the instillate pH = pK, + log (HC03-/~ Pco,) Values used for PK,, the dissociation constant, and (Y, the solubility coefficient, were 6.09 and 0.03, respectively. These values were determined for measurement of HCO,- in kidney tabular fluid (22) which has a composition similar to the gastric samples examined here. c) Gassing of the recovered instillate with Nz promoted formation of COZ from HCO,- and H+ (by removal of CO, formed intragastrically and during gassing). Acid or base remaining after gassing was determined by titration. Total concentrations of HCO,- and H+ in the instillates were then calculated total HCOI-: HC03- (CO,) plus HC03- (free) total H+: i) in samples acid after gassing with N2 H+ (CO& plus HCOS- (free) plus H+ (titrated) ii) in samples alkaline after gassing with N, H+ (CO,) plus HC03- (free) minus HCO1- (titrated) At pH values below 4.0, amounts of HC03- (free) were negligible. In samples alkaline aRer gassing, HC03(free) minus HCOI- (titrated) represents H+ reacting with HCO$- during the gassing and thus removed. In the above analysis of gastric HC03- and H+, it was assumed that gastric luminal CO, arises from reaction between secreted HCO,- and H+ and that diffusion of CO, out from or into the gastric lumen is slow. Experimental evidence for these assumptions will be presented (vide infra). Other analyses. Concentrations of Na+ and K+ were determined by flame photometry (Eppendorf, Netheler and Hinz, FGR) and Cl- was determined by titration (chloride titrator CMTlO, Radiometer, Copenhagen). The recovered solutions were always tested for hemoglobin and glucose (Labstix, Ames Co., Slough, England). These methods were checked with standard solutions and gave positive reactions for hemoglobin at a concentration of 3 pg+ml-l and for glucose at 500 pgernl-‘. Drugs and chemicals. All drug solutions were adjusted to pH 7.4 and isotonicity before use and injected intravenously in a volume of 1 mlkg-’ body wt. The histamine HZ-receptor antagonists, cimetidine and metiamide, were kind gifts from Dr. M. E. Parsons, Smith, Kline & French Laboratories, Ltd., Welwyn Garden, England. Carbachol (carbamylcholine chloride) and
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
HC03-
E537
SECRETION
atropin (as sulphate) were obtained ical Co., St. Louis, MO.
from Sigma Chem-
RESULTS
Intragastric C02. A prerequisite for the present method is that intraluminal CO, arises mainly from reaction between secreted HCO,- and H+ that CO, thus formed remains in the instilled solution. Thus diffusion of CO, out from and into the intragastric solution must be considered. Intragastric Pco, levels were determined over increasing time intervals in animals with spontaneous H+ secretion (Fig. 2). The pH of the gastric samples was low (~3.5) and no free HCO,- was present. The CO, concentration increased with time, and this relationship was linear with samples collected after 5- and lo-min intervals. A deviation from linearity occurred after time intervals of 15 min (Pco, - 30 mmHg) and longer, but values up to 70 mmHg were recorded in the 60-min instillates. Arterial blood PcoZ in samples obtained by cardiac puncture was measured in some experiments; these values did not exceed 42 mmHg. The deviation from linearity of Pco, with time at values below those in blood could indicate that some intraluminal CO, was removed, due to utilization by the parietal cells in the H+ secretory process (29, 30). The effects of inhibition of H+ secretion were therefore tested. This was done by intravenous injection of the histamine HZ-receptor antagonists, cimetidine or metiamide. Over a 3-h period in untreated stomachs (Fig. 3), the mean rate of H+ secretion fell from 102 to 47 whereas total HC03-, which was present peq*h+, throughout in the form of COZ, remained constant at approximately 35 ,ueq* h-l. Cimetidine (20 mgmkg-l) produced a more rapid decline in H+ rate, and a net
alkaline (total HC03- > total H+) secretion appeared (Fig. 4). Total HC03- increased somewhat after cimetidine from 36.0 to 41.2 .peq*h-l (P < O.Ol), which may indicate that some CO, was removed during H+ secretion. Cimetidine was therefore used to inhibit the H+ secretion in most experiments. Metiamide had a similar effect but the mean times required for the appearance of a net alkaline secretion were 145 t 13 min after cimetidine (20 mgekg-l, n = 13) and 195 t 11 min after metiamide (35 mg*kg+, n = 5). Cimetidine (mol wt 252) was thus a more potent inhibi1
I I
’
1
1
1
I
I
HCO;
(total
)
’
peq, 6’
20 0
1 20
0
I 40
I 60
1 80
1 100
I 120
1,’ 140
time
3. Rates of spontaneous H+ and Means + SE are given. n = 6.
HCO,-
FIG.
pig.
I’ I vwl
’ Clmetldtne
’
’
’
’
’
1 180
160
(min)
secretion
’
’
in guinea
’
I
(20 mg, kg-‘)
I
1 80
60
pC0;
(mm
Hg) 40
70 c
60 20 c
4
1
0
PC02 (mm 30
Hg)
PH 6.00
r
5 .oo
4 .oo
0
I 10
1 20
I 30
I 40
I 50 time
2. Means + SE of Pcoz in lo-ml after different times of instillation. There between intragastric Pcoz and time for duration. Acid secreting (noncimetidine-treated) n = 8 for all points. FIG.
I 60
J
(min)
gastric samples recovered was a linear relationship samples of 5- and lo-min animals were used.
PC02
10
I1 0
20
I 40
11 60
80
1’ 100
120
“1 140
2 .oo 160 time
4. Intravenous administration of cimetidine resulted in appearance of a net alkaline secretion total H+). Means + SE are given. n = 13. FIG.
3.00
180
(mm)
(20 mgekg-l) (total HCO,>
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E538
A.
tor of spontaneous H+ secretion in the guinea pig than metiamide (mol wt 242). Diffusion of CO,. The more rapid rise in COZ in samples collected after short time intervals of instillation (Fig. 2) might also reflect diffision of tissue or blood CO, into the gastric lumen, occurring when Pcoz in the latter was low. This possibility was tested by inhibition of H+ secretion, with a higher dose of cimetidine (40 mgekg-l) than that initially employed (20 mgakg-l). A reduction of the H+ secretory rate to values below 1 peq.10 min-l and reduction of Pco2 values to below 5 mmHg in 10 min were recorded without affecting the total output of HCO,- (free HCO,- + CO,). Ditision of CO, thus seems to be slow in the H+ inhibited mucosa. An additional test of the source of intragastric CO, was made by instilling exogenous HCl solutions (NaCl to isotonicity) into stomachs with steady rates of net alkaline secretion after cimetidine treatment (Fig. 5). Exogenous HCl caused a fall in pH together with an increase in CO,. Free HCO,- was reduced to zero but there was no significant change in total HCO,-. There was quantitative agreement between loss of exogenous H+ (i.e., neutralization), decrease in gastric free HC03-, and increase in CO, formation (Table 1). This effect very probably excludes rapid escape of CO, from the gastric lumen. Effects of carbachol on H+ and HC03- transport. Cholinergic stimuli increase H+ secretion in several
I
exogenous HCl (4.3 peq 1
GARNER
AND
G.
FLEMSTROM
species in vivo (18). In vitro, stimulation of both H+ secretion and alkalinization has been demonstrated (9). In a first series of experiments, the effect of carbachol (1 or 2 pgmkg-I) was investigated in animals with spontaneous net acid secretion (noncimetidine-treated). As shown in Fig. 6, the drug produced a marked increase in the rates of secretion of both H+ and HCO,-. The greater rise in total H+ was reflected by a fall in pH. The smaller increase in total HCO,- output all appeared in the form of CO,. Values returned to predrug levels within 20-30 min. In a second series of experiments, carbachol (2 pg=kg-l) was administered to animals pretreated with 20 mgmkg-l cimetidine to induce an alkaline net secretion (Fig. 7). Carbachol increased the mean output of total HCO,- more than that of total H+, and there was TABLE 1. Effects of exogenous HCL on gastric CO, and free HC03HCl Instilled
H+ Recovered
4.3 9.3
1.5 + 0.4 7.1 2 0.3
H+ Neutralized
Increase in Gastric CO,
Decrease in Gastric HCO,-
n
2.8 it@; 3 2.2 + 0:3
2.5 2 0.4 2.3 + 0.3
2.5 + 0.5 1.9 t 0.2
(5)
(6)
Values are means + SE. Consecutive lo-min gastric instillations of 10 ml of isotonic mannitol solution were performed in animals pretreated with cimetidine (20 mg kg-‘) until steady outputs of free HC03were obtained (see Fig. 5). A mannitol solution containing HCl was then instilled for one lo-min period followed by further lomin instillations of mannitol alone. The amount of exogenous HCl neutralized by the stomach was calculated as HCl instilled minus HCl recovered. No significant change in the gastric outputs of total HC03or Na+, K+, and Cl- occurred during these experiments. l
wq In each
sample
r
/
Carbachol (1 pg, kg-‘)
HCO; (total)
6 i
01
’
pC02 (mm 50
4.00l----J
0
20
GO
60
I
1
1
1
I
1
Hg)
PH
r
1
5.00
80 time
FIG. 5. Effect of instilling exogenous net alkaline secretion (total HCO,pretreated with 20 mgekg-’ cimetidine HCO,secretion was obtained. Note HCO,and increase in CO,.
I
(min)
HCl into a stomach with a > total H+). Animal was until a steady rate of free equivalent decrease in free
0
20
40
60
80
100
120
140 time
FIG. 6. Effect of carbachol (1 pg*kg-l) in net acid secreting (noncimetidine-treated) both secretions are stimulated and that than total HC09-.
160
(mtn)
on H+ and HCO,outputs stomach. Note that total H+ increases by more
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
peq,
HCO,-
Carbachol (2 pg, kg-‘)
6'
E539
SECRETION
T
(
1
1
pC0,
(mm
I
Hg)
pg
e kg-')
PH - 700 I
T
- 600
- 5.50
1 10
I 20
I 30
I 40 time
I 50
1
1
0
- 650
' 0
2
I
25 -
101
Carbachol (2 pgt kg-‘)
Carbachol
250
1
I
I
I
40
20
60
80
I
I
100 tbme (mln)
FIG. 8. This experiment illustrates that repeated injection of carbachol produced repeated stimulation of gastric secretions without affecting magnitude of response. Gastric net alkaline secretion (total HCO,- > total H+) was induced by pretreatment with cimetidine (20 fig-kg-l).
J 5.00
(mln)
I”
7. Effect of carbachol (2 pg*kg-l) on ion outputs in stomachs with a net alkaline secretion induced by cimetidine (20 mg/kg-l) or metiamide (35 mgekg-l). As with net acid secreting preparations (Fig. 6), carbachol stimulated both secretions, but in this case total HCO,- increased by more than total H+ and there was an increase in pH. Means + SE are given. n = 17. FIG.
an increase in pH. The rise in total HCO,- was mainly in the form of free HCOS-, the rise in CO, being smaller than that observed in the net acid secreting stomach (Fig. 6). The net influx of Na+, K+, and Cl- ions was also measured in animals with an alkaline net secretion (Fig. 7). Only mannitol solutions were instilled in the stomach. Under steady-state conditions, the relationship between the outputs of the ions studied was: Na+ > Cl- > HCO,- > H+ > K+. All these differences were significant (P C 0.01, n = 17). The output of Cl- was also greater than that of H+ plus K+ (P c 0.01). On stimulation by carbachol, however, there was no significant difference between the increase in Na+ (61 t 9 eq*h-l) and that in HCO,- (60 t 8) or between that in Cl- (76 t 13) and that in H+ plus K+ (69 t 10). Neither blood nor glucose were detected in the samples. Any major secretion of ions other than those measured here seems unlikely because no consistent excess of anions or cations was observed. Effect of atropin. In animals with a net alkaline secretion after administration of 20 mgekg-l cimetidine,
’
’
’
”
”
At ropine (100 pg. kg-‘)
Carbachol (2 pg, kg-‘)
’
’
Carbachol kg-‘)
(2 yg,
100 -
50 -
oi
’ 0
’
’
20
’
’
40
’
’
60
’
’
’
80 time
’
100 (min)
9. Injections of atropin (100 pg*kg-l) completely inhibitea stimulation of gastric HC03- and H+ secretions by carbachol (2 pg=kg-‘) but had no effect on basal ion outputs. FIG.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
E540 repeated injections of carbachol (2 ,ug*kg+) stimulated ion outputs without affecting the nature or magnitude of the response (Fig. 8). This provided a convenient means of testing the effect of possible inhibitors. Atropin (100 pgekg-l), injected 30 min prior to the stimulant completely inhibited the response to carbachol (n = 5) An example of one such experiment is shown in Fig. 9 This dose of atropin, however, did not affect the basal output of any of the ions measured. DISCUSSION
The aim of the present work was to develop a method to investigate HCO,- transport into the gastric lumen in vivo and to study some properties of this transport. It was found that the stomach secreted HCO,- at a steady basal rate and that this transport could be stimulated by low doses of carbachol. Atropin inhibited this response but did not affect the basal rate of HCO,secretion. Evaluation of method. In most animals there was some spontaneous secretion of HCl that converted HCO,- into CO,. It was necessary, therefore, to consider gastric intraluminal CO, in the analysis of gastric output of HCO,- (and H+) and to establish that this CO, arose mainly from reaction between secreted HCO,and H+. Diffusion of tissue or blood CO, into the gastric lumen was unlikely because the alkaline secretion that occurred during powerful inhibition of H+ secretion was mainly in the form of free HCO,- and very little CO, appeared. The quantitative agreement between release of CO, and decrease in free HCO,- when exogenous HCl was instilled seems to exclude also rapid escape of CO,. These observations are in agreement with previous studies in dogs (23) and amphibian mucosa (21), which have indicated that gastric mucosal permeability to CO, is low. In the net acid secreting (noninhibited) stomach, all HCO,- appeared in the form of CO,. Total HCO,- (CO, plus free HCOJ increased lo-15% when H+ secretion was inhibited. This result may be due to removal of some luminal CO, by the H+-secreting mucosa, resulting in a slight underestimate of the HCO,- (and H+) secretion. Removal of CO, during H+ secretion may reflect the utilization of CO, by the parietal cells in the H+ secretory process (29, 30) or be due to the greater mucosal blood flow during H+ secretion (5). Origin of-gastric HCO,-. Fundic (9, 10) and most of the antral (12, 13) transport of HCO,- in vitro is an active process and very probably originates from gastric surface epithelial cells. It is stimulated by carbachol, DB-cGMP and Ca2+ ions, whereas histamine, gastrin, and DB-CAMP have no effect. Carbachol (Fig. 7) and Ca2+ ions (unpublished results) stimulated HCO,transport also in the guinea pig, and basal secretion of HCO,- amounted to 5-10% of histamine-stimulated (maximal) H+ secretion (4) both in vitro and in vivo. These similarities may suggest that HCO,- transport in vivo is an active process. Some of the basal (atropin insensitive) HCO,- transport in the guinea pig may, however, reflect passive
A.
GARNER
AND
G.
FLEMSTROM
leakage of this ion through the mucosa. Passive diffusion of HCO,- contributes to gastric luminal alkalinization in vitro in antral but not in fundic epithelium and both these parts of the stomach were included in the present preparation. Gastric arterial injection of acetylcholine (5-50 pg*kg-l) or an artificially applied gastric arterial pressure of 200 mmHg, increases mucosal interstitial pressure and causes ultrafiltration of HCO,- into the dog gastric fundus (1, 2). However, this ultrafiltrate also contains glucose and lo-20 gl-l of protein. It bears a marked resemblance to the fluid appearing in the gastric lumen when the paracellular pathways of the mucosa are increased by exposure to high concentrations of unionized acetylsalicylic acid (6, 11, 16) or dithiothreitol (25). The technique for collecting samples in the present study minimized gastric mucosal trauma (15) and only low doses of carbachol (l-2 pgekg-l) injected into a peripheral vein, were employed. Neither blood nor glucose were detected during the 4- to 5-h duration of the experiments, and stable rates of ion transport were recorded. This finding very probably indicates that the permeability of the mucosa was normal. Previous work suggests some correlation between gastric secretion of H+ plus K+ and Cl- tra-nsport (27, 28). Furthermore, HCO,- and Na- transport were stimulated simultaneously in the present study, and it is possible that Na+ is- the passive coion during active HCO,- transport. Basal (unstimulated) transport of H+ plus K+ was, however, smaller than that of Cl-, and HCO,- was smaller than Na+. It is therefore likely that, in addition to parietal cell transport of (H+ + K+) Cland surface epithelial cell transport of HCO,- (+Na+), there was some passive diffusion of Na+ and Cl- into the gastric lumen. Passive components in gastric transepithelial transport of Cl- and Na+ have been reported before in the isolated amphibian mucosa (19). Role of gastric HCO,- transport. The present work provides direct evidence that transport of HCO,- into the gastric lumen occurs simultaneously with H+ secretion. Basal secretion of HCO,- (5-10% of maximal H+ secretion) is quantitatively sufficient to account for the well-known (24, 26) continuous loss of H+ ions from the lumen of the normal stomach. The results are in accordance with a previous hypothesis of Hollander (20) based, however, on indirect evidence from experiments in which only the net acid output was measured. Reaction between H+ and HCO,- should reduce the osmolarity of the mixed secretion. It has been reported (3) that instillation of isotonic HCl solution into the dog fundic pouch causes a reduction in osmolarity together with a decrease in H+ and an increase in Na+ concentrations. Some of the excess water may, however, diffuse along its chemical gradient into the mucosa (7). Due to the higher mobility and usual excess of the H+ ions, it is likely that neutralization of H+ by HCO,occurs in the immediate vicinity of the luminal surface (10). The net result of HCO,- (+Na+) transport might then appear as an interdiffusion between H+ and Na+ ions (31). Surface neutralization would reduce the con-
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
GASTRIC
HCO,-
E541
SECRETION
centration of H+ ions at the luminal surface and may protect the epithelium from damage. Acetazolamide, which inhibits gastric HC03- transport (9, 12) decreases the ability of the mucosa to resist luminal HCl (32). Adrenergic alpha receptor agonists (10) and acetylsalicylate (14, 16) inhibits gastric HC03- transport, and this may contribute to ulcerogenic properties of these agents.
The
authors
wish
to thank
Miss
Carin
Hulth
for skillful
technical
assistance-
The study was supported by grants from The Royal Society (London), the Swedish Medical Research Council (04X-3515), and the Medical Faculty of the University of Uppsala. A. Garner was a European Postdoctoral Fellow of The Royal Society (London). His present address: Dept. of Pharmacology, I. C.I. Ltd., Macclesfield SK10 4TG, England. Received
for
publication
27 June
1977.
REFERENCES 1. ALTAMIRANO, M., M. REQUENA, AND R. P. DURBIN. Effects of gastric arterial and venous pressures on gastric secretion in the dog. Am. J. Physiol. 227: 152-160, 1974. 2. ALTAMIRANO, M., M. REQUENA, AND T. C. PEREZ. Interstitial fluid pressure and alkaline gastric secretion. Am. J. Physiol. 229: 1421-1426, 1975. 3. BERKOWITZ, J. M., AND H. D. JANOWITZ. Alteration in composition of hydrochloric acid solutions by gastric mucosa. Am. J. Physiol. 210: 216-220, 1966. 4. BREIDENBACH, A. W., AND F. E. RAY. Effects of ascorbic acid on gastric secretion in guinea-pigs. Am. J. Physiol. 180: 637-640, 1955. 5. CHEUNG, L. Y. AND S. F. LOWRY. Canine gastric blood flow and oxygen consumption during cimetidine inhibition of acid secretion. Surg. Forum 27: 390-392, 1976. 6. DAVENPORT, H. W. Fluid produced by the gastric mucosa during damage by acetic and salicylic acids. Gastroenterology 50: 487499, 1966. 7. DURBIN, R. P. Coupling of water to solute movement in isolated gastric mucosa. Ciba Found. Symp. 38: Lung Liquids (new series). New York: Elsevier, 1976, p. 161-177. 8. FLEMSTR~M, G. Properties of isolated gastric antral and SCNinhibited fundic mucosa. In: Gastric Hydrogen Ion Secretion, edited by D. K. Kasbekar, G. Sachs, and W. S. Rehm. New York: Dekker, 1976, p. 102-116. 9. FLEMSTR~M, G. Active alkalinization by amphibian gastric fundic mucosa in vitro. Am. J. Physiol. 233: El-E12, 1977 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. PhysioZ. 2: ElE12, 1977. 10. FLEMSTR~M, G. Effect of catecholamines, Ca++ and gastrin on gastric HCO,secretion. Acta Physiol. Stand., Suppl.: Gastric Ion Transport, 1978, 81-90. 11. FLEMSTR~M, G., N. V. B. MARSDEN, AND W. RICHTER. Passive cutaneous anaphylaxis in guinea pigs elicited by gastric absorption of dextran induced by acetylsalicylic acid. Intern. Arch. Allergy AppZ. ImmunoZ. 51: 627-636, 1976. 12. FLEMSTR~M, G., AND G. SACHS. Ion transport by amphibian antrum in vitro. I. General characteristics. Am. J. PhysioZ. 228: 1188-1198, 1975. 13. FROMM, D., J. H. SCHWARTZ, R. ROBERSTON, AND R. FUHRO. Ion transport across isolated antral mucosa of the rabbit. Am. J. Physiol. 231: 1783-1789, 1976. 14. GARNER, A. Effects of acetylsalicylate on alkalinization, acid secretion and electrogenic properties in the isolated gastric mucosa. Acta PhysioZ. Stand. 99: 281-291, 1977. 15. GARNER, A. Influence of salicylates on the rate of accumulation of deoxyribonucleic acid in gastric washings from the guinea-
pig. ToxicoZ. AppZ. PharmacoZ. 42: 119-128, 1977. 16. GARNER, A. Mechanisms of action of aspirin on the gastric mucosa of the guinea pig. Acta PhysioZ. Stand., Suppl.: Gastric Ion Transport, 1978, 101-110. 17. GROSSMAN, M. I. The secretion of the pyloric glands of the dog. Intern. Congr . Physiol. Sci . , 21st, Buenos Aires, 1959, p. 226228. 18. HIRSCHOWITZ, B. I., AND G. SACHS. Lessons from the intact stomach. In: Gastric Secretion, edited by G. Sachs, E. Heinz, and K. J. Ullrich. New York: Academic, 1972, p. 7-34. 19. HOGBEN, C. A. M. Active transport of chloride by isolated frog gastric epithelium. Origin of the gastric potential. Am. J. Physiol. 180: 641-649, 1955. 20. HOLLANDER, F. The two-components mucous barrier. Arch. InternaL Med. 94: 107-120, 1954. 21. IMAMURA, A. The effects of carbon dioxide and bicarbonate on chloride fluxes across frog gastric mucosa. Biochim. Biophys. Acta 196: 245-253, 1970. 22. KARLMARK, B., AND B. G. DANIELSON. Titrable acid, PcoZ, bicarbonate and ammonium ions along the rat proximal tubule. Acta Physiol. &and. 91: 242-258, 1974. 23. KURTZ, L. D., AND B. B. CLARK. The inverse relationship of the secretion of hydrochloric acid to the tension of carbon dioxide in the stomach. GastroenteroZogy 9: 594-602, 1947. 24. MAKHLOUF, G. M., J. P. A. MCMANUS, AND W. I. CARD. A quantitative statement of the two-component hypothesis of gastric secretion. Gastroenterology 51: 149-171, 1966. 25. MUNRO, D. R. Route of protein loss during a model proteinlosing gastropathy in dogs. GastroenteroZogy 66: 960-972, 1974. 26. OBRINK, K. J. Studies on the kinetics of the parietal secretion of the stomach. Acta PhysioZ. &and. 15: Suppl. 51, 1948. 27. OBRINK, K. J. Sodium secretion as part of the formation of gastric juice. Uppsala J. Med. Sci. 80: 65-70, 1975. 28. RAY, T. K., AND L. L. TAGUE. Role of K+-stimulated ATPase in H+ and K+ transport by bullfrog gastric mucosa in vitro. Acta Physiol. Stand., Suppl.: Gastric Ion Transport, 1978, 283-292. 29. RUSSEL, J. C., AND K. KOWALEWSKI. The acid-base balance in gastric H+ secretion. Physiol. Chem. Physics 6: 269-274, 1974. 30. SANDERS, S. S., V. B. HAYNE, AND W. S. REHM. Normal H+ rates in frog stomach in absence of exogenous CO, and a note on pH stat method. Am. J. PhysioZ. 225: 1311-1321, 1973. 31. TEORELL, T. Electrolyte diffusion in relation to the acidity regulation of the gastric juice. GastroenteroZogy 9:. 425-443, 1947. 32. WERTHER, J. L., F. HOLLANDER, AND M. ALTAMIRANO. Effect of acetazolamide on gastric mucosa in canine vivo-vitro preparations. Am. J. Physiol. 209: 127-133, 1965.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (150.216.068.200) on January 11, 2019.
