menadione. CoQ 6 and CoQ10 (ubiquinone), duroquinone and durohydroquinone did not stimulate insulin release. CoQo's insulinotropism was enhanced in ...
Bioscience Reports, Vol. I1, No. 3, 1991
Stimulation of Insulin Release from Pancreatic Islets by Quinones Michael J. MacDonald Received May 17, 1991 Coenzyme Q (CoQ0) and other quinones were shown to be potent insulin secretagogues in the isolated pancreatic islet. The order of potency was CoQ0=benzoquinone=hydroquinonemenadione. C o Q 6 and CoQ10 (ubiquinone), duroquinone and durohydroquinone did not stimulate insulin release. CoQo's insulinotropism was enhanced in calcium-free medium and CoQ o appeared to stimulate only the second phase of insulin release. CoQo inhibited inositol mono-, bis- and trisphosphate formation. Inhibitors of mitochondrial respiration (rotenone, antimycin A, FCCP and cyanide) and the calcium channel blocker verapamil, did not inhibit CoQ0-induced insulin release. Dicumarol, an inhibitor of quinone reductase, did not inhibit CoQ0-induced insulin release, but it did inhibit glucose-induced insulin release suggesting that the enzyme and quinones play a role in glucose-induced insulin release. Quinones may stimulate insulin release by mimicking physiologicallyoccuring quinones, such as CoQl0 , by acting on the plasma membrane or in the cytosol. Exogenous quinones may bypass the quinone reductase reaction, as well as many reactions important for exocytosis. insulin secretion; pancreatic islets; coenzyme Qo; quinone reductase; insulin secretagogues; quinones. KEY WORDS:
Previous studies of pancreatic islets have implicated site II of the mitochondrial respiratory chain as being important for insulin release. Pancreatic islets (1) and insulinomas (2) are rich in the mitochondrial glycerol phosphate dehydrogenase. Succinate (when added to islets as its methyl ester to promote cellular penetration) has been shown to be an insulin secretagogue (3-5). Both succinate dehydrogenase, which is the first enzyme of succinate's metabolism, and glycerol phosphate dehydrogenase transfer electrons to ubiquinone (coenzyme Qa0) at site II of the electron transport chain. We, therefore, wondered if coenzyme Qlo, being a lipid-soluble mobile electron carrier, might exit the mitochondria and in some way interact in the cytosol and/or with the plasma membrane to participate in insulin secretion. As a first approach, we studied the effects of several water-soluble quinones, such as coenzyme Qo, on insulin release from isolated pancreatic islets. CoQo and other quinones without long hydrocarbon side chains did indeed stimulate insulin The Childrens Diabetes Center, University of Wisconsin Medical School, Madison, Wisconsin 53706 U.S.A. 165 0144-8463/91/0600-0165506.50/0 9 1991 Plenum Publishing Corporation
166
MacDonald
release, but quinones such as CoQ6 or CoQ10 did not stimulate insulin release. This may have been due to the limited solubility in aqueous solution of the quinones which have long hydrocarbon chains. Quinones m a y stimulate insulin release by mimicking physiologically-occurring quinones such as CoQ10. MATERIALS
AND METHODS
Pancreatic islets were isolated f r o m well-fed 200-250 g Sprague-Dawley rats by collagenase digestion (6, 7). Studies of insulin release (1, 8-10), 45Ca uptake (11), and inositol phosphate formation (12, 13) were as previously described. RESULTS Insulin Release One-half maximal stimulation of insulin release from pancreatic islets was attained at 50-100/~M CoQo. Maximal stimulation occurred at 100-500/~M CoQ0, and higher concentrations were progressively less stimulatory (Table 1 and Fig. 1). At 50 # M (which was near their limits of solubility), CoQ6, CoQx0 and durohydroquinone were not stimulatory. D u r o q u i n o n e was not stimulatory at concentrations of 50 # M and I m M , but benzoquinone and hydroquinone were roughly as stimulatory as CoQ0. Menadione has been reported to inhibit glucose-induced insulin release at concentrations b e l o w 50/~M (14). Although 50 # M menadione inhibited glucose-induced insulin release at a concentration of 1 mM, it stimulated insulin release in the absence of glucose (Table 1). Inhibitors Table 1. Stimulation by various quinones of insulin release from pancreatic islets and lack of inhibition by metabolic inhibitors. Results are the mean + S.D. and the number of replicates are in parentheses Quinone or other secretagogue None CoQo, 50/~M 0.5 mM 1 mM CO(~6, 50/tM CoQlo, 50 gM Duroquinone, 50/~M 1 mM Durohydroquinone, 50/~M Benzoquinone, 50/~M 1 mM Hydroquinone, 1 mM Menadione, 1 mM Glucose, 16.7 mM Glucose, 16.7 mM, plus menadione, 50 #M CoQo, 0.5 mM, plus menadione, 50/tM CoQo, 0.5 mM, plus rotenone, 1 ~uM CoQ0, 0.5 mM, plus FCCP, 10 #M Coenzyme Qo, 0.5 mM, plus sodium cyanide, 10 mM
Insulin release (/uU/5 islets/h) 29 + 12 (10) 169 + 29 (4) 385 + 33 (10) 293 9 64 (15) 38 + 9 (5) 34 + 7 (5) 42 zk10 (5) 51 + 17 (4) 36 • 6 (5) 124 + 60 (4) 260 • 59 (5) 295 + 31 (5) 132 + 66 (5) 16l + 19 (5) 32 + 21 (5) 367 • 46 (5) 386 i 30 (5) 367 + 46 (5) 37l + 40 (5)
Quinone-Induced Insulin Release
167
400
III
Z'5 9. J
19)
~(15)
"--
._ 200 ' ~
(5)
r
(~ ~ 100
__z
(14)
1!o
[CoenzymeQo] (mM) Fig. 1. Insulin released from pancreatic islets in the presence of various concentrations of CoQo.
of mitochondrial respiration had no effect on CoQ0-induced insulin release. Rotenone (1 #M), antimycin A (1/~M), FCCP (10/uM) and sodium cyanide (10 mM) had no effect on CoQo-induced insulin release (Table 1). Verapamil (10/~M), a calcium channel blocker, had no effect on CoQo-induced insulin release, but when islets were incubated in the presence of calcium-free medium, CoQo-induced insulin release was augmented. In companion experiments, glucose-induced insulin release was inhibited by calcium-free medium (Table 2). Table 2. Augmentation of the CoQo-induced insulin release by calcium-free medium and lack of inhibition by the calcium channel blocker verapamiL Results are the mean • (n = 5 replicate for each condition) Insulin release (/~U/5 islets/h)
Condition Experiment 1 No addition Coenzyme Qo, 0.1 mM Coenzyme Qo, 0.1 mM, plus Verapamil, 10/uM
8+0 237 + 90 264 + 76 Experiment 2
No addition Glucose, 16.7 mM Coenzyme Q0, 0.01 mM Coenzyme Q0, 0.1 mM Coenzyme Q0, 0.5 mM
Regular medium
Calcium-free medium
38 + 26 131 + 21 41 + 5 137 + 54 236 5:61
34 + 17 12 + 23 115 +41 595 + 10 554 + 81
168
MacDonald Table 3. Effect of dicumarol to inhibit CoQ0-induced insulin release. Results are the mean • ( n = 5 replicate incubations for each condition). Insulin release (tiU insulin/5 islets/h)
Condition No addition CoQo, 100 tiM CoQ0, 100 tiM + dicumarol, 10 tiM CoQo, 500 tiM CoQ0, 500 tiM + dicumarol, 10 tiM Glucose, 20 mM Glucose, 20 mM + dicumarol, 10 tiM
28 + 221 + 271 + 302 + 331 i 131 i 50 +
20 41 31 36 79 44 16
Dicumarol (10/~M or 50/~M), an inhibitor of quinone reductase, had no effect on insulin release induced by 0.1 mM and 0.5 mM CoQ0 in three seperate experiments, whereas glucose-induced insulin release was strongly inhibited (Table 3). Islets exposed to CoQ0 for one hour did not release insulin in response to glucose. CoQ0 stimulated primarily the second phase of insulin release. Insulin did not appear in the media in the first 15 min after CoQ0 was added to islets and maximal insulin release occurred between 30 and 60 min after CoQ0 was added to the islets. (0 + 0/~U/5 islets at 5 and 15 min, 128 + 42 and 399 + 139/zU/5 islets at 30 and 60 min, respectively).
Inositol Phosphate Formation CoQo was incubated with intact islets to test if its stimulation of insulin release was associated with increased inositol phosphate formation. Surprisingly, CoQ0 was a potent inhibitor of inositol phosphate formation during an incubation lasting 10 min. At a concentration of 0.1 mM, CoQ0 inhibited inositol mono-, bisand trisophosphate formation 40-60%and at 1 mM it inhibited 74-96%.
Islet 45Ca Uptake CoQo (0.5 mM) had no consistent effect o n 45Ca uptake by islets in three experiments. 45Ca uptake at 1, 5, 30 and 60 min after CoQ0 was added was not consistently different from that in the controls to which all ingredients but CoQ0 were added. DISCUSSION Quinone reductase appears not to be necessary for quinone-induced insulin release because dicumarol, which inhibits the islet quinone reductase (15) as it does the enzyme from other tissues (16, 17), does not inhibit CoQo-induced insulin release. Mitochondrial metabolism of quinones, such as by NADH dehydrogenase, is not likely to be the mechanism of quinone-induced release because inhibitors of oxidative phosphorylation, such as rotenone, FCCP and
Quinone-Induced Insulin Release
169
cyanide, do not inhibit CoQ0-induced insulin release. Thus, exogenous quinones can bypass any effects these steps might have in insulin release stimulated by intracellularly generated quinones. Quinones may stimulate insulin release by mimicking the action of physiologically-occurring quinones produced by mitochondrial metabolism of insulin secretagogues. However, quinones are known to cause the release of calcium from mitochondria (18-23) and this might also explain why quinones stimulate insulin release. Although cytosolic calcium content was not measured in the present study, it was observed that incubating islets in calcium-free medium, markedly potentiated quinone-induced insulin release. Calcium depletion might hypersensitize the excitation-contraction machinery of the beta cell to calcium released from mitochondria. Islets cultured for one hour with CoQ0 failed to release insulin in response to glucose. When quinones stimulate calcium release from mitochondria, the mitochondria swell and lose activity. Although a toxic effect may explain the refractoriness to glucose, it is important to emphasize that insulin appearing in the medium during one hour of exposure to CoQ0 is not leaked from damaged cells because cyanide, FCCP and rotenone, which poison cells, do not cause insulin to appear in the medium. Despite stimulating insulin release, quinones decreased inositol mono-, bisand trisphosphate formation during a very short exposure (
