Cholinergic Release of Gastrin by Feeding in Cats

GASTROENTEROLOGY 70:742-746, 1976 Copyright @ 1976 by The Williams & Wilkins Co.

Vol. 70, No.5

Printed in U.S.A.

CHOLINERGIC RELEASE OF GASTRIN BY FEEDING IN CATS S.-O. SVENSSON, S. EMAS, MARIANNE DORNER, AND

H.

KAESS

Department of Surgery, Karolinska Hospital, Department of Pharmacology, Karolinska Institute, Stockholm, Sweden, and University Clinic of Medicine, Heidelberg, West Germany

The effect of atropine on acid secretion and serum gastrin level after feeding was studied in 8 cats with denervated fundic pouches (Heidenhain pouch, HP). Serum gastrin concentration was measured by radioimmunoassay. In the control experiments, the mean peak acid response occurred 45 min after feeding and the mean peak gastrin concentration 15 min after feeding. After the 1st postprandial hr, acid secretion declined but the gastrin level remained elevated. Atropine, 0.1 mg per kg intravenously 5 min before feeding, depressed both acid secretion and the increase in serum gastrin concentration during the 1st postpra.ndial hr, In both the control and atropine experiments, acid secretion and serum gastrin level were correlated significantly during the 1st postprandial hr, but no correlation was found later in the experiments. The reduction by atropine of the acid and serum gastrin responses varied among the cats but were correlated significantly. It is concluded that a cholinergic mechanism is implicated in the release of gastrin by feeding in cats during the 1st postprandial hr. Atropine reduces the HP acid response to feeding at least in part by depressing the release of gastrin. In a previous study on cats with vagally denervated fundic pouches (Heidenhain pouch, HP), intravenous atropine markedly reduced the maximal acid response to single feeding but did not alter an equally high peak acid response to pentagastrin. 2 These findings suggested either that endogenously released gastrin was more susceptible than pentagastrin to inhibition by atropine at the oxyntic cells, or that atropine depressed the release of gastrin, or both. To elucidate whether atropine inhibits the acid response to feeding by depressing gastrin release, the effect of atropine on postprandial acid secretion and imm unoreactive serum gastrin level was determined in another set of HP cats. The experiments might also reveal whether the HP acid response to feeding is related to the postprandial serum gastrin level. The postprandial increase of serum gastrin level, measured by radioimmunoassay, has been reported to be enhanced rather than depressed by atropine in man 3 - 5 and unaltered in dogs. 6 , 7 These findings were interpreted as indicating that the mechanism for gastrin release by feeding was noncholinergic or highly resistant to atropine. No study has appeared demonstrating a significant relationship between acid secretion and serum gastrin level after feeding. Received June 16, 1975. Accepted November 18, 1975. This paper was presented at the 6th Meeting of the European Gastro-Club, Erlangen, West Germany, 1974.' The authors are grateful to Mrs. I. Rystedt-Nilsson, Miss G. Ranelycke, and Miss U. Teckentrup for expert help. This study was supported by the Swedish Medical Research Council (Grant 2324), Rut and Richard Julin's Foundation, the Karolinska Institute, Stockholm, Sweden, and Deutsche Forschungsgemeinschaft. 742

Methods Eight adult cats (2.8 to 4.3 kg) were provided with separated vagally denervated fundic pouches (Heidenhain pouch, HP) and gastric fistulas (GF) of the innervated main stomach. The G F was placed just proximal to the pyloric gland area near the greater curvature.' Plastic cannulas' were used for the HP and GF. Secretory studies started, at the earliest, 3 weeks after surgery. Before experiments, food but not water was withheld for 18 hr. In two exper~ments on each cat a single intravenous injection of 100 mg per kg body weight of 2-deoxy-o-glucose (2DG) (Sigma Chemical Co., St. Louis, Mo.) stimulated acid secretion from both the GF and HP. Closing the GF (two experiments on each cat) abolished the acid response to 2DG from the HP by acid inhibition of gastrin release. 9 The oxyntic and pyloric gland area of the stomach was therefore considered vagally innervated and the HP was considered vagally denervated. 9 Preliminary dose-response experiments on the 8 cats confirmed the prevo us finding 2 that the HP acid response to feeding with 25 or 50 g of meat did not differ. In the present study the cats were therefore fed 25 g of meat with the GF closed. The ground meat contained 21 % of protein, 3% of fat, 1 % of ashes, and 75% of water. In 12 experiments on the 8 cats, 0.1 mg per kg body weight of atropine sulfate was injected intravenously 5 min before feeding, and in 10 control experiments atropine was exchanged for saline. Gastric secretion from the HP was collected in 15-min periods, and acid output was determined by titration with 0.01 N NaOH solution with phenolphthalein as indicator. Basal secretion was collected for 30 min; after feeding, secretion was collected for 3 hr. Venous blood samples for gastrin determinations were collected 15 and 5 min before feeding and 15, 30, 45, 60, 90, 120, and 180 min after feeding through a plastic catheter flushed with heparin solution. The blood samples (3.5 ml) were allowed to coagulate, and the serum was separated by centrifugation,

May 1976

743

CHOLINERGIC RELEASE OF GASTRIN BY FEEDING

immediately deep-frozen, and stored at -20°C until assayed for gastrin. Between blood sampling, saline was infused through the catheter by an infusion pump (12.5 ml per hr). Owing to the blood loss, 2 weeks were allowed for recovery between the feeding experiments. The cats remained in excellent health throughout the study. Serum gastrin concentration was determined by radioimmunoassay according to Yalow and Berson,'o using rabbit antiserum to synthetic human gastrin I (lCI Pharmaceuticals, Cheshire, England). The details of the assay technique have been described previously. 11 The sensitivity of the assay technique was 5 pg of gastrin per ml of serum. All gastrin assays were performed in triplicates. The serum samples from the control and atropine experiments on the same cat were determined in the same assay. In assays of human sera, the coefficient of variation within assays (35 triplicate assays) was 3%. " The peak acid output and the peak serum gastrin concentration in each experiment is the highest value during the 1st hr after feeding. To reduce the differences between cats in acid output and serum gastrin concentration, respectively, the acid responses and gastrin concentrations in both the control and atropine experiments are expressed in percentage of the mean peak control values for each cat. The mean curves for percentage acid output and gastrin concentration in the 8 cats are calculated from the percentage mean values of each cat. Differences between peak values or percentage values for acid output or gastrin concentration in the control and atropine experiments were evaluated by the paired t-test (two-tailed). 12 In cats with two control or atropine experiments, the mean values were used in the analysis. To reveal any relationship between acid output and mean gastrin level with time during the initial and concluding four 15-min periods after feeding, the percentage values in the control and atropine experiments were analyzed by the linear (r) and rank correlation coefficients. 12 In cats with two control or atropine experiments the mean values were used. The mean gastrin level during a 15-min period is the mean of the gastrin concentration at the onset and end of the period. For the statistical analysis this value was then expressed in percentage of the peak gastrin concentration in the control experiments. The same statistical methods were applied to analyze the relationship between the effect of atropine on the peak acid output and peak gastrin concentration, and on the maximal increase of acid output and gastrin concentration (peak minus mean basal values), expressed in percentage of corresponding control values .

I FEEDING ATRo.PINE 0.1 mg / kg i.V. (. ) OR SALINE i.v. (0)

I

10.0. ~

.£ ~

80.

::l Cl.

60.

::l

40.

~

0 0

~

20.

...J

« Ul

0.

2

3

Ho.URS

~

1. Mean acid response (n = 8) to 25 g of meat in 10 control (open circles) and 12 atropine experiments (filled circles) on 8 HP cats (GF closed). Acid output is expressed in percentage of the peak acid output in the controls for each cat. Mean 100% value: 0.15 mEq per 15 min . Vertical lines = standard error of mean. The degree of significance of difference between control and atropine experiments is indicated as follows: no plus, P > 0.05; + ,0.01 < P < 0.05; + +, 0.001 < P < 0.01; +++ , P < 0.001. FIG.

0.30

o

FEEDING

•

ATROPINE FEEDING

240 220

200

0.25

180 c

'e

160

0.20

!!?

.....

~

++

120

0.15

E

100

~

l( ~

5

80

0.10

o

~

a 0.05

~

140

60

E

.....

gI

a.

z 0:

~ (J)


+++

40 20

~


~

Results 0,--,-Acid secretion. In the control experiments, feeding ACID GASTRIN stimulated acid secretion from the HPwithin 15 min FIG. 2. Mean peak acid output and mean peak gastrin concentra(fig. 1), and the peak acid output occurred in the second tion (n = 8) during the 1st postprandial hr in control and atropine to fourth 15-min periods. The mean peak acid output for experiments on 8 HP cats . Number of experiments and symbols as in the 8 cats, the 100% value, was 0.15 ± 0.02 (SEM) mEq legend to figure 1. per 15 min (fig. 2). After the 1st postprandial hr, the mean curve for acid output decreased to about 50% and cats; in 2 cats atropine almost eliminated the acid then remained around 40 % throughout the experiments response, in 4 cats the peak acid output was reduced by (fig. 1) . 75 to 60%, and in 2 cats by 35 and 15%. Serum gastrin concentration. In the control experiThe acid response to feeding was significantly lower in the atropine than in the control experiments (fig. 1), but ments, the mean basal gastrin concentration was 81 :±: 15 the time for peak output was unchanged. The mean peak pg per ml. The peak serum gastrin concentration ocacid output was significantly (P < 0.001) reduced, by curred in the first to fourth 15-min periods after feeding 60% or to 0.06 ± 0.02 mEq per 15 min (fig. 2). The and the concentration then remained around 80% of the reduction of acid secretion by atropine varied among the peak concentration throughout the experiments (fig. 3).

744

SVENSSON ET AL.

I

feeding (fig. 6, table 1:4, P < 0.05). According to the linear regression lines (figs. 5 and 6), the HP acid response to feeding is completely inhibited, when the peak gastrin concentration is reduced by about 50% (fig. 5), or the maximal increase of gastrin concentration by about 80% (fig. 6).

FEEDING

ATROPINE OJ mg / kg i.v. (. ) OR SALINE iv ( 0 )

I

100 ~

S 80

o

~ 60 u ~ 40 cr f-

~ 20

(:J

...J


0

2

Vol . 70, No.5

3

HOURS

(f)

~ FIG. 3. Mean serum gastrin concentration (n = 8) after feeding of 25 g of meat in control and atropine experiments on 8 HP cats. Gastrin concentration is expressed in percentage of the peak gastrin concentration in the controls for each cat. Mean 100% value: 201 pg of gastrin per ml of serum. Further explanations in legend to figure l.

The mean peak gastrin concentration, the 100% value, was 201 ± 45 pg per ml (fig. 2). Basal gastrin concentration accounted for 40% of peak concentration . The maximal increase of gastrin concentration above basal after feeding varied among the cats and ranged from 51 to 414 pg per ml (median 78 pg per ml). Before atropine was given, the mean basal serum gastrin concentration was 78 ± 15 pg per ml, which did not differ from that in the controls. The gastrin concentration was significantly lower during the 1st hr after feeding in the atropine than the control experiments (fig. 3). The maximal increase of gastrin concentration above basal during the 1st postprandial hr was reduced by 54%. The increase was, however, significant (P < 0.01) also in the atropine experiments and ranged in the 8 cats from 4 to 164 pg per ml (median 40 pg per ml). One hour after feeding, the concentration gradually increased further and did not differ from that in the control experiments (fig. 3). The mean peak gastrin concentration during the 1st postprandial hr was 35% lower in the atropine than in the control experiments (P < 0.01), or 133 ± 29 pg per ml (fig. 2). The reduction of peak gastrin concentration varied among the cats; in 6 cats it was 30 to 50% and in 2 cats less than 20%. Acid secretion and serum gastrin concentration. During the initial four 15-min periods after feeding, there was a significant (P < 0.001) positive correlation between acid output and mean serum gastrin level in both the control and atropine experiments (fig. 4, table 1:1). A similar relationship (P < 0.01 and P < 0.001) was obtained, when corrections were made for basal acid secretion and basal gastrin level (table 1:2). During the last six 15-min periods of the experiments, no significant correlations were found , however, between acid secretion and gastrin level. The effect of atropine on peak acid output and peak gastrin concentration was correlated significantly in the 8 cats (fig. 5, table 1:3, P < 0.001), as well as the effect on maximal increase of acid secretion and gastrin concentration (peak minus mean basal values) after

Discussion The present study on cats confirms previous findings in the healthy man 3. 5.13,14 and in dogs 6 , 7,1 5 that single feeding increased the immunoreactive serum gastrin level. Acid secretion and serum gastrin level have been recorded simultaneously after feeding in man 13 and in dogs with vagally innervated 6 , 7 or denervated fundic pouches,15 but no relationship was reported between postprandial serum gastrin level and acid secretion. Different gastrin components have been revealed in serum. 16, 17 Using the nomenclature of Yalow and Berson,18 the heptadecapeptide gastrin (HG) and big gastrin (BG) are released by feeding l8 whereas the big big gastrin (BBG )-the main gastrin component in fasting serum in man, dog, and hog-appears not to be released by feeding. 19 The half-time disappearance rates for HG and BG are in the order of minutes, and for BBG around 1.5 hr.16. 20 At equal serum gastrin levels, the HG is about 5 times more potent in stimulating acid secretion in dogs than is BG.20 The potency of BBG has not yet been determined. Feeding probably also releases different gastrins with different disappearance rates and different potencies to stimulate acid secretion in cats. Despite this, high positive correlation was found between serum gastrin level and acid secretion from the HP during the initial hour. During the 2nd and 3rd postprandial hr, acid secretion declined while the gastrin level remained elevated, and hence no correlation was found. The absent correlation between acid secretion and gastrin level after the 1st postprandial hr could suggest an increase of immunoreactive gastrins with low potencies to stimulate acid secretion, or a release of an ~

~ (:J

z

5

t!l100

u.

ATROPINE r -0.806

CONTROL r =0.735

ffi 80 I-

~ 60 I-

..

:::)

~ 40 :::)

~ 20

~ 20

40 60

80

KX)

20

80

100

GASTRIN CONC. AFTER FEEDING IN %

FiG. 4. Relationship between 15-min acid output and mean serum gastrin level during the initial four 15-min periods after feeding in control and atropine experiments on 8 HP cats, expressed in percentage of the peak control value for each cat. Mean 100% values as in figures I and 3. Four I5-min periods with no acid response in the control, and five in the atropine experiments, are excluded. Control experiments (left): n = 28, P < 0.001; atropine experiments (right): n = 27, P < 0.001.

745

CHOLINERGIC RELEASE OF GASTRIN BY FEEDING

May 1976

1. Linear correlation coefficients (r) for relationships between acid secretion and serum gastrin level (I and 2) , and between effect of atropine on acid secretion and gastrin concentration (3 and 4) during 1st hr after feeding in 8 HP cats. Analysis based on percentage values (see Methods) . Zero values for acid secretion excluded. Analysis by rank correlation gave similar results

TABLE

Control experiments

Variables"

No.

1. Acid output/15 min vs mean gastrin level/ IS min 2. As 1 minus mean basal values 3. Peak acid output vs peak gastrin concentration 4. As 3 minus mean basal values a

Atropine experiments p

No.

P

28

0.735

<0.D01

27

0.806

<0.D01

27 8

0.528

< 0.01

22 8

0.758 0 .942

< 0.001 <0.001

8

0.759

<0.05

8

Numbers refer to Results section.

a:::-l 100 wo

tg:


80

1-0

60

Su Cl-u.-

6* oz

r z O.942

...

40

~; z 20

~8 ww

Cl-u.-

0 20 40 60 80 100 PEAK GASTRIN AFTER FEEDING IN % OF CONTROL

FIG. 5. Comparison of the effect of atropine on peak acid output and peak gastrin concentration after feeding in 8 HP cats. The output and concentration is expressed in percentage of the peak control values for each cat. Mean 100% values as in figures 1 and 3. Basal gastrin concentration accounts for 40 % of peak gastrin concentration . Number of experiments as in figure 1. The effects of atropine on peak acid output and gastrin concentration were significantly (P < 0.(01) correlated. a::: w Iu.-


I--l =>0 Cl-a::: 1-1=>z 00

Ou

-u.~o u.-* o~

W

100

r= 0 .759

80 60 40 20

Ul(!)


~i5

U W ZW _u.-

0 20

40

60

80

100

INCREASE OF GASTRIN AFTER FEEDING IN % OF CONTROL

FIG . 6. Comparison of the effect of atropine on maximal increase of acid output and of serum gastrin concentration (peak minus mean basal values) after feeding in 8 HP cats. The increases are expressed in percentage of the maximal increases in the control experiments on each cat. Mean 100 % value for the increase of acid output: 0.15 mEq per 15 min ; of gastrin concentration: 120 pg of gastrin per ml of serum. Number of experiments as in figure 1. The effects of atropine were significantly (P < 0 0. 5) correlated.

inhibitor that antagonizes the action of the released gastrin at the oxyntic cells. Physiological amounts of acid in the intact duodenum do not inhibit acid secretion from the HP in cats stimulated by exogenous gas-

trin,21 but the effect of duodenal acidification on acid secretion stimulated by endogenous gastrin has not been reported. Atropine markedly inhibited both the HP acid secretion and the increase of serum gastrin concentration during the 1st hr after feeding. The inhibition varied among the cats, but the inhibition of peak acid output and peak gastrin concentration was highly correlated (fig_ 5). It has been reported recently by the authors that atropine inhibited the acid response to feeding in cats but did not alter a comparable high acid response ' to pentagastrin,2 suggesting that endogenous gastrin was more susceptible than pentagastrin to inhibition by atropine at the oxyntic cells, or that atropine inhibited the release of gastrin, or both_ The high positive correlation between the inhibition of acid secretion and gastrin increase by atropine during the 1st postprandial hr in the present study supports the suggestion that atropine inhibits HP acid response to feeding at least in part by depressing the release of gastrin_ Atropine .might affect gastrin release by inhibition of vagal release of gastrin and by removing cholinergic tone which reduces the sensitivity of the gastrin cells to local chemicaP2, 23 and mechanicaP' stimulation by food. In any case, the effect of atropine shows that a cholinergic mechanism is implicated in the release of gastrin by feeding in cats. Mter the 1st postprandial hr, the gastrin level increased in the atropine experiments to the level in the control experiments, but acid secretion remained low. The late increase in serum gastrin level could suggest that atropine no longer was effective to depress gastrin release, although it reduced acid secretion, implying that gastrin release was more resistant than acid secretion to inhibition by atropine, or that the increase in gastrin level represented noncholinergic release of gastrin with low potency to stimulate acid secretion. In dogs with vagally innervated fundic pouches, atropine inhibited the acid response to feeding but did not prevent the increase in serum gastrin concentration .6, 7 It was concluded that gastrin release by feeding is noncholinergic or highly resistant to atropine. However, in both studies, each on 2 dogs, the postprandial increase in gastrin level appeared later in the atropine than in the control experiments, but the differences were not discussed. In the present experiments, atropine significantly depressed the increase in serum gastrin level only during the 1st postprandial hr. Therefore, the time for

746

SVENSSON ET AL.

the atropine injection might be of importance. In the dog experiments, atropine (0.10 mg per kg) was injected intravenously 45 min before 6 or subcutaneously (0.15 mg per kg) 20 min before feeding 7 and in the present experiments it was injected intravenously 5 min before feeding. The discrepancy between their findings and the authors' might, however, also represent a species difference. , In man, atropine in a total dose of 0.6 mg 3 , 4 or 1 mg 5 given intramuscularly 30 min before feeding enhanced the postprandial increase in serum gastrin level. Acid secretion was not recorded in these experiments. Two plausible explanations can be given for the discrepancy between the findings in man and cat. First, acid secretion may be more susceptible than gastrin release to the inhibitory action of the doses of atropine employed in man, leading to a rise in gastric pH which facilitates cholinergic and/or noncholinergic release of gastrin.25 Second, low doses of atropine might block a cholinergic inhibitory mechanism for gastrin release in man. In conclusion, the present study has demonstrated for the first time that a significant positive relationship exists between serum gastrin level, measured by radioimmunoassay, and acid secretion after feeding, and that a cholinergic mechanism is implicated in gastrin release during the 1st hr after feeding in cats. The discrepancy between these findings in cats and previous findings in man and dogs might be explained by different experimental designs and/or different doses of atropine. Further studies are needed to reveal whether species differences exist in the effect of atropine on gastrin release after feeding in cats, dogs, and man. REFERENCES 1. Emas S, Svensson S-O, Dorner M , et al : Inhibition of acid and gastrin response to feeding by atropine in cats. Acta Hepatogastroenterol (Stuttg) 22: 59-60, 1975 2. Svensson S-O, Emas S: Acid response to graded doses of pentagastrin and feeding, and effect of atropine in Heidenhain pouch cats. Scand J Gastroenterol 9:281-286, 1974 3. Korman MG, Soveny C, Hansky J: Effect of food on serum gastrin evaluated by radioimmunoassay. Gut 12:619- 624, 1971 4. Korman MG, Soveny C, Hansky J: Serum gastrin in duodenal ulcer. Part I. Basal levels and effect of food and atropine. Gut 12:899- 902, 1971 5. Walsh JH, Yalow RS, Berson SA: The effect of atropine on plasma gastrin response to feeding. Gastroenterology 60:16-21, 1971 6. Nilsson G, Simon J , Yalow RS, et al: Plasma gastrin and gastric acid responses to sham feeding and feeding in dogs. Gastroenterology 63:51-59, 1972 7. Csendes A, Walsh JH, Grossman MI: Effects of atropine and of

Vol.70,No.5

antral acidification on gastrin release and acid secretion in response to insulin and feeding in dogs. Gastroenterology 63:257-263, 1972 8. Emas S: Gastric secretory responses to repeated intravenous infusions of histamine and gastrin in non anesthetized and anesthetized gastric fistula cats. Gastroenterology 39:771-782, 1960 9. Eisenberg MM, Emas GS, GroEsman MI: Comparison of the effect of 2-deoxY-D-glucose and insulin on gastric acid secretion in dogs . . Surgery 60:111-117, 1966 10. Yalow RS, Berson SA: Radioimmunoassay of gastrin . Gastroenterology 58:1-14, 1970 11. Emas S, Svensson S-O, Dorner M, et al: Acid secretion and serum gastrin following insulin and 2-deoxY-D-glucose in duodenal ulcer patients. Scand J Gastroenterol 9:629-637, 1974 12. Snedecor GW, Cochran WG: Statistical Methods, Sixth edition . Ames, Iowa, Iowa State University Press, 1969 13. Fordtran JS, Walsh JH: Gastric acid secretion rate and buffer content of the stomach after eating. Results in normal subjects and in patients with duodenal ulcer. J Clin Invest 52:645-657, 1973 14. Becker HD, Reeder DD, Thompson JC: Effect of truncal vagotomy with pyloroplasty or with antrectomy on food-stimulated gastrin values in patients with duodenal ulcer. Surgery 74:580-586, 1973 15. Walsh JH, Csendes A, Grossman MI: Effect of truncal vagotomy on gastrin release and Heidenhain pouch acid secretion in response to feeding in dogs. Gastroenterology 63:593-600, 1972 16. Straus E , Yalow RS : Studies on the distribution and degradation. of heptadecapeptide, big, and big big gastrin. Gastroenterology 66:936-943, 1974 17. Rehfeld JF: Gastrins in serum. A review of gastrin radioimmunoanalysis and the discovery of gastrin heterogeneity in serum. Scand J Gastroenterol 8:577-583, 1973 18. Yalow RS, Berson SA: Further studies on the nature of immunoreactive gastrin in human plasma. Gastroenterology 60:203- 214, 1971 19. Yalow RS, Wu N: Additional studies on the nature of big big gastrin. Gastroenterology 65:19-27, 1973 20. Walsh JH, Debas HT, Grossman MI: Pure human big gastrin. Immunochemical properties, disappearance half time, and acidstimulating action in dogs. J Clin Invest 54:477-485, 1974 21. Emas S, Svensson S-O, Borg I: Effect of duodenal acidification or exogenous secretin on gastric acid secretion stimulated by histamine, pentagastrin or human gastrin I in conscious cats. Digestion 5:17-30, 1971 22. Emas S, Vagne M , Grossman MI: Heidenhain pouch response to antral stimulation before and after antral denervation. Proc Soc Exp Bioi Med 132:1162-1166, 1969 23. Tepperman BL, Walsh JH, Preshaw RM: Effect of antral denervation on gastrin release by sham feeding and insulin hypoglycemia in dogs. Gastroenterology 63:973-980, 1972 24. Debas HT, Konturek SJ, Walsh JH, et al: Proof of a pyloro-oxyntic reflex for stimulation of acid secretion. Gastroenterology 66:526- 532, 1974 25. Grossman MI: How radioimmunoassay has added to our knowledge about gastrin. Metabolism 22:1033-1037, 1973