Effect of fatty acids on isolated pancreatic islets H. N. JORDAN AND R. W. PHILLIPS Department of PhysiologY c ” and Biophysics, Fort ColLinxs, Colorado 80523
Colorado
JORDAN, H. N., AND R. W. PHILLIPS. Efrect of futtvc ucids OII isokted ovine pancreatic iskts. Am. J. Physiol. 234(2): E16ZE167, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 3(2): E162-El67, 1978. -A procedure was developed for the isolation of intact islets of Langerhans from sheep pancreas. The pancreas was disrupted by syringe injection of Hanks solution followed by collagenase incubation and islet separation by sedimentation. The islets were incubated in varying concentrations of glucose and butyrate. The rate of insulin release was approximately linear while the glucose and butyrate concentrations were increased. In additional studies at 2.5 and 5.0 mM levels of substrate concentration, the stimulation of insulin had the following pattern: octanoate > hexanoate > butyrate, whereas beta-hydroxybutyrate, lactate, acetate, and propionate had only slight stimulatory effects that were not statistically significant. Decanoate did not alter insulin release from isolated islets. These data confirm earlier in vivu reports that fatty acids stimulate pancreatic hormone release in sheep and that the stimulus is related to chain length of the fatty acid through C-8 but that C-10 has no effect. A hypothesis was suggested to explain these results based on chain length, solubility, and plasma membrane alterations.
insulin
release; volatile
fatty acids; glucose; in vitro
DERIVE A LARGE PART oftheir metabolic energy requirements from acetic, propionic, and butyric acids. These volatile fatty acids are produced by rumen fermentation and are absorbed into the portal circulation. In sheep, according to Manns and Boda (21) and Bell et al. (Z), butyric and propionic acids are potent stimulators of insulin release, whereas acetic acid has little effect. Because it has not been shown whether these fatty acids act directly on pancreatic islets we have designed experiments to: a) isolate the islets of Langerhans from acinar tissue; b) determine whether or not fatty acids directly affect these isolated islets through the release of insulin; and c) study the relationship of fatty acid chain length to the quantity of insulin released. SHEEP
MATERIALS
AND
METHODS
To establish the location and relative concentration of islets in sheep pancreata, a gland was totally excised and fixed in Bouin’s solution; various areas were sectioned, stained (7), and examined microscopically. The
ovine
State University,
islets appeared uniformly distributed throughout the pancreas. Sheep approximately 6 mo of age were anesthetized with sodium pentobarbital, and a right laporotomy was performed on each. The pancreas was removed and immediately placed in ice-cold Hanks solution. In addition, Hanks solution was infused with a needle and syringe at various sites in the body of the pancreas until it was grossly distended. This step was found to be essential to allow for the subsequent digestion with collagenase. Excess fat and blood vessels that became clearly visible after infusion of Hanks solution were removed, and the pancreatic tissue was finely chopped with scissors. The technique for isolation of pancreatic islets of the rat as described by Lacy and Kostianovsky (17) was modified for the larger sheep pancreas. The pieces of tissue were transferred to a 50-ml centrifuge tube. Fragments that floated were removed because they consisted predominantly of adipose tissue. The remaining tissue was gently centrifuged for 1 min and the supernatant removed by aspiration The residue was transferrred to 125-ml Erlenmeyer flask containing ZOO mg collagenase (Calbiochem, B grade) in 10 ml of Hanks solution. The flask was securely stoppered and shaken in a Dubnoff metabolic shaker at 35OCfor 25-30 min, Incubation time was determined in preliminary experiments by sampling the contents at 2-min intervals and continuing incubation until the islet tissue appeared to be substantially free of acinar fragments. After incubation, the solution was poured into two 50-ml centrifuge tubes, and about 30 ml of ice-cold Hanks solution was added to each tube to inhibit collagenase action. Islets and fragments of acinar tissue were separated by centrifugation, and the supernatant was discarded. The residue in each tube was resuspended in 20 ml of the same solution and again centrifuged, the washing being rejected. This washing was repeated a 3rd time before suspension of tissue in 10 ml of solution in a black-base petri dish surrounded by crushed ice for islet separation. Tissue fragments were examined with a dissecting microscope at x 7 magnification; under these conditions, the islets were visible as white spheres. The islets were picked up individually with a braking pipette and transferred to a small cold test tube. After completion of this initial separation, the contents of the test tube
El62
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
were pipetted into a second petri dish and further diluted with Hanks solution. A second separation of the islet tissue to remove occasional fragments of acinar tissue was carried out, the islets again being placed in a small test tube in Hanks solution. On examination of the islets under a higher magnification, only those islets free of acinar tissue were transferred into incubation flasks for the appropriate experimental procedures. The total time from pancreatic removal to the isolation of islets ready for experimentation was approximately 120 min. The islets were incubated in groups of 10 in 0.5 ml of modified Krebs bicarbonate solution containing bovine plasma albumin (Sigma Chemical Co.) at 1 mg/ml; pH was maintained at 7.4. Flasks similar to those described by Keen et al. (14) were used. With each set of experiments, 10 control islets were incubated in the medium alone as well as in the test solutions. The flasks were capped, gassed for 2 min with 95% 0, and 5% COZ, and incubated in a Dubnoff metabolic shaker at 72 oscillations/min at 37°C for a period of 90 min. Isolated islets were incubated with glucose and the sodium salts (buffered to pH 7.4 with HCl) of a number of short- and medium-chain fatty acids at 2.5 and 5.0 mM concentrations: acetic, propionic, butyric, hexanoic, octanoic, and decanoic, as well as lactic acid. The shorter-chain fatty acid salts are completely miscible in water and would be free in solution as the minimal albumin concentration would not bind. One of the ketone bodies, beta-
FIG.
1. Morphology
of unstained
islets
separated
El63
ISLETS
from
sheep
hydroxybutyric acid, was also used as an incubant. The incubation fluid was transferred in 100 ~1 aliquots to 10 x 75 mm disposable test tubes and frozen for the subsequent determination of insulin. Insulin was assayed in duplicate by a radioimmunoassay procedure (9) with crystalline bovine insulin as a standard. Glucose was determined by a glucose oxidase method (4). RESULTS
A modification of the sedimentation procedure (17) was used to isolate the islets of Langerhans from sheep pancreata, and with considerable practice a large number could be isolated from a single pancreas. Only islets exceeding 100 ,um in diameter were used for these studies. Under a dissection microscope against a black background, they appeared as a grayish-white spherical or ovoid structure; the appearance of unstained islets is shown in Fig. 1. By light microscopy, the islets appeared intact. The beta cells contained numerous secretory granules, and their appearance was similar to islets observed in sheep pancreata not treated with collagenase. Insulin release by isolated islets influenced by varying concentrations of butyrate and glucose is shown in Fig. 2. Glucose at the 0.5 mg/ml level appeared to have no effect, a result that is not surprising because this concentration is at the lower end of the normal blood
pancreas
compressed
under
a cover
slip.
Magnification
X 160
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
El64
H. N. JORDAN
glucose concentration in sheep. Normal plasma glucose in sheep would be between 0.5 and 1.0 mglml and rarely greater than 0.7 mg/ml. Even under conditions in which glucose was significantly raised above normal glucose levels, the insulin response was minimal. As the glucose level in the test medium was increased, however, insulin release was observed to increase in proportion to the glucose concentration (Fig. 2). Butyrate caused a relatively greater increase in insulin release than glucose (Fig. 2) and indicated that shortchain fatty acids do indeed have a direct effect on sheep islets. To determine the effect of various acids on insulin release, isolated islets were incubated with buffered sodium salts of a number of short- and medium-chain fatty acids in addition to butyrate, as well as lactic acid and beta-hydroxybutyric acid (Figs. 3 and 4). Two
AND
R. W. PHILLIPS
concentrations of each acid were used, and a dose response may be easily noted by comparison between the figures. Neither lactic, acetic, nor beta-hydroxybutyric, nor propionic acids appear to have an effect on insulin release because the results were not greatly different from the control incubation. It is interesting that decanoic acid does not cause insulin release. As the fatty acid chain length increases up to an B-carbon chain length, insulin release is increased. The lo-carbon fatty acid has no apparent effect on the islets. DISCUSSION
In vitro studies of insulin release with the use of either pieces of sheep pancreas or perfused sheep pancreas have been limited. However, methods have recently become available for the isolation of intact islets
500
420
c d 4
2 c:
FIG. 2. Insulin released from isolated islets incubated in vitro in presence of varying concentrations of butyrate and glucose. Numbers within bars represent total number of sheep. Vertical line at top of each bar represents mean k SE. Probability that values are different from control are listed at bottom of appropriate bar.
220
180
H
140
100 60
20 Blk.
1.5mM Butyrate
500
m
480
-
460
-
440
L
420
-
400
-
380
L
360
-
340 320
-
300
er
280
c
260
-
240
+
220
r
200
+
1RO
m
160
-
140 120
c
100 80
e c
60
c
20
-
2.5mM Butyratc PC 0.01
5.olnM Butyrate P-- 0.01
O.Smg/ml
1omM Butyrate PC 0.001
l.Omq/ml
Glucose
Glut
l.Smq/ml
ose
3.0mg/ml GlUCOSC~ Pf’ 0.1
Glucose
FIG. 3. Effect of fatty acid chain length on insulin release from isolated islets incubation in vitro at a concentration of 2.5 mM. Three sheep were used in each group of trials. Vertical line at top of each bar represents mean + SE. Probability that values are different from control are listed at bottom of appropriate bar.
f
t
Blk.
2.5mM
2.5mM 1,s L’ t a t c
Acetate
2.5mM Propionate
2.smM
2.5mM
Hydroxy-
But
yrate
butyratv
P-
0.05
2.5mM Hcxanoatta P* 0.02
2.smM
2.5mM
Oc tanoatt P* 0.01
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
El65
ISLETS
800 7GO 720 680 640 600 560 2 3; *r
480
g 2
440 400
/
520
FIG. 4. Effect of fatty acid chain length on insulin release from isolated islets incubated in vitro at a concentration of 5.0 mM. Three sheep were used for each group of experiments. Vertical line at top of each bar represents mean * SE. Probability that values are different from control are tested at bottom of appropriate bar.
360 374 2 Id2
320 280 240 200 160 120 80
t
40
Blk.
5.OmM Lactate
5.OlllM Acetate
5.omM Hydroxybutyrate
S.OmM Propionate
5.OlTLM Rut
yrate
P’
0.05
of Langerhans from mammalian pancreata in large numbers by the use of collagenase (17). Islets separated by such methods appear to r&pond to factors that regulate insulin synthesis and secretion (12, 17, 24). The proportionate rise in insulin release to glucose concentration agrees with earlier findings (17, 31, 36) that glucose induces an immediate and pronounced release of insulin from isolated rat pancreatic islets. The magnitude of the increase in insulin release from sheep islets with increasing glucose concentration is relatively low when compared to other species. The greater insulin release stimulated here in vitro by butyrate as opposed to glucose confirmed the in vivo results of Phillips et al. (25, 26), Manns et al. (21, ZZ), Bell et al. (Z), and Jones et al. (13). Although in this experiment the concentration of butyrate was greater than the levels seen in peripheral blood of sheep, the above-listed in vivo findings indicate that butyrate at physiological levels will stimulate insulin release (22). The nonstimulatory effect of lactic, acetic, and betahydroxybutyric acids on insulin release reported here is in agreement with studies in which slices of sheep pancreas were used (10, 11). The lack of response to acetic and beta-hydroxybutyric acids would indicate that metabolites of butyric and the longer-chain fatty acids are not involved with insulin release. Malaisse and Malaisse-Lagae (19) felt that metabolism of noncarbohydrate substances such as fatty acids and ketones caused the resultant insulin release. This concept is hard to accept because of the comparative lack of action of beta-hydroxybutyrate, acetate, and decanoate found in these experiments. The lack of effect of longer-chain fatty acids on insulin release in sheep has been previously demonstrated in vivo (1, 26). Sanbar et al. (30) found that octanoic acid affects glucose concentration and utilization in dogs. Subsequently, Sanbar and Martin (31) and Montague and Taylor (24) reported that butyrate and octanoate
5.OmM Hexanoatc P*- 0.01
IL1
5.OmM &tar-mate
5.Oll-M kc
atma
t e
PC 0.001
caused a marked insulin release from rat islets, a result that clearly indicates that the effect of fatty acids is not species-specific, More recently, Pi-Sunyer (27) reported that octanoate did not have a stimulatory effect of insulin release from pancreatic slices. The water-lipid solubility of the fatty acids would seem to be implicated. Unfortunately, this will not clarify all aspects of the results. Acetic, propionic, and butyric acids are completely miscible in water; hexanoic and octanoic acids are fairly soluble, whereas decanoic acid is only slightly soluble. The same variation in biological handling of fatty acids of increasing chain length can be seen if one considers the fate of absorbed fatty acids. Through C-8 most absorbed fatty acids travel directly to the liver via the portal system. Longer-chain fatty acids are incorporated into triglycerides and chylomicrons, enter the lacteals, and travel via lymphatics to enter the vascular system past the liver. This difference in handling has traditionally been related to solubility, Problems become evident when direct comparisons between in vitro and in vivo experiments are made. One such problem in these experiments is the relatively low level of albumin that was added to the incubating system. It would be capable of binding only a minor fraction of the added fatty acid, whereas under normal physiological conditions most of the fatty acid is bound to ablumin (33). However, earlier in vivo experiments with large quantities of short- and medium-chain FFA during periods of time when sheep had a very elevated level of endogenous FFA produced analogous results (25, 26). In those experiments, it would seem unlikely that the injected short-chain fatty acids would have replaced the more tightly bound endogenous long-chain fatty acids from their binding sites (33). Thus, one could assume that both bound and unbound short-chain fatty acids have the same effect on islet function. Other research has shown that short- and mediumchain fatty acids may directly affect tissues. Propionate,
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
El66 butyrate, and hexanoate, but not octanoate or decanoate, alter the morphological and growth characteristics in HeLa cells in tissue culture (6), an action thought to be due to changes in the properties of the plasma membrane. Butyrate has been shown to increase cyclic AMP levels in several different cell culture experiments. The action appears to be an effect of stimulation of adenyl cyclase, not inhibition of phosphodiesterase (28, 32). Cha nges in cyclic AMP have been convincingly linked as coupling factors between cell excitation and hormonal release although insulin secretion may also occur without an increase in cyclic AMP (29). In the experiments reported here, it is possible that butyrate as well as hexanoate and octanoate may act to directly alter cyclic AMP levels in islet cells and cause increased secretion. Phillips et al. (25, 26) had proposed an indirect mechanism for the glucogenic behavior of butyrate involving glucagon release and suggested that the increase in blood glucose after butyrate infusion arises from hepatic glycogenolysis and not from the conversion of butyrate to glucose. At this time, it would seem that fatty acids may cause secretion by both alpha and beta cells giving a simultaneous insulin and glucagon release, which would explain both the observed hyperinsulin and hyperglycemic responses (21, 22, 25, 26). Conversely, the fatty acids may act directly on the alpha cells and the subsequent release of insulin may be caused by increased glucagon as has been noted (8, 18, 34). We find this second hypothesis less attractive as the increase in insulin release due to fatty acids is of such magnitude that it does not seem likely to be caused by glucagon. By and large, glucagon-induced insulin release is not this great (15, 18). Other theories have been suggested; for instance, mammalian beta cells are believed to be freely permeable to glucose (5) and to phosphorylate it at a rate significantly higher than either pancreatic acini or liver (23). Montague and Taylor (24) found that shortchain fatty acids increased the quantity of glucose-6phosphate in the islet. They suggested that glucose-6phosphate may be regulated by fatty acids and in this
H. N.
JORDAN
AND
R. W. PHILLIPS
way influence insulin release. A reasonable assumption is that insulin release can be controlled by factors intimately related to glucose metabolism within the beta cell (16). However, this hypothesis does not preclude the possibility that short- and medium-chain fatty acids may be more effective in initiating insulin release in sheep. Whether this effect is controlled through changes in glycolytic intermediates by a direct effect of the fatty acids on islet cell membranes and morphology or by some other action such as changing CAMP levels is not clear. Another aspect to consider is that glucose levels are normally much higher in blood than are the short- or medium-chain length fatty acids, Whereas some of the glucose levels used in this study were within physiological ranges, the fatty acid levels were all higher than those normally encountered in ovine arterial blood. Normal levels of volatile fatty acids were not used because of their very low concentration in blood. In vivo studies (22) showed that physiological levels of short-chain fatty acids can stimulate insulin release without concomitant hyperglycemia or hyperketonemia. It seems plausible that in sheep, in contrast to man, shortand medium-chain fatty acids play a greater role in the control of insulin and glucagon release than does glucose. Unfortunately, the mechanism by which elevated levels of short- and mediumlength fatty acids stimulate hormone release from islets has not been clarified by this study. With slight modifications the islet isolation procedure described can be adapted for the isolation of pancreatic islets from other large animals to study the control of various substances on insulin and glucagon release. According to Lacy et al. (18) and Buchanan et al. (3), proteolytic destruction of released insulin has not been observed in isolated islets of the rat. Therefore, this technique eliminates some of the difficulties associated with the use of pancreatic slices and perfused pancreata because only islets are present in the incubation flasks. Received
for publication
14 February
1977.
REFERENCES 1. ASH, R. W., R. J. PENNINGTON, AND R. S. REID. The effect of short chain fatty acids on blood glucose concentration in sheep. Biochem. J. 90: 353-360, 1964. 2. BELL, J. P., L. A. SALAMONSE, G. W. HOLLAND, E. A. ESPINER, D. W. BEAVEN, AND D. S. HART. Autotransplantation of the pancreas in sheep: insulin secretion from the transplant. J. Endocrinol. 48: 511~-525, 1970. 3. BUCHANAN, K. D., J. E. VANCE, AND R. H. WILLIAMS. Insulin and glucagon release from isolated islets of Langerhans. Effect of enteric factors. Diabetes 18: 381-386, 1969. 4. CAMPBELL, L. A., AND D. S. KRONFELD. Estimates of low concentrations of plasma glucose using glucose oxidase. Am. J. Vet. Res. 22: 587-589, 1961. 5. COORE, H. G., AND P. J. RANDLE. Regulation of insulin secretion studies with pieces of rabbit pancreas incubated in vitro. Bio&em. J. 93: 66-78, 1964. 6. GINSBERG, E., D. SALOMON, T. SREEVALSAN, AND E. FREESE. Growth inhibition and morphological changes caused by lipophilic acids in mammalian cells. PFOC. Nat/. Acad. Sci. U.S. 70: 2545-2461, 1973. 7. GOMORI, G. Aldehyde fuchsin: a new stain for elastic tissue. A 111. J. C/in. Pathol. 20: 665-666, 1950.
G. M., L. L. BENNETT, D. F. SMITH, AND F. G. SCHMID. Effect of pulse administration of glucose or glucagon on insulin secretion in vitro. Metabolism 16: 222-233, 1967. HALES, C. N., AND P. J. RANDLE. Immunoassay of insulin with insulin antibody precipitate. Biochem. J. 88: 137-146, 1963. HERTENDLY, F., L. MACHLIN, AND D. M. KIPNIS. Further studies on the regulation of insulin and growth hormone secretion in the sheep. Endocrinology 84: 192-199, 1969. HERTENDLY, F., L. MACHLIN, Y. TAKAHASHI, AND D. M. KIPNIS. Insulin release from sheep pancreas in vitro. J. Endocrinol. 41: 605-606, 1968. HOWELL, S. L., AND K. W. TAYLOR. The secretion of newly synthesized insulin in vitro. Biochem. J. 102: 922-927, 1967. JONES, K. L., R. L. BELL, J. M. OYLER, AND D. D. GOETSH. Hyperglycemic effects of sodium butyrate in normal and pancreatectomized sheep. Am. J. Vet. Res. 31: 81412, 1970. KEEN, H., J. B. FIELD, AND 1. H. PASTAN. A simple method for in vitro metabolic studies using small volumes of tissue and medium. Metabolism 12: 143-147, 1963. KETTERER, H., A. M. EISENTRAUT, AND R. H. UNGER. Effect upon insulin secretion of physiologic doses of glucagon administered via the portal vein. Diabetes 16: 283-288, 1967.
8. GRODSKY,
9.
10.
11.
12. 13.
14.
15.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
El67
ISLETS
16. LACY, P. E. Quantitative histochemistry of the islets of Langerhans. I. Lactic, malic glucose-6-phosphate and 6-phosphogluconate dehydrogenase activities in beta cells and acini. Diabetes 11: 96-100, 1962. 17. LACY, P. E., AND M. KOSTIANOVSKY. A method for the isolation of intact islets of Langerhans from rat pancreas. Diabetes 16: 35-39, 1967. 18. LACY, P. E., D. A. YOUNG, AND C. J. FINK. Studies on insulin secretion in vitro from isolated islets of the rat pancreas. Endocrinology 83: 1155-1161, 1968. 19. MALAISSE, W. J., AND F. MALAISSE-LAGAE. Stimulation of insulin secretion by noncarbohydrate metabolites. J. Lab. C&n. Med. 72: 438-448, 1968. 20. MALAISSE, W. J., F. MALAISSE-LAGAE, AND II. MAYHEW. A possible role for the adenyl cyclase system in insulin secretion. J. Clin. Invest. 46: 1724-1734, 1967. 21. MANNS, J. G., AND J. M. BODA. Insulin release by acetate, propionate, butyrate and glucose in lambs and adult sheep. An2. J. Physiol. 212: 747-755, 1969. 22. MANNS, J. G., 3. M. BODA, AND R. F. WILLES. Probable role of propionate and butyrate in control of insulin secretion in sheep. Am. J. Physiol. 212: 756-764, 1967. 23. MATSCHINSKY, F. M., AND J. E. ELLERMAN. Metabolism of glucose in the islets of Langerhans. J. Biol. Chem. 243: 27302736, 1968. 24. MONTAGUE, W., AND K. W. TAYLOR. Regulation of insulin secretion by short chain fatty acids. Nature 217: 853, 1968. 25. PHILLIPS, R. W., A. L. BLACK, AND F. MOLLER. Butyrate induced glycogenolysis in hypoglycemic lambs. Life Sci. 4: 521-525, 1965. 26. PHILLIPS, R. W., W. A. HOUSE, R. A. MILLER, J. L. MOTT, AND D. L. SOOBY. Fatty acid, epinephrine, and glucagon hypergly-
27.
28. 29.
30.
31.
32.
33. 34.
35. 36.
cemia in normal and depancreatized sheep. Am. J. Physiol. 217: 1265-1268, 1969. PI-SUNYER, F. X. Effect of long and medium chain fatty acids on insulin secretion from pieces of rat pancreas. Proc. Sot. Exptl. Biol. Med. 149: 693-697, 1975. PRASAD, K. N., AND P. K. SINHA. Effect of sodium butyrate on mammalian cells in culture: a review. In uitro 12: 125-132, 1976. RASMUSSEN, H., AND D. B. P. GOODMAN. Relationships between calcium and cyclic nucleotides in cell activation. Physiol. Rev. 57: 421-509, 1977. SANBAR, S. S., G. HETENI, JR., N. FORBOTH, AND J. R. EVANS. Effects of infusion of octanoate on glucose concentration and rates of glucose production and utilization in dogs. Metabolism 14: 1311-1323, 1965. SANBAR, S. S., AND J. M. MARTIN. Stimulation by octanoate of insulin release from isolated rat pancreas. Metabolism 16: 482484, 1967. SHEPPARD, J. R., AND K. N. PRASAD. Cyclic AMP levels and the morphological differentiation of mouse neuroblastoma cells. Life Sci. 12 Part II: 431-439, 1973. SPECTOR, A. A. Fatty acid binding to plasma abiumin. J. Lipid Res. 16: 165-179, 1975. SUSSMAN, K. F., AND G. D. VAUGHAN. Insulin release after ACTH, glucagon and adenosine-3-phosphate kyclic AMP) in the perfused isolated rat pancreas. Diabetes 16: 449-454, 1967. UMBREIT, W. W., R. H. BURRIS, AND J- F. STAUFFER. Manometric Techniques (3rd ed.). Minneapolis: Burgess, 1957, pm 338. VANCE, J. D., K. D. BUCHANAN, D. R. CHALLONAR, AND R. H. WILLIAMS. The effect of glucose concentration on insulin and glucagon release from isolated islets of Langerhans of the rat. Diabetes 17: 187-193, 1968.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
Colorado
JORDAN, H. N., AND R. W. PHILLIPS. Efrect of futtvc ucids OII isokted ovine pancreatic iskts. Am. J. Physiol. 234(2): E16ZE167, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 3(2): E162-El67, 1978. -A procedure was developed for the isolation of intact islets of Langerhans from sheep pancreas. The pancreas was disrupted by syringe injection of Hanks solution followed by collagenase incubation and islet separation by sedimentation. The islets were incubated in varying concentrations of glucose and butyrate. The rate of insulin release was approximately linear while the glucose and butyrate concentrations were increased. In additional studies at 2.5 and 5.0 mM levels of substrate concentration, the stimulation of insulin had the following pattern: octanoate > hexanoate > butyrate, whereas beta-hydroxybutyrate, lactate, acetate, and propionate had only slight stimulatory effects that were not statistically significant. Decanoate did not alter insulin release from isolated islets. These data confirm earlier in vivu reports that fatty acids stimulate pancreatic hormone release in sheep and that the stimulus is related to chain length of the fatty acid through C-8 but that C-10 has no effect. A hypothesis was suggested to explain these results based on chain length, solubility, and plasma membrane alterations.
insulin
release; volatile
fatty acids; glucose; in vitro
DERIVE A LARGE PART oftheir metabolic energy requirements from acetic, propionic, and butyric acids. These volatile fatty acids are produced by rumen fermentation and are absorbed into the portal circulation. In sheep, according to Manns and Boda (21) and Bell et al. (Z), butyric and propionic acids are potent stimulators of insulin release, whereas acetic acid has little effect. Because it has not been shown whether these fatty acids act directly on pancreatic islets we have designed experiments to: a) isolate the islets of Langerhans from acinar tissue; b) determine whether or not fatty acids directly affect these isolated islets through the release of insulin; and c) study the relationship of fatty acid chain length to the quantity of insulin released. SHEEP
MATERIALS
AND
METHODS
To establish the location and relative concentration of islets in sheep pancreata, a gland was totally excised and fixed in Bouin’s solution; various areas were sectioned, stained (7), and examined microscopically. The
ovine
State University,
islets appeared uniformly distributed throughout the pancreas. Sheep approximately 6 mo of age were anesthetized with sodium pentobarbital, and a right laporotomy was performed on each. The pancreas was removed and immediately placed in ice-cold Hanks solution. In addition, Hanks solution was infused with a needle and syringe at various sites in the body of the pancreas until it was grossly distended. This step was found to be essential to allow for the subsequent digestion with collagenase. Excess fat and blood vessels that became clearly visible after infusion of Hanks solution were removed, and the pancreatic tissue was finely chopped with scissors. The technique for isolation of pancreatic islets of the rat as described by Lacy and Kostianovsky (17) was modified for the larger sheep pancreas. The pieces of tissue were transferred to a 50-ml centrifuge tube. Fragments that floated were removed because they consisted predominantly of adipose tissue. The remaining tissue was gently centrifuged for 1 min and the supernatant removed by aspiration The residue was transferrred to 125-ml Erlenmeyer flask containing ZOO mg collagenase (Calbiochem, B grade) in 10 ml of Hanks solution. The flask was securely stoppered and shaken in a Dubnoff metabolic shaker at 35OCfor 25-30 min, Incubation time was determined in preliminary experiments by sampling the contents at 2-min intervals and continuing incubation until the islet tissue appeared to be substantially free of acinar fragments. After incubation, the solution was poured into two 50-ml centrifuge tubes, and about 30 ml of ice-cold Hanks solution was added to each tube to inhibit collagenase action. Islets and fragments of acinar tissue were separated by centrifugation, and the supernatant was discarded. The residue in each tube was resuspended in 20 ml of the same solution and again centrifuged, the washing being rejected. This washing was repeated a 3rd time before suspension of tissue in 10 ml of solution in a black-base petri dish surrounded by crushed ice for islet separation. Tissue fragments were examined with a dissecting microscope at x 7 magnification; under these conditions, the islets were visible as white spheres. The islets were picked up individually with a braking pipette and transferred to a small cold test tube. After completion of this initial separation, the contents of the test tube
El62
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
were pipetted into a second petri dish and further diluted with Hanks solution. A second separation of the islet tissue to remove occasional fragments of acinar tissue was carried out, the islets again being placed in a small test tube in Hanks solution. On examination of the islets under a higher magnification, only those islets free of acinar tissue were transferred into incubation flasks for the appropriate experimental procedures. The total time from pancreatic removal to the isolation of islets ready for experimentation was approximately 120 min. The islets were incubated in groups of 10 in 0.5 ml of modified Krebs bicarbonate solution containing bovine plasma albumin (Sigma Chemical Co.) at 1 mg/ml; pH was maintained at 7.4. Flasks similar to those described by Keen et al. (14) were used. With each set of experiments, 10 control islets were incubated in the medium alone as well as in the test solutions. The flasks were capped, gassed for 2 min with 95% 0, and 5% COZ, and incubated in a Dubnoff metabolic shaker at 72 oscillations/min at 37°C for a period of 90 min. Isolated islets were incubated with glucose and the sodium salts (buffered to pH 7.4 with HCl) of a number of short- and medium-chain fatty acids at 2.5 and 5.0 mM concentrations: acetic, propionic, butyric, hexanoic, octanoic, and decanoic, as well as lactic acid. The shorter-chain fatty acid salts are completely miscible in water and would be free in solution as the minimal albumin concentration would not bind. One of the ketone bodies, beta-
FIG.
1. Morphology
of unstained
islets
separated
El63
ISLETS
from
sheep
hydroxybutyric acid, was also used as an incubant. The incubation fluid was transferred in 100 ~1 aliquots to 10 x 75 mm disposable test tubes and frozen for the subsequent determination of insulin. Insulin was assayed in duplicate by a radioimmunoassay procedure (9) with crystalline bovine insulin as a standard. Glucose was determined by a glucose oxidase method (4). RESULTS
A modification of the sedimentation procedure (17) was used to isolate the islets of Langerhans from sheep pancreata, and with considerable practice a large number could be isolated from a single pancreas. Only islets exceeding 100 ,um in diameter were used for these studies. Under a dissection microscope against a black background, they appeared as a grayish-white spherical or ovoid structure; the appearance of unstained islets is shown in Fig. 1. By light microscopy, the islets appeared intact. The beta cells contained numerous secretory granules, and their appearance was similar to islets observed in sheep pancreata not treated with collagenase. Insulin release by isolated islets influenced by varying concentrations of butyrate and glucose is shown in Fig. 2. Glucose at the 0.5 mg/ml level appeared to have no effect, a result that is not surprising because this concentration is at the lower end of the normal blood
pancreas
compressed
under
a cover
slip.
Magnification
X 160
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
El64
H. N. JORDAN
glucose concentration in sheep. Normal plasma glucose in sheep would be between 0.5 and 1.0 mglml and rarely greater than 0.7 mg/ml. Even under conditions in which glucose was significantly raised above normal glucose levels, the insulin response was minimal. As the glucose level in the test medium was increased, however, insulin release was observed to increase in proportion to the glucose concentration (Fig. 2). Butyrate caused a relatively greater increase in insulin release than glucose (Fig. 2) and indicated that shortchain fatty acids do indeed have a direct effect on sheep islets. To determine the effect of various acids on insulin release, isolated islets were incubated with buffered sodium salts of a number of short- and medium-chain fatty acids in addition to butyrate, as well as lactic acid and beta-hydroxybutyric acid (Figs. 3 and 4). Two
AND
R. W. PHILLIPS
concentrations of each acid were used, and a dose response may be easily noted by comparison between the figures. Neither lactic, acetic, nor beta-hydroxybutyric, nor propionic acids appear to have an effect on insulin release because the results were not greatly different from the control incubation. It is interesting that decanoic acid does not cause insulin release. As the fatty acid chain length increases up to an B-carbon chain length, insulin release is increased. The lo-carbon fatty acid has no apparent effect on the islets. DISCUSSION
In vitro studies of insulin release with the use of either pieces of sheep pancreas or perfused sheep pancreas have been limited. However, methods have recently become available for the isolation of intact islets
500
420
c d 4
2 c:
FIG. 2. Insulin released from isolated islets incubated in vitro in presence of varying concentrations of butyrate and glucose. Numbers within bars represent total number of sheep. Vertical line at top of each bar represents mean k SE. Probability that values are different from control are listed at bottom of appropriate bar.
220
180
H
140
100 60
20 Blk.
1.5mM Butyrate
500
m
480
-
460
-
440
L
420
-
400
-
380
L
360
-
340 320
-
300
er
280
c
260
-
240
+
220
r
200
+
1RO
m
160
-
140 120
c
100 80
e c
60
c
20
-
2.5mM Butyratc PC 0.01
5.olnM Butyrate P-- 0.01
O.Smg/ml
1omM Butyrate PC 0.001
l.Omq/ml
Glucose
Glut
l.Smq/ml
ose
3.0mg/ml GlUCOSC~ Pf’ 0.1
Glucose
FIG. 3. Effect of fatty acid chain length on insulin release from isolated islets incubation in vitro at a concentration of 2.5 mM. Three sheep were used in each group of trials. Vertical line at top of each bar represents mean + SE. Probability that values are different from control are listed at bottom of appropriate bar.
f
t
Blk.
2.5mM
2.5mM 1,s L’ t a t c
Acetate
2.5mM Propionate
2.smM
2.5mM
Hydroxy-
But
yrate
butyratv
P-
0.05
2.5mM Hcxanoatta P* 0.02
2.smM
2.5mM
Oc tanoatt P* 0.01
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
El65
ISLETS
800 7GO 720 680 640 600 560 2 3; *r
480
g 2
440 400
/
520
FIG. 4. Effect of fatty acid chain length on insulin release from isolated islets incubated in vitro at a concentration of 5.0 mM. Three sheep were used for each group of experiments. Vertical line at top of each bar represents mean * SE. Probability that values are different from control are tested at bottom of appropriate bar.
360 374 2 Id2
320 280 240 200 160 120 80
t
40
Blk.
5.OmM Lactate
5.OlllM Acetate
5.omM Hydroxybutyrate
S.OmM Propionate
5.OlTLM Rut
yrate
P’
0.05
of Langerhans from mammalian pancreata in large numbers by the use of collagenase (17). Islets separated by such methods appear to r&pond to factors that regulate insulin synthesis and secretion (12, 17, 24). The proportionate rise in insulin release to glucose concentration agrees with earlier findings (17, 31, 36) that glucose induces an immediate and pronounced release of insulin from isolated rat pancreatic islets. The magnitude of the increase in insulin release from sheep islets with increasing glucose concentration is relatively low when compared to other species. The greater insulin release stimulated here in vitro by butyrate as opposed to glucose confirmed the in vivo results of Phillips et al. (25, 26), Manns et al. (21, ZZ), Bell et al. (Z), and Jones et al. (13). Although in this experiment the concentration of butyrate was greater than the levels seen in peripheral blood of sheep, the above-listed in vivo findings indicate that butyrate at physiological levels will stimulate insulin release (22). The nonstimulatory effect of lactic, acetic, and betahydroxybutyric acids on insulin release reported here is in agreement with studies in which slices of sheep pancreas were used (10, 11). The lack of response to acetic and beta-hydroxybutyric acids would indicate that metabolites of butyric and the longer-chain fatty acids are not involved with insulin release. Malaisse and Malaisse-Lagae (19) felt that metabolism of noncarbohydrate substances such as fatty acids and ketones caused the resultant insulin release. This concept is hard to accept because of the comparative lack of action of beta-hydroxybutyrate, acetate, and decanoate found in these experiments. The lack of effect of longer-chain fatty acids on insulin release in sheep has been previously demonstrated in vivo (1, 26). Sanbar et al. (30) found that octanoic acid affects glucose concentration and utilization in dogs. Subsequently, Sanbar and Martin (31) and Montague and Taylor (24) reported that butyrate and octanoate
5.OmM Hexanoatc P*- 0.01
IL1
5.OmM &tar-mate
5.Oll-M kc
atma
t e
PC 0.001
caused a marked insulin release from rat islets, a result that clearly indicates that the effect of fatty acids is not species-specific, More recently, Pi-Sunyer (27) reported that octanoate did not have a stimulatory effect of insulin release from pancreatic slices. The water-lipid solubility of the fatty acids would seem to be implicated. Unfortunately, this will not clarify all aspects of the results. Acetic, propionic, and butyric acids are completely miscible in water; hexanoic and octanoic acids are fairly soluble, whereas decanoic acid is only slightly soluble. The same variation in biological handling of fatty acids of increasing chain length can be seen if one considers the fate of absorbed fatty acids. Through C-8 most absorbed fatty acids travel directly to the liver via the portal system. Longer-chain fatty acids are incorporated into triglycerides and chylomicrons, enter the lacteals, and travel via lymphatics to enter the vascular system past the liver. This difference in handling has traditionally been related to solubility, Problems become evident when direct comparisons between in vitro and in vivo experiments are made. One such problem in these experiments is the relatively low level of albumin that was added to the incubating system. It would be capable of binding only a minor fraction of the added fatty acid, whereas under normal physiological conditions most of the fatty acid is bound to ablumin (33). However, earlier in vivo experiments with large quantities of short- and medium-chain FFA during periods of time when sheep had a very elevated level of endogenous FFA produced analogous results (25, 26). In those experiments, it would seem unlikely that the injected short-chain fatty acids would have replaced the more tightly bound endogenous long-chain fatty acids from their binding sites (33). Thus, one could assume that both bound and unbound short-chain fatty acids have the same effect on islet function. Other research has shown that short- and mediumchain fatty acids may directly affect tissues. Propionate,
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
El66 butyrate, and hexanoate, but not octanoate or decanoate, alter the morphological and growth characteristics in HeLa cells in tissue culture (6), an action thought to be due to changes in the properties of the plasma membrane. Butyrate has been shown to increase cyclic AMP levels in several different cell culture experiments. The action appears to be an effect of stimulation of adenyl cyclase, not inhibition of phosphodiesterase (28, 32). Cha nges in cyclic AMP have been convincingly linked as coupling factors between cell excitation and hormonal release although insulin secretion may also occur without an increase in cyclic AMP (29). In the experiments reported here, it is possible that butyrate as well as hexanoate and octanoate may act to directly alter cyclic AMP levels in islet cells and cause increased secretion. Phillips et al. (25, 26) had proposed an indirect mechanism for the glucogenic behavior of butyrate involving glucagon release and suggested that the increase in blood glucose after butyrate infusion arises from hepatic glycogenolysis and not from the conversion of butyrate to glucose. At this time, it would seem that fatty acids may cause secretion by both alpha and beta cells giving a simultaneous insulin and glucagon release, which would explain both the observed hyperinsulin and hyperglycemic responses (21, 22, 25, 26). Conversely, the fatty acids may act directly on the alpha cells and the subsequent release of insulin may be caused by increased glucagon as has been noted (8, 18, 34). We find this second hypothesis less attractive as the increase in insulin release due to fatty acids is of such magnitude that it does not seem likely to be caused by glucagon. By and large, glucagon-induced insulin release is not this great (15, 18). Other theories have been suggested; for instance, mammalian beta cells are believed to be freely permeable to glucose (5) and to phosphorylate it at a rate significantly higher than either pancreatic acini or liver (23). Montague and Taylor (24) found that shortchain fatty acids increased the quantity of glucose-6phosphate in the islet. They suggested that glucose-6phosphate may be regulated by fatty acids and in this
H. N.
JORDAN
AND
R. W. PHILLIPS
way influence insulin release. A reasonable assumption is that insulin release can be controlled by factors intimately related to glucose metabolism within the beta cell (16). However, this hypothesis does not preclude the possibility that short- and medium-chain fatty acids may be more effective in initiating insulin release in sheep. Whether this effect is controlled through changes in glycolytic intermediates by a direct effect of the fatty acids on islet cell membranes and morphology or by some other action such as changing CAMP levels is not clear. Another aspect to consider is that glucose levels are normally much higher in blood than are the short- or medium-chain length fatty acids, Whereas some of the glucose levels used in this study were within physiological ranges, the fatty acid levels were all higher than those normally encountered in ovine arterial blood. Normal levels of volatile fatty acids were not used because of their very low concentration in blood. In vivo studies (22) showed that physiological levels of short-chain fatty acids can stimulate insulin release without concomitant hyperglycemia or hyperketonemia. It seems plausible that in sheep, in contrast to man, shortand medium-chain fatty acids play a greater role in the control of insulin and glucagon release than does glucose. Unfortunately, the mechanism by which elevated levels of short- and mediumlength fatty acids stimulate hormone release from islets has not been clarified by this study. With slight modifications the islet isolation procedure described can be adapted for the isolation of pancreatic islets from other large animals to study the control of various substances on insulin and glucagon release. According to Lacy et al. (18) and Buchanan et al. (3), proteolytic destruction of released insulin has not been observed in isolated islets of the rat. Therefore, this technique eliminates some of the difficulties associated with the use of pancreatic slices and perfused pancreata because only islets are present in the incubation flasks. Received
for publication
14 February
1977.
REFERENCES 1. ASH, R. W., R. J. PENNINGTON, AND R. S. REID. The effect of short chain fatty acids on blood glucose concentration in sheep. Biochem. J. 90: 353-360, 1964. 2. BELL, J. P., L. A. SALAMONSE, G. W. HOLLAND, E. A. ESPINER, D. W. BEAVEN, AND D. S. HART. Autotransplantation of the pancreas in sheep: insulin secretion from the transplant. J. Endocrinol. 48: 511~-525, 1970. 3. BUCHANAN, K. D., J. E. VANCE, AND R. H. WILLIAMS. Insulin and glucagon release from isolated islets of Langerhans. Effect of enteric factors. Diabetes 18: 381-386, 1969. 4. CAMPBELL, L. A., AND D. S. KRONFELD. Estimates of low concentrations of plasma glucose using glucose oxidase. Am. J. Vet. Res. 22: 587-589, 1961. 5. COORE, H. G., AND P. J. RANDLE. Regulation of insulin secretion studies with pieces of rabbit pancreas incubated in vitro. Bio&em. J. 93: 66-78, 1964. 6. GINSBERG, E., D. SALOMON, T. SREEVALSAN, AND E. FREESE. Growth inhibition and morphological changes caused by lipophilic acids in mammalian cells. PFOC. Nat/. Acad. Sci. U.S. 70: 2545-2461, 1973. 7. GOMORI, G. Aldehyde fuchsin: a new stain for elastic tissue. A 111. J. C/in. Pathol. 20: 665-666, 1950.
G. M., L. L. BENNETT, D. F. SMITH, AND F. G. SCHMID. Effect of pulse administration of glucose or glucagon on insulin secretion in vitro. Metabolism 16: 222-233, 1967. HALES, C. N., AND P. J. RANDLE. Immunoassay of insulin with insulin antibody precipitate. Biochem. J. 88: 137-146, 1963. HERTENDLY, F., L. MACHLIN, AND D. M. KIPNIS. Further studies on the regulation of insulin and growth hormone secretion in the sheep. Endocrinology 84: 192-199, 1969. HERTENDLY, F., L. MACHLIN, Y. TAKAHASHI, AND D. M. KIPNIS. Insulin release from sheep pancreas in vitro. J. Endocrinol. 41: 605-606, 1968. HOWELL, S. L., AND K. W. TAYLOR. The secretion of newly synthesized insulin in vitro. Biochem. J. 102: 922-927, 1967. JONES, K. L., R. L. BELL, J. M. OYLER, AND D. D. GOETSH. Hyperglycemic effects of sodium butyrate in normal and pancreatectomized sheep. Am. J. Vet. Res. 31: 81412, 1970. KEEN, H., J. B. FIELD, AND 1. H. PASTAN. A simple method for in vitro metabolic studies using small volumes of tissue and medium. Metabolism 12: 143-147, 1963. KETTERER, H., A. M. EISENTRAUT, AND R. H. UNGER. Effect upon insulin secretion of physiologic doses of glucagon administered via the portal vein. Diabetes 16: 283-288, 1967.
8. GRODSKY,
9.
10.
11.
12. 13.
14.
15.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
EFFECT
OF
FATTY
ACIDS
ON
OVINE
PANCREATIC
El67
ISLETS
16. LACY, P. E. Quantitative histochemistry of the islets of Langerhans. I. Lactic, malic glucose-6-phosphate and 6-phosphogluconate dehydrogenase activities in beta cells and acini. Diabetes 11: 96-100, 1962. 17. LACY, P. E., AND M. KOSTIANOVSKY. A method for the isolation of intact islets of Langerhans from rat pancreas. Diabetes 16: 35-39, 1967. 18. LACY, P. E., D. A. YOUNG, AND C. J. FINK. Studies on insulin secretion in vitro from isolated islets of the rat pancreas. Endocrinology 83: 1155-1161, 1968. 19. MALAISSE, W. J., AND F. MALAISSE-LAGAE. Stimulation of insulin secretion by noncarbohydrate metabolites. J. Lab. C&n. Med. 72: 438-448, 1968. 20. MALAISSE, W. J., F. MALAISSE-LAGAE, AND II. MAYHEW. A possible role for the adenyl cyclase system in insulin secretion. J. Clin. Invest. 46: 1724-1734, 1967. 21. MANNS, J. G., AND J. M. BODA. Insulin release by acetate, propionate, butyrate and glucose in lambs and adult sheep. An2. J. Physiol. 212: 747-755, 1969. 22. MANNS, J. G., 3. M. BODA, AND R. F. WILLES. Probable role of propionate and butyrate in control of insulin secretion in sheep. Am. J. Physiol. 212: 756-764, 1967. 23. MATSCHINSKY, F. M., AND J. E. ELLERMAN. Metabolism of glucose in the islets of Langerhans. J. Biol. Chem. 243: 27302736, 1968. 24. MONTAGUE, W., AND K. W. TAYLOR. Regulation of insulin secretion by short chain fatty acids. Nature 217: 853, 1968. 25. PHILLIPS, R. W., A. L. BLACK, AND F. MOLLER. Butyrate induced glycogenolysis in hypoglycemic lambs. Life Sci. 4: 521-525, 1965. 26. PHILLIPS, R. W., W. A. HOUSE, R. A. MILLER, J. L. MOTT, AND D. L. SOOBY. Fatty acid, epinephrine, and glucagon hypergly-
27.
28. 29.
30.
31.
32.
33. 34.
35. 36.
cemia in normal and depancreatized sheep. Am. J. Physiol. 217: 1265-1268, 1969. PI-SUNYER, F. X. Effect of long and medium chain fatty acids on insulin secretion from pieces of rat pancreas. Proc. Sot. Exptl. Biol. Med. 149: 693-697, 1975. PRASAD, K. N., AND P. K. SINHA. Effect of sodium butyrate on mammalian cells in culture: a review. In uitro 12: 125-132, 1976. RASMUSSEN, H., AND D. B. P. GOODMAN. Relationships between calcium and cyclic nucleotides in cell activation. Physiol. Rev. 57: 421-509, 1977. SANBAR, S. S., G. HETENI, JR., N. FORBOTH, AND J. R. EVANS. Effects of infusion of octanoate on glucose concentration and rates of glucose production and utilization in dogs. Metabolism 14: 1311-1323, 1965. SANBAR, S. S., AND J. M. MARTIN. Stimulation by octanoate of insulin release from isolated rat pancreas. Metabolism 16: 482484, 1967. SHEPPARD, J. R., AND K. N. PRASAD. Cyclic AMP levels and the morphological differentiation of mouse neuroblastoma cells. Life Sci. 12 Part II: 431-439, 1973. SPECTOR, A. A. Fatty acid binding to plasma abiumin. J. Lipid Res. 16: 165-179, 1975. SUSSMAN, K. F., AND G. D. VAUGHAN. Insulin release after ACTH, glucagon and adenosine-3-phosphate kyclic AMP) in the perfused isolated rat pancreas. Diabetes 16: 449-454, 1967. UMBREIT, W. W., R. H. BURRIS, AND J- F. STAUFFER. Manometric Techniques (3rd ed.). Minneapolis: Burgess, 1957, pm 338. VANCE, J. D., K. D. BUCHANAN, D. R. CHALLONAR, AND R. H. WILLIAMS. The effect of glucose concentration on insulin and glucagon release from isolated islets of Langerhans of the rat. Diabetes 17: 187-193, 1968.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on December 25, 2018.
