- Email: [email protected]
Multiple subcutaneous injections of somatostatin induce tachyphylaxis of the suppression of plasma insulin, but not glucagon, in the rat
Regulatory Peptides, 4 (1982) 333-339
333
Elsevier BiomedicalPress
Multiple subcutaneous injections of somatostatin induce tachyphylaxis of the suppression of plasma insulin, but not glucagon, in the rat F. M~irki, U . M . B u c h e r a n d J . - C . R i c h t e r Research Laboratories, Pharmaceuticals Division, CIBA-GEIGY Ltd., Basle, Switzerland
(Received31 August 1982; acceptedfor publication I September1982)
Summary The time course of pancreatic effects of somatostatin was studied over a period of 2 h in unanesthetized unrestrained rats after administration of the peptide by intravenous infusion and by single and multiple subcutaneous injections. During infusion of 10 and 30 ~ g / k g per min, somatostatin continuously suppressed plasma insulin and plasma glucagon. Plasma glucose was significantly increased at the lower dose, but not affected at the higher dose. Single subcutaneous injections of 0.3 and 3 m g / k g decreased plasma insulin and glucagon dose-dependently for 20-60 rain without affecting plasma glucose. Multiple subcutaneous injections of somatostatin (one to four doses of 3 mg/kg, administered at intervals of 30 min) caused an initial decrease of plasma insulin (at 30 min), a rebound-increase at 60 and 90 min, and a final return to control values by 120 min. Plasma glucagon remained continuously suppressed. Plasma glucose increased significantly at 60 and 90 min and tended to return towards control values thereafter. In conclusion, pancreatic B cells - but not A cells - of the rat develop tachyphylaxis to somatostatin within 2 h after multiple subcutaneous injections of the peptide. By this mode of administration, 'selective' suppression of plasma glucagon can be achieved with somatostatin in the rat. glucagon selectivity; intravenous infusion; plasma glucose
Address correspondence to: Dr. F. M~irki, CIBA-GEIGY Ltd., K-125.410, CH-4002 Basle, Switzerland;
Telephone (061) 362725. 0167-0115/82/0000-0000/$02.75 © 1982 ElsevierBiomedicalPress
334 Introduction
While somatostatin - a tetradecapeptide inhibiting the secretion of a variety of hormones [2] - is extensively utilized as a valuable research tool, therapeutic application of the peptide, e.g., for the treatment of diabetes [1], is curtailed by its short biological half-life [8]. Chemical analogues with a long(er) duration of action may, therefore, eventually provide a practicable alternative to the current mode of administration of somatostatin, i.e., continuous intravenous infusion. In the course of a research program to develop such long-acting analogues we explored several modifications of our routine test model for somatostatins, the unanesthetized rat with basal hormone release [6]. We studied, for instance, whether it is possible to mimic the time course of pancreatic effects of a long-acting analogue by a series of multiple injections of (short-acting) somatostatin, using the subcutaneous route. The results of these experiments are reported here.
Methods
The study was performed in male rats (Tif: RAI f (SPF)) from Tierfarm Sisseln A G with body weights of 350-450 g (intravenous infusion) and 170-200 g (subcutaneous administration). The animals were kept at 22 + 2°C with a 12 h light/12 h dark cycle and fed with Nafag 890 rat chow (Nafag, 9202 Gossau, Switzerland) and tap water ad libitum. For the infusion experiments rats were prepared with a chronic arterial and venous infusion catheter (feed-through-swivel-system ZABONA®). Both catheters consisted of two polyethylene tubes (PE 10 and PE 20, Clay-Adams) of the appropriate length that had been joined together in a current of hot air. The catheters were implanted under pentobarbitone anesthesia, filled with sterile physiological saline and closed peripherally with a steel pin. The arterial catheter was introduced into the left femoral artery and advanced to the descending limb of the aorta, the venous catheter was placed in the vena cava inferior via the femoral vein. The subcutaneous sections of both catheters were brought to the nape of the neck, where they passed out through the skin. A healing period of 7-10 days was allowed before commencement of the experiments. During the experiments the rats remained unanesthetized and unrestrained, and could freely move in the cage (250 cm2). After a control period of 30 min saline or somatostatin was administered intravenously during 120 min (test period) by means of an infusion pump (Braun-Melsungen, infusion rate 2.25 ml/h). Arterial blood samples (1 ml each) were withdrawn at times zero, 30, 60 and 120 rain for the assay of plasma insulin, glucagon and glucose concentrations. To exclude any hypovolemic counterregulation after blood withdrawal, an equal volume of arterial blood from an untreated donor rat was injected immediately after each withdrawal. The protocol for experiments with single subcutaneous injection has been described previously [6]. In the experiments with multiple subcutaneous injections separate groups of eight rats received a total of four injections each, administered at
335 intervals of 30 min. Animals were killed 30 min after the last injection (120 min after first injection). Treatment was as follows: group A received three injections of saline (1 ml/kg) followed by one injection of somatostatin (3 m g / k g per ml); group B, two injections of saline followed by two injections of somatostatin; group C, one injection of saline followed by three injections of somatostatin; group D, four injections of somatostatin; control group, four injections of saline. Blood sample treatment and assay methods for plasma insulin, glucagon and glucose have been described previously [6]. Statistical significance was evaluated using Student's t-test.
Results and Discussion
This study compares the time course of effects of single and multiple subcutaneous injections of somatostatin with that of a constant intravenous infusion, measuring plasma levels of insulin, glucagon and glucose in unanesthetized, unrestrained rats. Results obtained in the infusion experiments with saline (control) and l0 and 30 ~ g / k g per min of somatostatin, administered over a period of 2 h, are shown in Fig. 1. The absence of significant changes of all three variables during saline infusion indicates that the experimental procedure did not impose a measurable stimulatory or stress effect on the unanesthetized animals. Infusion of somatostatin caused a marked, significant decrease of plasma insulin throughout the experiment. Since the two doses administered are relatively high, and since both produced identical effects, it may be assumed that maximal hormone suppression was achieved. The decrease of plasma glucagon was relatively smaller than that of plasma insulin. A similar observation has been made previously when somatostatin was administered to rats by single subcutaneous or intravenous injection [6]. Both doses produced practically the same, presumably maximal, glucagon suppression. A considerably lower dose of somatostatin (0.7 ~ g / k g per min) had been administered in the study of Lins et al. [5]; while still able to suppress plasma insulin, this dose had failed to significantly decrease plasma glucagon during a 2-h infusion. Plasma glucose was significantly increased by somatostatin at l0 p~g/kg per rain, whereas virtually no change occurred at 30 ~ g / k g per min. This result was not expected in view of nearly identical changes of plasma insulin and glucagon, respectively, after the two different doses of somatostatin. However, levels of pancreatic hormones in the portal vein may be more important determinants of the glycemic response than peripheral levels, as measured in our experiment. Studies in dogs indicate indeed that portal concentrations of insulin and glucagon may change markedly under somatostatin infusion, while little or no change of suppressed hormone levels is detected in peripheral plasma [7]. The dose and the injection interval for the experiment with multiple subcutaneous injections of somatostatin were selected on the basis of an experiment with a single injection. As shown in Fig. 2, single somatostatin doses of 0.3 and 3 m g / k g suppressed plasma insulin for about 20-30 min and up to about 45 min, respec-
336
8OO
~ ~ .~
INSULIN
INSULIN
120
6O0
O Z 0.
4OO
.
.
.
.
• • D_
-
80 I
2OO
s ~1S I
s
40 O,
I
"1"
I
0
0.5
I
I
I
(HOURS)
2
i
I i 5 10
0
I 20
I 30
I I 45 (MIN) 60
GLUCAGON 120 ..I
a. 20
•
eo
I
I
I
0
0.5
1
0
100
n,
0
I
(HOURS)
I
2
Dz 0 (..)
80 6O
I
o T" 0
10
LU
J
,
i' 5
'r 10
'i 20
120 ~
8"
I 30
I 45
(MIN)
I 60
GLUCOSE
S.
. oo,,~°.O -'~..
100
8O
0"1"1 o~
gGON
GLUCOSE
O Z
,-
I
I
I
i
,,~
0
OS
1
(HOURS)
2
0
5
I
I
I
10
20
:30
I 45 TIME (MIN)
I 60
Fig. 1. Effect of infusion of somatostatin on plasma insulin, glucagon and glucose in unanesthetized, unrestrained rats. Key: O, control, saline 2.25 m l / h (n = 4); O, somatostatin l0 p.g/kg per min (n = 3); I , somatostatin 30 ~ g / k g per min (n = 6). Mean_+ S.E.M. Statistical significance of difference against control: • P < 0.05, 0 0 P < 0.01. Fig. 2. Effect of a single subcutaneous injection of somatostatin on plasma insulin, glucagon and glucose in unanesthetized rats. The number of animals used at each time point is indicated by the symbols I (n =8), II (n = 16) and III (n = 24). The dose of 0.3 m g / k g is represented by ©, and 3 m g / k g by O. Results are expressed as percent of control values (equal number of rats injected with saline).
tively, with a maximal effect around 5-10 min after administration. Plasma glucagon showed a relatively smaller, but somewhat longer-lasting decrease, whereas plasma glucose was not significantly affected. It was therefore anticipated that multiple subcutaneous injections of 3 m g / k g somatostatin, administered at intervals of 30 min, might cause continuous suppression of insulin as well as glucagon, comparable to the effect of the intravenous infusion (see Fig. 1). This was, in fact, observed with respect to plasma glucagon; 30 min after one injection, and 30 min after the last of two, three or four injections of somatostatin (i.e., up to 2 h since the first injection), plasma glucagon was decreased to a similar extent, i.e., by about 40% (Fig. 3).
337
Plasma insulin, however, showed a paradoxical response. It was decreased 30 min after a single injection of somatostatin (Fig. 3A), but increased by 40-50% above control values after two and three injections (60 and 90 min after the first somatostatin injection; Fig. 3B, C). Finally, control values were reached after four injections (120 rain after the first somatostatin injection; Fig. 3D). Thus, after the initial suppression which was followed by a phase of a rebound-increase, complete tachyphylaxis occurred. This result indicates that the response of the pancreatic B cell to somatostatin is decisively influenced by the route of administration of the peptide. Within the time period studied, intravenous infusion leads to continued suppression, whereas multiple subcutaneous injection results in a biphasic response followed by tachyphylaxis. By contrast, glucagon secretion from the A cell remains continuously suppressed irrespective of the route of somatostatin administration. A possibly related phenomenon has been observed in a study by Lien and Garsky
+40 INSULIN +20
0
i
-20
-40 (Z
O O
GLUCAGON
P" Z
0
L
<
~-~ z z 0
-40 ee
+20
ee
ee
GLUCOSE
1 A I
B I
C I
D I
30
60
90
120
TIME SINCE FIRST S S INJECTION (MIN)
Fig. 3. Changes of plasma insulin, glucagon and glucose after multiple subcutaneous injections of somatostatin (SS) in unanesthetized rats. One (A), two (B), three (C) or four (D) injections of SS (3 m g / k g ) were administered at 30 min intervals. Animals were killed 30 min after the last injection. Mean changes against control + S.E.M. (n = 8). Statistical significance: • P < 0.05, ~ P < 0.01.
338
[4]. The authors reported that (single) subcutaneous and intravenous injection of a somatostatin analogue suppressed plasma insulin, but not glucagon, whereas intravenous infusion decreased insulin as well as glucagon. They concluded that the route of administration is a determining factor for the selectivity of pancreatic hormone suppression in the rat. Comparable studies have so far not been reported in man, because the effect of a single injection of somatostatin in this species is too short-lived [3]. However, in view of the therapeutic potential of selective glucagon suppression in the treatment of diabetes [1], it should be of interest to explore the effect of a long-acting somatostatin analogue after multiple subcutaneous injections in man. On the basis of our results in the rat it appears feasible that this mode of administration might confer 'selective' glucagon suppression, if and as long as tachyphylaxis of human B cells to this analogue occurs. Plasma glucose was not affected after one injection of somatostatin (Fig. 3A), but significantly increased after two and three injections (Fig. 3B, C). A smaller, no longer significant increase after the fourth injection (Fig. 3D) could indicate that hyperglycemia may be temporary and may level off with continued treatment. In conclusion, this study demonstrates a radically different response of pancreatic B cells of the rat to somatostatin, when the peptide is administered by different routes. Tachyphylaxis develops within two hours after multiple subcutaneous injections, whereas full suppression of insulin release is maintained during intravenous infusion. By contrast, A cell secretion remains suppressed during administration of somatostatin by either route. Consequently, 'selective' suppression of glucagon release can be achieved in the rat by multiple subcutaneous injections of (non-selective) somatostatin.
Acknowledgements We thank Jutta Motz, Domenico Ammaturo and Roland Frutiger for technical assistance and Jeannine Schweitzer for typing the manuscript.
References 1 Gerich, J.E., Schultz, T.A., Lewis, S.B. and Karam, J.H., Clinical evaluation of somatostatin as a potential adjunct to insulin in the management of diabetes mellitus, Diabetologia, 13 (1977) 537-5,14. 2 Hansen, A.P. and Lundbaek, K., Somatostatin: a review of its effects especially in human beings, Diab~te M~tab. (Paris), 2 (1976) 203-218. 3 Leblanc, H. and Yen, S.S.C., Comparison of cyclic and linear forms of somatostatin in the inhibition of growth hormone, insulin and glucagon secretion, J. Clin. Endocrinol. Metab., 40 (1975) 906-908. 4 Lien, E.L. and Garsky, V.M., Route of administration as a determining factor in the selectivity of hormone suppression of a somatostatin analog in rats, Horm. Metab. Res., 13 (1981) 675-678. 5 Lins, P.E., Petersson, B. and Efendi~:, S., Effects of short-term and prolonged infusions of somatostatin on endocrine pancreas, body weight and food intake in rats, Acta Physiol. Scan&, 110 (1980) 267-275. 6 Miirki, F., Kamber, B., Rink, H. and Sieber, P., Non-selective inhibition of basal glucagon release by [D-Cyst4]-analogues of somatostatin in the rat, J. Endocrinol., 81 (1979) 315-323.
339 7 Polonsky, K., Jaspan, J., Pugh, W., Dhorajiwala, J., Abraham, M., Blix, P. and Moossa, A.R., Insulin and glucagon breakthrough of somatostatin suppression. Importance of portal vein hormone measurements. Diabetes, 30 (1981) 664-669. 8 Sheppard, M., Shapiro, E., Pimstone, B., Kronheim, S., Berelowitz, M. and Gregory, M., Metabolic clearance and plasma half-disappearance time of exogenous somatostatin in man, J. Clin. Endocrinol. Metab., 48 (1979) 50-53.
333
Elsevier BiomedicalPress
Multiple subcutaneous injections of somatostatin induce tachyphylaxis of the suppression of plasma insulin, but not glucagon, in the rat F. M~irki, U . M . B u c h e r a n d J . - C . R i c h t e r Research Laboratories, Pharmaceuticals Division, CIBA-GEIGY Ltd., Basle, Switzerland
(Received31 August 1982; acceptedfor publication I September1982)
Summary The time course of pancreatic effects of somatostatin was studied over a period of 2 h in unanesthetized unrestrained rats after administration of the peptide by intravenous infusion and by single and multiple subcutaneous injections. During infusion of 10 and 30 ~ g / k g per min, somatostatin continuously suppressed plasma insulin and plasma glucagon. Plasma glucose was significantly increased at the lower dose, but not affected at the higher dose. Single subcutaneous injections of 0.3 and 3 m g / k g decreased plasma insulin and glucagon dose-dependently for 20-60 rain without affecting plasma glucose. Multiple subcutaneous injections of somatostatin (one to four doses of 3 mg/kg, administered at intervals of 30 min) caused an initial decrease of plasma insulin (at 30 min), a rebound-increase at 60 and 90 min, and a final return to control values by 120 min. Plasma glucagon remained continuously suppressed. Plasma glucose increased significantly at 60 and 90 min and tended to return towards control values thereafter. In conclusion, pancreatic B cells - but not A cells - of the rat develop tachyphylaxis to somatostatin within 2 h after multiple subcutaneous injections of the peptide. By this mode of administration, 'selective' suppression of plasma glucagon can be achieved with somatostatin in the rat. glucagon selectivity; intravenous infusion; plasma glucose
Address correspondence to: Dr. F. M~irki, CIBA-GEIGY Ltd., K-125.410, CH-4002 Basle, Switzerland;
Telephone (061) 362725. 0167-0115/82/0000-0000/$02.75 © 1982 ElsevierBiomedicalPress
334 Introduction
While somatostatin - a tetradecapeptide inhibiting the secretion of a variety of hormones [2] - is extensively utilized as a valuable research tool, therapeutic application of the peptide, e.g., for the treatment of diabetes [1], is curtailed by its short biological half-life [8]. Chemical analogues with a long(er) duration of action may, therefore, eventually provide a practicable alternative to the current mode of administration of somatostatin, i.e., continuous intravenous infusion. In the course of a research program to develop such long-acting analogues we explored several modifications of our routine test model for somatostatins, the unanesthetized rat with basal hormone release [6]. We studied, for instance, whether it is possible to mimic the time course of pancreatic effects of a long-acting analogue by a series of multiple injections of (short-acting) somatostatin, using the subcutaneous route. The results of these experiments are reported here.
Methods
The study was performed in male rats (Tif: RAI f (SPF)) from Tierfarm Sisseln A G with body weights of 350-450 g (intravenous infusion) and 170-200 g (subcutaneous administration). The animals were kept at 22 + 2°C with a 12 h light/12 h dark cycle and fed with Nafag 890 rat chow (Nafag, 9202 Gossau, Switzerland) and tap water ad libitum. For the infusion experiments rats were prepared with a chronic arterial and venous infusion catheter (feed-through-swivel-system ZABONA®). Both catheters consisted of two polyethylene tubes (PE 10 and PE 20, Clay-Adams) of the appropriate length that had been joined together in a current of hot air. The catheters were implanted under pentobarbitone anesthesia, filled with sterile physiological saline and closed peripherally with a steel pin. The arterial catheter was introduced into the left femoral artery and advanced to the descending limb of the aorta, the venous catheter was placed in the vena cava inferior via the femoral vein. The subcutaneous sections of both catheters were brought to the nape of the neck, where they passed out through the skin. A healing period of 7-10 days was allowed before commencement of the experiments. During the experiments the rats remained unanesthetized and unrestrained, and could freely move in the cage (250 cm2). After a control period of 30 min saline or somatostatin was administered intravenously during 120 min (test period) by means of an infusion pump (Braun-Melsungen, infusion rate 2.25 ml/h). Arterial blood samples (1 ml each) were withdrawn at times zero, 30, 60 and 120 rain for the assay of plasma insulin, glucagon and glucose concentrations. To exclude any hypovolemic counterregulation after blood withdrawal, an equal volume of arterial blood from an untreated donor rat was injected immediately after each withdrawal. The protocol for experiments with single subcutaneous injection has been described previously [6]. In the experiments with multiple subcutaneous injections separate groups of eight rats received a total of four injections each, administered at
335 intervals of 30 min. Animals were killed 30 min after the last injection (120 min after first injection). Treatment was as follows: group A received three injections of saline (1 ml/kg) followed by one injection of somatostatin (3 m g / k g per ml); group B, two injections of saline followed by two injections of somatostatin; group C, one injection of saline followed by three injections of somatostatin; group D, four injections of somatostatin; control group, four injections of saline. Blood sample treatment and assay methods for plasma insulin, glucagon and glucose have been described previously [6]. Statistical significance was evaluated using Student's t-test.
Results and Discussion
This study compares the time course of effects of single and multiple subcutaneous injections of somatostatin with that of a constant intravenous infusion, measuring plasma levels of insulin, glucagon and glucose in unanesthetized, unrestrained rats. Results obtained in the infusion experiments with saline (control) and l0 and 30 ~ g / k g per min of somatostatin, administered over a period of 2 h, are shown in Fig. 1. The absence of significant changes of all three variables during saline infusion indicates that the experimental procedure did not impose a measurable stimulatory or stress effect on the unanesthetized animals. Infusion of somatostatin caused a marked, significant decrease of plasma insulin throughout the experiment. Since the two doses administered are relatively high, and since both produced identical effects, it may be assumed that maximal hormone suppression was achieved. The decrease of plasma glucagon was relatively smaller than that of plasma insulin. A similar observation has been made previously when somatostatin was administered to rats by single subcutaneous or intravenous injection [6]. Both doses produced practically the same, presumably maximal, glucagon suppression. A considerably lower dose of somatostatin (0.7 ~ g / k g per min) had been administered in the study of Lins et al. [5]; while still able to suppress plasma insulin, this dose had failed to significantly decrease plasma glucagon during a 2-h infusion. Plasma glucose was significantly increased by somatostatin at l0 p~g/kg per rain, whereas virtually no change occurred at 30 ~ g / k g per min. This result was not expected in view of nearly identical changes of plasma insulin and glucagon, respectively, after the two different doses of somatostatin. However, levels of pancreatic hormones in the portal vein may be more important determinants of the glycemic response than peripheral levels, as measured in our experiment. Studies in dogs indicate indeed that portal concentrations of insulin and glucagon may change markedly under somatostatin infusion, while little or no change of suppressed hormone levels is detected in peripheral plasma [7]. The dose and the injection interval for the experiment with multiple subcutaneous injections of somatostatin were selected on the basis of an experiment with a single injection. As shown in Fig. 2, single somatostatin doses of 0.3 and 3 m g / k g suppressed plasma insulin for about 20-30 min and up to about 45 min, respec-
336
8OO
~ ~ .~
INSULIN
INSULIN
120
6O0
O Z 0.
4OO
.
.
.
.
• • D_
-
80 I
2OO
s ~1S I
s
40 O,
I
"1"
I
0
0.5
I
I
I
(HOURS)
2
i
I i 5 10
0
I 20
I 30
I I 45 (MIN) 60
GLUCAGON 120 ..I
a. 20
•
eo
I
I
I
0
0.5
1
0
100
n,
0
I
(HOURS)
I
2
Dz 0 (..)
80 6O
I
o T" 0
10
LU
J
,
i' 5
'r 10
'i 20
120 ~
8"
I 30
I 45
(MIN)
I 60
GLUCOSE
S.
. oo,,~°.O -'~..
100
8O
0"1"1 o~
gGON
GLUCOSE
O Z
,-
I
I
I
i
,,~
0
OS
1
(HOURS)
2
0
5
I
I
I
10
20
:30
I 45 TIME (MIN)
I 60
Fig. 1. Effect of infusion of somatostatin on plasma insulin, glucagon and glucose in unanesthetized, unrestrained rats. Key: O, control, saline 2.25 m l / h (n = 4); O, somatostatin l0 p.g/kg per min (n = 3); I , somatostatin 30 ~ g / k g per min (n = 6). Mean_+ S.E.M. Statistical significance of difference against control: • P < 0.05, 0 0 P < 0.01. Fig. 2. Effect of a single subcutaneous injection of somatostatin on plasma insulin, glucagon and glucose in unanesthetized rats. The number of animals used at each time point is indicated by the symbols I (n =8), II (n = 16) and III (n = 24). The dose of 0.3 m g / k g is represented by ©, and 3 m g / k g by O. Results are expressed as percent of control values (equal number of rats injected with saline).
tively, with a maximal effect around 5-10 min after administration. Plasma glucagon showed a relatively smaller, but somewhat longer-lasting decrease, whereas plasma glucose was not significantly affected. It was therefore anticipated that multiple subcutaneous injections of 3 m g / k g somatostatin, administered at intervals of 30 min, might cause continuous suppression of insulin as well as glucagon, comparable to the effect of the intravenous infusion (see Fig. 1). This was, in fact, observed with respect to plasma glucagon; 30 min after one injection, and 30 min after the last of two, three or four injections of somatostatin (i.e., up to 2 h since the first injection), plasma glucagon was decreased to a similar extent, i.e., by about 40% (Fig. 3).
337
Plasma insulin, however, showed a paradoxical response. It was decreased 30 min after a single injection of somatostatin (Fig. 3A), but increased by 40-50% above control values after two and three injections (60 and 90 min after the first somatostatin injection; Fig. 3B, C). Finally, control values were reached after four injections (120 rain after the first somatostatin injection; Fig. 3D). Thus, after the initial suppression which was followed by a phase of a rebound-increase, complete tachyphylaxis occurred. This result indicates that the response of the pancreatic B cell to somatostatin is decisively influenced by the route of administration of the peptide. Within the time period studied, intravenous infusion leads to continued suppression, whereas multiple subcutaneous injection results in a biphasic response followed by tachyphylaxis. By contrast, glucagon secretion from the A cell remains continuously suppressed irrespective of the route of somatostatin administration. A possibly related phenomenon has been observed in a study by Lien and Garsky
+40 INSULIN +20
0
i
-20
-40 (Z
O O
GLUCAGON
P" Z
0
L
<
~-~ z z 0
-40 ee
+20
ee
ee
GLUCOSE
1 A I
B I
C I
D I
30
60
90
120
TIME SINCE FIRST S S INJECTION (MIN)
Fig. 3. Changes of plasma insulin, glucagon and glucose after multiple subcutaneous injections of somatostatin (SS) in unanesthetized rats. One (A), two (B), three (C) or four (D) injections of SS (3 m g / k g ) were administered at 30 min intervals. Animals were killed 30 min after the last injection. Mean changes against control + S.E.M. (n = 8). Statistical significance: • P < 0.05, ~ P < 0.01.
338
[4]. The authors reported that (single) subcutaneous and intravenous injection of a somatostatin analogue suppressed plasma insulin, but not glucagon, whereas intravenous infusion decreased insulin as well as glucagon. They concluded that the route of administration is a determining factor for the selectivity of pancreatic hormone suppression in the rat. Comparable studies have so far not been reported in man, because the effect of a single injection of somatostatin in this species is too short-lived [3]. However, in view of the therapeutic potential of selective glucagon suppression in the treatment of diabetes [1], it should be of interest to explore the effect of a long-acting somatostatin analogue after multiple subcutaneous injections in man. On the basis of our results in the rat it appears feasible that this mode of administration might confer 'selective' glucagon suppression, if and as long as tachyphylaxis of human B cells to this analogue occurs. Plasma glucose was not affected after one injection of somatostatin (Fig. 3A), but significantly increased after two and three injections (Fig. 3B, C). A smaller, no longer significant increase after the fourth injection (Fig. 3D) could indicate that hyperglycemia may be temporary and may level off with continued treatment. In conclusion, this study demonstrates a radically different response of pancreatic B cells of the rat to somatostatin, when the peptide is administered by different routes. Tachyphylaxis develops within two hours after multiple subcutaneous injections, whereas full suppression of insulin release is maintained during intravenous infusion. By contrast, A cell secretion remains suppressed during administration of somatostatin by either route. Consequently, 'selective' suppression of glucagon release can be achieved in the rat by multiple subcutaneous injections of (non-selective) somatostatin.
Acknowledgements We thank Jutta Motz, Domenico Ammaturo and Roland Frutiger for technical assistance and Jeannine Schweitzer for typing the manuscript.
References 1 Gerich, J.E., Schultz, T.A., Lewis, S.B. and Karam, J.H., Clinical evaluation of somatostatin as a potential adjunct to insulin in the management of diabetes mellitus, Diabetologia, 13 (1977) 537-5,14. 2 Hansen, A.P. and Lundbaek, K., Somatostatin: a review of its effects especially in human beings, Diab~te M~tab. (Paris), 2 (1976) 203-218. 3 Leblanc, H. and Yen, S.S.C., Comparison of cyclic and linear forms of somatostatin in the inhibition of growth hormone, insulin and glucagon secretion, J. Clin. Endocrinol. Metab., 40 (1975) 906-908. 4 Lien, E.L. and Garsky, V.M., Route of administration as a determining factor in the selectivity of hormone suppression of a somatostatin analog in rats, Horm. Metab. Res., 13 (1981) 675-678. 5 Lins, P.E., Petersson, B. and Efendi~:, S., Effects of short-term and prolonged infusions of somatostatin on endocrine pancreas, body weight and food intake in rats, Acta Physiol. Scan&, 110 (1980) 267-275. 6 Miirki, F., Kamber, B., Rink, H. and Sieber, P., Non-selective inhibition of basal glucagon release by [D-Cyst4]-analogues of somatostatin in the rat, J. Endocrinol., 81 (1979) 315-323.
339 7 Polonsky, K., Jaspan, J., Pugh, W., Dhorajiwala, J., Abraham, M., Blix, P. and Moossa, A.R., Insulin and glucagon breakthrough of somatostatin suppression. Importance of portal vein hormone measurements. Diabetes, 30 (1981) 664-669. 8 Sheppard, M., Shapiro, E., Pimstone, B., Kronheim, S., Berelowitz, M. and Gregory, M., Metabolic clearance and plasma half-disappearance time of exogenous somatostatin in man, J. Clin. Endocrinol. Metab., 48 (1979) 50-53.