- Email: [email protected]
Inhibition by l-valine and l-norleucine of 3-phenylpyruvate-induced insulin release
M6moires originaux
BIOCHIMIE. 1984, ~ , 353-360
Inhibition by L-valine and L-norleucine of 3-phenylpyruvate-induced insulin release. A. S E N E R and W.J. MALAIS~E.
Laboratory of Experimental Medicine, Brussels Free University School of Medicine, Brusselv, Belgium. (Refu le 8-12-1983, acceptd le 21-3-I984).
R6sum6 - - La sdcrdtion d~nsuline provoqude par le 3-phdnylpyruvate dans les ilots pancrdatiques
de rat est inhibde par la L-valine, la L-norleucine ou i'aminooxyacdtate. L'effet inhibiteur de ces trois agents coincide avec une moindre stimulation par ie 3-phdnylpyruvate de la production de 14C02 par des flots prd-incubds en prdsence de L-[ UJ4C] glutamine. Le 3-phdnylpyruvate augmente la vitesse de conversion de la L-valine en 2-ketoisovaldrate et de la L-norleucine en 2-ketocaproate. Cependant, le 3-phdnylpyruvate, qui augmente la ddcarboxylation oxydative du 2-ketoisovaldrate, inhibe la production de I~C02 par les ilots exposds d la D, L-[]-I4C] norleucine. Ces rdsuhats indiquent que des nutriments sdcrdtagogues distincts (tels que le 3-phdnyipyruvate et la Lnorleucine) dont chacun est capable de stimuler la sdcrdtion d'insuline, peuvent agir de manidre antagoniste sur le processus sdcrdtoire lorsqu'ils sont utilisds en combinaison. Les rdsultats du prdsent travail soulignent dgalement le r61e des oxydations mitochondriales et du transfert intracellulaire d'dquivalents rdduits dans la riposte sdcrdtoire au 3-phdnyipyruvate et f~ la norleucine. Mots-cl6s : ilots pancr6atiques / 3-ph6nylpyruvate / L-valine / L-norleucine / L-leucine / aminooxyac6tate / s6cr6tion d'insuline.
Insulin release induced by 3-phenylpyruvate in isolated rat pancreatic islets was inhibited by L-valine, L-norleucine or aminooxyacetate. The inhibitory effect of these three agents coincided with a lesser stimulation by 3-phenylpyruvate ofl4C02 output from islets prelabelled with L-[ U-uC] glutamine. Conversely, 3-phenylpyruvate augmented the rate of conversion of L-valine to 2-ketoisovalerate and that of L-norleucine to 2-ketocaproate. However, 3-phenylpyruvate, which increased 2-ketoisovalerate oxidative decarboxylation, inhibited t~C02 production by islets exposed to D, L-[1J4C] norleucine. These findings reveal that distinct nutrient secretagogues (e.g. 3-phenylpyruvate and L-norleucine), which are each able to stimulate insulin release, may act antagonistically upon the secretory process when used in combination. The present results also emphasize the relevance of both mitochondrial oxidation and intracellular transfer of reducing equivalents as determinants of the secretory response to such nutrients as 3-phenylpyruvate and norleucine. Summary
-
-
Key-words : pancreatic islets / 3-phenylpyruvate / L-valine / L-norlencine / L-leucine / aminooxyacetate / insulin release.
Introduction The 2-ketoacid 3-phenylpyruvate stimulates insulin release from isolated rat pancreatic islets
[1], even in the absence of another exogenous nutrient [2]. This stimulatory effect was recently ascribed to facilitation of a transamination process between 3-phenylpyruvate and endogenous
A. Sener and W.J. Malaisse
354
amino acids, the latter being converted to their corresponding 2-keto acids and further metabolized in the islet cells [2, 3]. However, 3-phenyipyruvate inhibits insulin release induced by high concentrations of either D-glucose (16.7 or 27.8 mM) or 2-ketoisocaproate (10 raM) [2]. This inhibitory effect coincides with a decrease in D-[U-~4C]-glucose or [U-t4C] 2-ketoisocaproate oxidation by the islets and was attributed to inhibition by 3-phenylpyruvate of the transfer of 2-keto acids (e.g. pyruvate produced by glycolysis) into the mitochondria [2]. These findings led us to examine, in the present study, the influence of L-valine, L-norleucine and aminooxyacetate upon insulin release provoked by 3-phenylpyruvate. ~-norleucine and /.-valine were selected for such a study because these amino acids do not stimulate insulin release in the absence of another exogenous nutrient, whereas their deamination product, 2-ketocaproate and 2-ketoisovaleratc are, respectively, a potent insulin secretagogue as efficient as 2-ketoisocaproate [4, 5] and a weak insulinotropic agent, only efficient in the presence of other nutrient [6] or non-nutrient [7] secretagogues. Aminooxyacetate was used because it inhibits extramitochondrial transamination reactions in the islets as in other tissues [8, 9].
Materials
and
1.0 ml of a bicarbonate-buffered incubation medium [10]. For measuring the productior~ of ~4CO2 and labelled acidic metabolites, groups of 15-20 islets each were incubated for 120min in 40-50 pa of the same incubation medium. The output of J4CO2 from islets labelled Outing 30 rain preincubatien with L-[U-J4C]glutamine (I.0 mM) was measured over 30 min incubation in gr~,ups of 8 islets placed in 100.u.l of a medium deprived of exogenous labelled nutrient [13]. L-[U-~4C]glutamine, L-[1-14C]leucine, L-[I-'4C]valine and L-[U-~C] valine were obtained from New England Nuclear (Boston, Massachussetts, USA). O,L-[1-14C]norleucine, prepared as described elsewhere [14], was a gift from Drs. S. Lenzen, H. Formanek and U. Panten (University of G6ttingen~ G6ttingen, F.R.G.). In the presence of the latter but not other radioactive amino acids, 3-phenylpyruvate augmented the control value (no islet) for both ~4CO2 output and the production of labelled acidic metabolites, even when the incubation medium contained antibiotics (pen;ciHin 50 l~g/ml and streptomycin 100 ttg/mi). In the presence of 3-phenylpyruvate, the blank value for ~4CO2 output wa~ increased four-fold and tha,t for labelled acidic metabolites about 28 % higher than in the absence of the pyruvate analogue. All results are expressed as the mean value (_+ SEM) together with the number of individual observations (in parentheses). The statistical significance of differences between mean values was assessed by use of Student's t-test.
methods Results
All experiments were performed with pancreatic islets derived from fed albino rats. The methods used to measure insulin release [10], the production of 14CO2 [11] and labelled acidic metabolites [12] from e~ogenous nutrients, and the output of "CO2 from islets prelabelled with L-[U-'4C]glutamine [13] are described in detail elsewhere. Briefly, for measurement of insulin release, groups of 8 islets were incubated for 90 rain in
Insulin release L-valine (P < 0.005) or ~-norleucine (P < 0.001) inhibited insulin release induced by 3-phenylpyruvate (Table I). In the presence of 3-phenylpyruvate, aminooxyacetate exerted an even more marked inhibitory effect, the rate of
TABLE I Effect of 3-phenylpyruvate and other agems u?on insulin release 3-phenylpyruvate (raM) . 10.0 10.0 10.0 10.0 I 0.0
L-valine (raM) .
__ 10.0 _ _ _
.
L-norleucine (raM)
Aminooxyacetate (raM)
. --. 10.0
----
_ 10.0
5.0 5.0
Insulin release (~tU/90 rain per islet) 10.4 38.4 26.0 22.1
4- 2.0 + 2.I _+ 4.0 + 2.3
(68)
(67) (24)
(43) 10.8 + 2.8 (43) 15.0 5:: 2.7 (44)
Mean values (± SEM) are shown together with the number of individual observations (in parentheses).
3-phenylpyr~vate-induced insulin release insulin release becoming comparable to that found in the absence of exogenous nutrient. In the presence of l:oth L-norleucine and aminooxyacetate, the secretory rate induced by 3-phenylpyruvate was significantly lower (P < 0.05) than in the presence of Lonorleucine, but not significantly different (P > 0.2) from that recorded in the presence of aminooxyacetate.
Metabolism of leucine, norleucine and valine As judged from the production of ;4CO2 and labelled acidic metabolites by islets exposed to L-[I-~4C] leucine, D,L-[I-~4C] norleucine or L-El24C]valine, 3-phenylpyruvate increased significantly (P < 0.00I in all cases) the rate of conversion of these amino acids to their corresponding 2-keto acids (Table II). The relative magnitude of the enhancing action of 3-phenylpyruvate was somewhat lower with O,L-norleucine ( + 58.7 :i: 12.4%) than with either L-leucine ( + 90.8 _+ 11.9 %) or L-valine ( + 94.3 + 10.4%). The production of 2-[1-'4Cl keto acids from either L-[I-m4C]leucine, O,L-[l-14C]norleucine or L-[l-~4C]-valine was also markedly increased by 3-phenylpyruvate. The pyruvate analogue, however, affected in a variable manner the production of ~4CO~ from these [1-'4C] amino acids. Thus, whereas 3-phenylpyruvate increased by 40.8 _+ 9.0 % (P < 0.001) the production of ~4CO: from L-[I-~4C]valine, the pyruvate analogue barely increased the generation of ~4CO2 from
355
L-[I-I~C] leucine ( + 29.5 +_ 16.4 %; P < 0.1) and markedly decreased ( - 49.3 + 7.5 %; P < 0.001) the output of ~4CO2 by islets exposed to Dd.-[I-14C] norleucine. We had previously shown that 3-phenylpyruvate fails to affect significantly the production of ~4CO2 from islets exposed to/.-[U-~4C] leucine [2]. No [U-~4C] norleucine was available to study the fate of this amino acid. We were able, however, to compare the effect of 3-phenylpyruvate upon the metabolism of L-[I-~4C]valine and /.-[U-~4C]valine, respectively. In the absence of 3-phenylpyruvate, the yield of ~CO2 was 2.71 + 0.17 times higher in islets exposed to L-[U-~4C]~Jaline than in those incubated in the rJresence of L-[I-14C] valine (Table II). Relative to the control value found in the absence of 3-phenylpyJuvate, the rate of L-valine conversion to 2-kctoisovalerate was increased by the pyruvate analogue to 194.3 + 10.4 and 226.9 +_ 16.8 %, as judged from the data obtained with L-[i-I4C] valine and L-[UJ4C] valine, respectively (P < 0.001 in both cases). These estimates take into account the incorporation of 14C into both CO2 and acidic metabolites derived from L-[~C] valine. In the absence and presence of 3-phenylpyruvate, respectively, the recovery of ~4C-atoms in acidic metabolites and CO2 produced by islets exposed to L-[U-~4C]valine averaged, when e~pressed as valine residues, 73.6 _+ 10.0 % and 85.8 + 6.4 % of the rate of L-valine conver-
TABLE II Effect of 3-phenylpyruvate upon the metabolism of amino acids 3-phenylpyruvate
Nutrient (mM) '4CO, output (pmol/120 min per islet) t.-[I-~4Clleucine (10.0) O.L-[l-:4C]norleucine (20.0) L-[|-14C]valine (10.0) L-[U-m4C]valine (10.0)
Nil
71.1 64.1 26.0 70.4
_ _ + _
3.9 (32) 2.7 (19) 1.3 (39) 4.3 (16)
'4C-keto acid production (pg-atoms of 14C/120 min per islet) L-[I-'4C] leucine (10.0) 44.4 _+ 5.7 (22) O.L-[l-~4C]norleucine (20.0) 60.7 _ 4.1 (19) L-[I-~4CIvaline (10.0) 7.6 +_ 1.2 (39) L-[U-~4C]valine (t00) 53.2 + 13.2 (16)
I0.0 mM
92.1 32.5 36.6 66.6
_ + +_ +
7.7 (19) 4.0(19) 2.1(27) 4.1 (16)
128.3 + 12.0 (10) 165.6 _ 14.3 (21) 28.7 +_ 2.8 (24) 213.8 - 20.4 (16)
Mean values (+ SEM) are shown together with the number of individual observations (in parentheses).
A. Sener and W.J. Malaisse
356
L-valin. ~-133.6:k1.8]--~2-K IV~
,sobutyryl-CoA~~r1"1±~9~-~C02
(7"6+1"2)
~--~3.8 +3.i~--acidic
3-phenylpyruvate L-phenylalanine //~~C L-valine --~65.3+3.5~2~2-KIY~isobutyryl-CoA--~ f _ • (28"7+2"8)
metabolites
02
~1Z6 + 5.8~-~acidic L
B
rnetabolites
FIG. 1. -- Schematic riew of the metabolism of L-valine (IO mM) in pancreatic islets incubated in the absence or presence of 3-phenylpyru rate (10 m M). Flow raies (rectangles) or values for the accumulation of 2-ketoisovalerate (2-Kiv; parentheses) are expressed as pmol/120 min per islet. The data in the right part of the figure are expressed as 4-C residues (i.e. as isobutyr~te equivalents).
sion to 2-ketoisovalerate as judged from the data collected in the presence of L-[IJ4C] valine (Table IS). As judged from the data obtained in the presence of t.-[UJ4C] valine and L-[IJ4C] valine, respectively, the absolute value for the ~4C-atoms not recovered as either ~4CO2 or labelled acid metabolites in islets exposed to L-[UJ4C] valine was virtually identical in the absence (44.4 +_ 17.7 pg-atoms of ~4C/120 min per islet) and presence (46.1 + 27.6 pg-atoms of ~4C/120 min per islet) of 3-phenylpyruvate. The pyruvate analogue not only increased the rate of L-valine conversion to 2-ketoisovalerate, but also affected the fate of 2-ketoisovalerate formed from exogenous L-valine. Thus, whereas 3-phenylpyruvate increased about fourfold the amount of 2-[124C] ketoisovalerate recovered in the islets and incubation medium, the pyruvate analogue increased by no more than 40 % the rate ,,..,¢
~
fl
141"~'I l r , ~ a . , , ~ g , , , , . ~ , , . . ~ l , ~ . , ~ t , ~
,-la,,'..-J~-I~,e~w-,,iat;nr.
with an increased accumulation (P < 0.05) of acidic me~abolites derived from isobutyryl-CoA (+ 13.8 +_. 6.1 pmol/120 min per islet) but failed to cause any increase in the production of ~4CO2 attributable to the catabolism of this CoA ester (Fig. 1). On the contrary, the production of ~4CO2 from isobutyryl-CoA was slightly but significantly (P < 0.02) decreased by 3-phenylpyruvate. These findings, which are compatible with the accumumr:tion of 3-hydroxyisobutyric acid, indicate that, relative to the rate of generation of isobutyrylCoA, its further oxidative catabolism was inhibited by 3-phenylpyruvate from a control value of 42.7 :!: 3.3 to 20.5 _+ 3.0 % (P < 0.001).
Production o f 14C0z by islets prelabelled with L-[UJ4 C] glutamine In the last series of experiments, we examined tI,hlaa~.,
liQia
!). Moreover, the 3-phenylpyruvate-induced increase in the rate of generation of isobutyryl-CoA (4- 10.6 __ 2.5 pmol/120 min per islet) coincided
;a ana afh,,~n,-.,a a La~.,aav~
~v a "
"l_r~h,an*d n.,,-~,.t o t ~ a J a l . I J l *,a v ~aL~w ,
-.I-paa~..,a
~_~ol;n~
L,- v ¢~aaal~.~
L-norleucine and aminooxyacetate upon the productioh of '4CO2 by islets prelabelled with L-[U-~4C] glutamine (Table III). During preincu-
TABLE I I I
Effect of 3-phenylpyruvate and other agents upon " C02 outputfrom isletsprelabeiled with L-[U-"C]glutamine Agent(s) Nil 3-phenylpyruvate (I0 m M ) 3-phenylpyruvate (I0 m M ) + L-valine(I0 m M ) 3-phenylpyruvate (I0 m M ) + L-norleucine(I0 m M ) 3-phenylpyruvate (I0 m M ) + aminooxyacetate (5 m M ) Aminooxyacetate (5 raM) Antimycin A (0.01m M )
Z4CO2output 14.45 26.63 21.33 23.28 19.96 14.40 5.65
_ ± ± +
0.22 (49) 0.34 (22) 0.99 (12) 0.45 (I0)
± 0.68 (10)
± 0.98 (lO) ± 0.22 (44)
Mean values (_+ SEM) for "CO2 output are expressed as percent of the "C content of the islets and are shown together with the number of individual observations (in parentheses).
3-phenylpyruvate-induced insulin release bation, the incorporation of radioactivity into islets corresponded to 9.53 + 0.15 pmol of L-[U-~4Clglutamine/30 min per islet (n = 102). The output of 14CO2 during incubation was markedly increased by 3-phenylpyruvate (P < 0.001). Thus, after correction for the data found in the presence of antimycin A, the output of ~4CO2 in the presence of 3-phenylpyruvate averaged 238 + 4 % of the mean basal value. We had previously show that neither /.-valine nor L-norleucine affected significantly ~4CO2 output from islets prelabelled with L-[U-~4C]glutamine [15]. However, in the presence of 3-phenylpyruvate, both amino acids significantly decreased the output of ~4CO2 (P < 0.001). After correction for the basal value, the stimulant action of 3-phenylpyruvate was decreased by 43.5 _+ 8.1 and 27.5 + 3.7 % in the presence of L-valine and ~-norleucine, respectively. Aminooxyacetate failed to affect ~4CO2 output in the absence of 3-phenylpyruvate, but significantly decreased ~4CO2 output (P < 0.001) in the presence of the pyruvate analogue. Nevertheless, in the presence of both 3-phenylpyruvate and aminooxyacetate, the output of ~4COz remained significantly higher than its basal value (P < 0.001)..in ,:,,:her words, aminooxyacetate impaired but failed to abolish the enhancing action of 3-phenylpyruvate upon m4CO2output by islets prelabelled with L-[U-;4C]glutamine.
Discussion The pyruvate analogue, 3-phenylpyruvate, is thought to stimulate insulin release by facilitating the catabolism of endogenous amino acids in the islet cells [2, 3]. It could be expected, therefore, that exogenous amino acids should, as a rule, enhance 3-phenylpyruvate-stimulated insulin release [16]. However, the present results reveal that L-valine or L-norleucine inhibit insulin release induced by 3-phenyipyruvate. Any biochemical explanation for such a situation should take into account three closely related aspects of the relationship between insulin release and fuel metabolism in the islet cells, namely (i) the difference in behaviour of distinct amino acids in the absence of 3-phenylpyruvate, (ii) the reciprocal influence of amino acids and 3-phenylpyruvate upon their respective metabolism, and (iii) the alteration by aminooxyacetate of the metabolic and secretory response to the pyruvate analogue. These three aspects will be discussed here in succession.
3 57
Metabolism of distinct amino acids From the data in Table II, it would appear that the rate of conversion of exogenous amino acids to their respective 2-keto acids displayed the following hierarchy : L-leucine = t.-norleucine > L-valine, at least if the assumption is made that only the t-isomer of norleucine is metabolized in the islet cells [14l. The validity of the latter assumption may well be questioned since unlabelled o-nodeucine was previously found to inhibit 2-ketoisocaproate-stimulated insulin release in a manner comparable to that seen with either L-nodeucine or the natural branched chain amino acids L-valine and L-isoleucine [12]. Several factors could account for differences in the rate of transamination of distinct amino acids. First, the activity of amino acid aminotransferase in islet homogenates is known to differ greatly with distinct amino acids [14, 17]. Second, the ratio between extracellular, cytosolic and mitochondrial concentration does not need to be identical with all amino acids. Third, L-ieucine, as distinct from L-norleucine or L-valine, activates glutamate dehydrogenase in the islets [18] and, hence, facilitates the regeneration of 2-ketoglutarate from glutamate [19]. Last, the catabolism of L-norleucine or L-valine, as distinct from that of L-leucine, may generate 2-keto acids (e.g. oxalacetate or pyruvate) susceptible to act as transamination partners in the initial conversion of these exogenous amino acids to 2-ketocaproate or 2-ketoisovalerate. Thus, by analogy with the metabolism of L-leucine, L-norleucine could be converted to 2-ketocaproate and then to valeryl-CoA, which itself could yield by I~-oxidation acetyl-CoA and proprionyl-CoA, the latter being then converted in a stepwise manner to succinyl-CoA, succinate, fumarate, malate and either pyruvate or oxalacerate. Likewise, L-valine, which is also converted to succinyl-CoA [20], could as well yield pyruvate or oxalacetate as transamination partners. However, L-valine may be less efficient than L-norleucine in this respect to the extent that part of the 2-ketoisovalerate flux throul~h the branched chain 2-ketoacid dehydrogenasc may be lost as [t-hydroxyisobutyrate (see below). The finding that the rate of conversion of D./.-[l-14C]-norleucine (20 raM) to 2-ketocaproate and ~4CO2 was not greatly different from the rate of /.-[1-14C]leucine (10 mM) conversion to 2-ketoisocaproate and 14CO2 does not necessarily detract from the view that the secretory response to these amino acids tightly depends on their capacity to augment oxidative fluxes in the islet
358
A. Sener and W.J. Malaisse
cells [14]. Indeed, L-leucine, at variance from ~-norleucine, by activating glutamate dehydrogenase, facilitates the catabolism of endogenous amino acids [19, 21]. Such a facilitating action is probably responsible for the fact that 3-phenylpyruvate fails to enhance L-leucine-stimulated insulin output [2, 16]. The much lower rate of transamination and oxidation of L-valine, relative to that of L-leucine or L-norleucine, is in good agreement with the fact that L-valine fails to stimulate insulin release, even in the presence of such nutrients as o-glucose or L-leucine [22].
Reciprocal effects of 3-phenylpyruvate and amino acids upon their respective metabolism The most simple explanation for the inhibitory action of L-valine or L-norleucine upon 3-phenylpyruvate-stimulated insulin release would be that these amino acids, by facilitating the extramitochondrial conversion of 3-phenylpyruvate to L-phenylalanine, decrease the access of 3-phenylpyruvate to the mitochondria, in which the pyruvate analogue would otherwise be optimally located for facilitating the transamination and further oxidative catabolism of endogenous amino acids [16]. Two independent series of observations support such a hypothesis. First, 3-phenylpyruvate indeed increased the rate of conversion of exogenous amino acids to their corresponding 2-keto acids. Second, L-vallne and L-leucine indeed impaired the stimulant action of 3-phenylpyruvate upon 14CO2 output from islets pre!abeUed with L-[U-m4C]g!utamine. In both cases, the metabolic changes appeared somewhat more marked with L-valine than L-leucine. Thus, in good agreement with the relative rate of transamination of these two amino acids in islet homogenates incubated with either 3-phenylpyrovate [3] or 2-ketoglutarate [17], L-valine was more efficient than L-norleucine in either its susceptibility to undergo increased transamination in the islets exposed to 3-phenylpyruvate or its capacity to decrease 3-phenylpyruvate-stimulated 14CO2 output from islets prelabelled L-[U-14C]glutamine. It could be argued that, at variance with these metabolic data, the rate of insulin release tended to be somewhat lower in the presence of both 3-phenylpyruvate and t.-norleucine than in the simultaneous presence of 3-phenylpyruvate and L-valine, as already observed in a prior series of experiments [16]. However, under these conditions, the rate of insulin release does not depend solely on the fate of 3-phenylpyruvate, but also on that of L-valine or L-norleucine.
Our data clearly indicate that 3-phenylpyruvate dramatically affected the metabolism of these amino acids. In the case of D,L-[I-14C] norleucine, the production of x4CO2 was impaired by 3-phenylpyruvate, despite a sizeable increase in the rate of conversion of norleucine to 2-ketocaproate. By analogy with the effect of 3-phenylpyruvate upon the oxidation of 2-[U-~4C]-ketoisocaproate [2], this supports the view that the direct or indirect transamination between L-norleucine and 3-phenylpyruvate occurs mainly in the cytosol and, hence, coincides with a severe inhibition by 3-phenylpyruvate of the transport of 2-ketocaproate, derived from norleucine, into the mitochondria. In the absence of 3-phenylpyruvate, L-norleucine fails to stimulate insulin release [22], suggesting that the catabolism of this amino acid and, if any, its influence upon the catabolism of endogenous nutrients arc not sufficient to increase 02 consumption above the threshold value required for stimulation of insulin release [23]. Since 3-phenylpyruvate further decreased the oxidation of norleucine, it is not surprizing that the catabolism of this amino acid was not sufficient to compensate for its inhibitory action upon the metabolic response to the pyruvate analogue. In
contrast
to
its
inhibitory
effect
on
O,L-[I-14C]norleucine oxidation, 3-phenylpyruvate increased the production of m4CO2 from L-[l-14C] valine. It is conceivable, therefore, that an increased generation of reducing equivalents at the level of the reaction catalyzed by the branched chain 2-keto acid dehydrogenase compensated, to a limited extent, for the vaiine-induced alteration of the metabolic response to 3-phenylpyruvate. Incidentally, the relative magnitude of the 3-phenylpyruvate-induced increment in ~4COe production from L-[l-~4C]valine was much less marked than would be expected from the relative magnitude of the increase in the rate of conversion of L-[l-14C] valine to 2-ketoisovalerate (Fig. 1). This again suggests that 3-phenylpyruvate inhibits the transport of 2-keto acids into the mitochondria. The comparison of the data collected with L-[1-14C] valine and L-[U-m4C]valine, respectively, also revealed that 3-phenylpyruvate impairs the further oxidative metabolism of isobutyryl-CoA, relative to its rate of generation. This could be due, inter alia, to both the increased catabolism of endogenous amino acids and the induction of a more reduced redox state in the islets exposed to 3-phenylpyruvate [3]. Our data indicate that, especially in the presence of 3-phenylpyruvate, a significant amount of acidic metabolites generated from
3 59
3-phenylpyruvate-induced insulin release
isobutyrvl-Coa escaped full oxidation. By analogy with the situation recently identified in isolated hepatocytes, this would be compatible with the accumulation of 3-hydroxyisobutyrate [20].
that this drug failed to reverse the inhibition of 3-phenylpyruvate-stimulated insulin release by L-norleucine. Conclusions
Inhibitory effect o f a m i n o o x y a c e t a t e
In the islets, aminooxyacetate inhibits extramitochondrial transamination reactions [8]. At first glance, the fact that aminooxyacetate abolished 3-phenylpyruvate-induced insulin release could suggest, therefore, that such extramitochondrial transamination reactions represent an essential determinant of the secretory response to the pyruvate analogue. This interpretation, however, is invalidated by two series of observations. First, in the absence of exogenous nutrients, aminooxyacetate fails to affect the basal output of NH~ [24], indicating that the transamination of endogenous amino acids takes place in the mitochondria. Likewise, aminooxyacetate failed to affect ~4CO2 output from islets preincubated with I.-[U-~4C] glutamine and incubated in the absence of exogenous nutrient. Second, in the islets prelabelled with L-[U-~4C]glutamine, aminooxyacetate impaired, but failed to abolish, the enhancing action of 3-phenylpyruvate upon ~4CO2 output. This again indicates that the facilitating action of 3-phenylpyruvate upon the metabolism of endogenous amino acids is not solely a cytosolic process. Moreover, even the partial inhibition by aminooxyacetate of 3-phenylpyruvatestimulated ~4CO2 output from the prelabelled islets does not necessarily indicate that the pyruW I L I I glutamate ....... •,,~,~ is engaged, together __..,L (derived from exogenous !-[U-]4C] glutamine, [see ref. 25]), in a cytosolic transamination process. Indeed, the full oxidation of I-[U-~4C] glutamine is dependent on a transfer of malate from the mitochondria to the cytosol [25, 26] and the latter process may itself be impaired in islets exposed to aminooxyacetate [8]. As an alternative explanation, we believe that the inhibitory effect of aminooxyacetate upon insulin release induced by 3-phenylpyruvate could be due to impairment of the malate-aspartate shuttle involved in the intracellular transfer of reducing equivalents between the mitochondrial and cytosolic compartments. This view is consistent with the fact that aminooxyacetate impairs the ionic and secretory response to such nutrient secretagogues as 2-ketoisocaproate, ew:n though the oxidative metabolism of 2-ketoisocaproate is unaffected by the transamination inhibitor [27, 28]. The proposed mode of action of aminooxyacetate would also account for the fact ~,,'~t~
,~...
n
!,-.
. . . .
"
In conclusion, the present data provide further support to the view that the mitochondrial oxidative catabolism of endogenous amino acids and the transfer of reducing equivalents to the cyto~oi represent two essential biochemical determiuants of the secretory response of pancreatic islets to 3-phenylpyruvate. The present work also iUustrates the fact that two distinct nutrients (e.g. 3-phenylpyruvate and L-norleucine), which are each able to stimulate insulin release under suitable experimental conditions [2, 22], may act antagonistically to one another when used in combination. In this respect, the situation encountered in islets exposed to both 3-phenylpyruvate and Lnorleucine (or L-valine) is reminiscent of that previously characterized in islets exposed to both 2-ketoisocaproate and a branched chain amino acid [12]. Such converging observations do not solely emphasize the concept that the regulation of oxidative events represents the fundamental mechanism by which the B-cell identifies circulating nutrients as insulin secretagogues [29, 30]. These observations also indicate that due attention should now be paid to the metabolic interaction between distinct substrates (glucose, amino acids, fatty acids, ketone bodies, lactate and pyruvate) in order to gain a better understanding of the regulation of insulin release by the heterogenous constellation of circulating nutrients.
Acknowledgements This work was supported by grants from the Belgian Foundation for Scientific Medical Research and Belgian Ministry of Scientific Policy. The authors wish to thank J. Schoonheydt, A. Tinant and M. Urbain for technical assistance and C. Demesmaeker for secretarial help. We are most grateful to Drs. S. Lenzen, H. Formanek and U. Panten for the generous gift Of D.L-[I-14C]norleucine.
REFERENCES I. Panten, U. & Langer, J. (1981) Biochem. J., 198, 353-~56. 2. Sener, A., Welsh, M., Lebrun, P., Garcia-Morales, P., Saceda, M., Malaisse-Lagae, F., Herchuelz, A., Valverde, I., Hellerstr6m, C. & Malaisse, W.J. (1983) Biochem. J., 210, 913-919.
360
A. Sener and W.J. Malaisse
3. Malaisse, W.J., Sener, A., Welsh, M., MalaisseLagae, F., Hellerstr6m, C. & Christophe, J. (1983) Biochem. J., 210, 921-927. 4. Hutton, J.C., Sener, A., Herchuelz, A., Atwater, I., Kawazu, S., Boschero, A.C., Somers, G., Devis, G. & Malaisse, W.J. (1980) Endocrinology. 106, 203-219. 5. Sener, A., Hutton, J.C. & Malaisse, W.J. (1981) Biochim. Biophys. Acta, 677, 32-38. 6. Hutton, J.C., Atwater, I. & Malaisse, W.J. (1980) Horm. Metab. Res., suppl, series 10, 31-37. 7. Sener, A. & Malaisse, W.J. (1979) Eur. J. Biochem., 98, 141-147. 8. Malaisse, W.J., Malaisse-Lagae, F. & Sener, A. (! 982) Endocrinology. 111, 392-397. 9. Corneil, N.W., Crow, K.E. & Witefoot, R.P. (1981) Biochem. J., 198, 219-223. 10. Malaisse, W.J., Brisson, G. & Malaisse-Lagae, F. (1970) J. Lab. Clin. Med.o 76, 895-902. 11. Carpinelli, A.R., Sener, A., Herchuelz, A. & Malaisse, W.J. (1980) Metabolism, 29, 540-545. 12. Hutton, J.C., Sener, A. & Malaisse, W.J. (1980) J. Biol. Chem., 255, 7340-7346. 13. Malaisse, W.J., Sener, A., Malaisse-Lagae, F., Hutton, J.C. & Christophe, J. (1981) Biochim. Biophys. Acta. 677, 39-49. 14. Lenzen, S., Formanek, H. & Panten, U. (1982) J. Biol. Chem., 257, 6631-6633. 15. Sener, A., Malaisse-Lagae, & Malaisse, W.J. (1981) Proc. Natl. Acad. Sci. USA, 78, 5460-5464. 16. Sener, A., Malaisse-Lagae, F. & Malaisse, W.J. (1983) Acta diabetol, lat., 20, 205-209.
17. Sener, A., Owen, A., Malaisse-Lagae, F. & Malaisse, W.J. (1982) Horm. Metab. Res~, !4, 405 -409. 18. Sener, A. & Malaisse, W.J. (1980) Nature, 288, 187-189. 19. Malaisse, W.J., Sener, A., Malaisse-Lagae, F., Welsh, M., Matthews, D.E., Bier, D.M. & Hdlerstr/~m, C. (1982) J. Biol. Chem., 257, 8731-8737. 20. Corkey, B.E., Martin-Requero, A., Walajtys-Rode, E., Williams, R.J. & Williamson, J.R. (1982) J. Biol. Chem., 257, 9668-9676. 21. Malaisse, W.J., Malaisse-Lagae, F. & Sener, A. (1984) Biochim. Biophys. Acta, 797, 194-202. 22. Sener, A., Somers, G., Devis, G. & Malaisse, W.J. (1981) Diabetologia, 21, 135-142. 23. Hutton, J.C. & Malaisse, W.J. (1980) Diabetologia, 18, 395-405. 24. Sener, A., Best, L., Malaisse-Lagae, F. & Malaisse, W.J. (1984) Arch. Biochem. Biophys., 229, 155-169. 25. Malaisse, W.J., Sener, A., Carpinelli, A.R., Anjaneyulu, K., Lebrun, P., Herchuelz, A. & Christophe, J. (1980) Mol. Cell. Endocrinol., 20, 171 - 189. 26. Sener, A., Malaisse-Lagae, F. & Malaisse, W.J. (! 982) Biochem. J., 202, 309-3 ! 6. 27. Hutton, J.C., Sener, A. & Malaisse, W.J. (1979) Biochem. J.. 184, 291-301. 28. Lebrun, P., Malaisse, W.J. & Herchuelz, A. (1983) Am. J. Physiol., 245, E38-E46. 29. Malaisse, W.J., Sener, A., Herchuelz, A. & Hutton, J.C. (1979) Metabolism, 28, 373-386. 30. Malaisse, W.J. (I 983) Diab. M~tab., 9, 313-320.
BIOCHIMIE. 1984, ~ , 353-360
Inhibition by L-valine and L-norleucine of 3-phenylpyruvate-induced insulin release. A. S E N E R and W.J. MALAIS~E.
Laboratory of Experimental Medicine, Brussels Free University School of Medicine, Brusselv, Belgium. (Refu le 8-12-1983, acceptd le 21-3-I984).
R6sum6 - - La sdcrdtion d~nsuline provoqude par le 3-phdnylpyruvate dans les ilots pancrdatiques
de rat est inhibde par la L-valine, la L-norleucine ou i'aminooxyacdtate. L'effet inhibiteur de ces trois agents coincide avec une moindre stimulation par ie 3-phdnylpyruvate de la production de 14C02 par des flots prd-incubds en prdsence de L-[ UJ4C] glutamine. Le 3-phdnylpyruvate augmente la vitesse de conversion de la L-valine en 2-ketoisovaldrate et de la L-norleucine en 2-ketocaproate. Cependant, le 3-phdnylpyruvate, qui augmente la ddcarboxylation oxydative du 2-ketoisovaldrate, inhibe la production de I~C02 par les ilots exposds d la D, L-[]-I4C] norleucine. Ces rdsuhats indiquent que des nutriments sdcrdtagogues distincts (tels que le 3-phdnyipyruvate et la Lnorleucine) dont chacun est capable de stimuler la sdcrdtion d'insuline, peuvent agir de manidre antagoniste sur le processus sdcrdtoire lorsqu'ils sont utilisds en combinaison. Les rdsultats du prdsent travail soulignent dgalement le r61e des oxydations mitochondriales et du transfert intracellulaire d'dquivalents rdduits dans la riposte sdcrdtoire au 3-phdnyipyruvate et f~ la norleucine. Mots-cl6s : ilots pancr6atiques / 3-ph6nylpyruvate / L-valine / L-norleucine / L-leucine / aminooxyac6tate / s6cr6tion d'insuline.
Insulin release induced by 3-phenylpyruvate in isolated rat pancreatic islets was inhibited by L-valine, L-norleucine or aminooxyacetate. The inhibitory effect of these three agents coincided with a lesser stimulation by 3-phenylpyruvate ofl4C02 output from islets prelabelled with L-[ U-uC] glutamine. Conversely, 3-phenylpyruvate augmented the rate of conversion of L-valine to 2-ketoisovalerate and that of L-norleucine to 2-ketocaproate. However, 3-phenylpyruvate, which increased 2-ketoisovalerate oxidative decarboxylation, inhibited t~C02 production by islets exposed to D, L-[1J4C] norleucine. These findings reveal that distinct nutrient secretagogues (e.g. 3-phenylpyruvate and L-norleucine), which are each able to stimulate insulin release, may act antagonistically upon the secretory process when used in combination. The present results also emphasize the relevance of both mitochondrial oxidation and intracellular transfer of reducing equivalents as determinants of the secretory response to such nutrients as 3-phenylpyruvate and norleucine. Summary
-
-
Key-words : pancreatic islets / 3-phenylpyruvate / L-valine / L-norlencine / L-leucine / aminooxyacetate / insulin release.
Introduction The 2-ketoacid 3-phenylpyruvate stimulates insulin release from isolated rat pancreatic islets
[1], even in the absence of another exogenous nutrient [2]. This stimulatory effect was recently ascribed to facilitation of a transamination process between 3-phenylpyruvate and endogenous
A. Sener and W.J. Malaisse
354
amino acids, the latter being converted to their corresponding 2-keto acids and further metabolized in the islet cells [2, 3]. However, 3-phenyipyruvate inhibits insulin release induced by high concentrations of either D-glucose (16.7 or 27.8 mM) or 2-ketoisocaproate (10 raM) [2]. This inhibitory effect coincides with a decrease in D-[U-~4C]-glucose or [U-t4C] 2-ketoisocaproate oxidation by the islets and was attributed to inhibition by 3-phenylpyruvate of the transfer of 2-keto acids (e.g. pyruvate produced by glycolysis) into the mitochondria [2]. These findings led us to examine, in the present study, the influence of L-valine, L-norleucine and aminooxyacetate upon insulin release provoked by 3-phenylpyruvate. ~-norleucine and /.-valine were selected for such a study because these amino acids do not stimulate insulin release in the absence of another exogenous nutrient, whereas their deamination product, 2-ketocaproate and 2-ketoisovaleratc are, respectively, a potent insulin secretagogue as efficient as 2-ketoisocaproate [4, 5] and a weak insulinotropic agent, only efficient in the presence of other nutrient [6] or non-nutrient [7] secretagogues. Aminooxyacetate was used because it inhibits extramitochondrial transamination reactions in the islets as in other tissues [8, 9].
Materials
and
1.0 ml of a bicarbonate-buffered incubation medium [10]. For measuring the productior~ of ~4CO2 and labelled acidic metabolites, groups of 15-20 islets each were incubated for 120min in 40-50 pa of the same incubation medium. The output of J4CO2 from islets labelled Outing 30 rain preincubatien with L-[U-J4C]glutamine (I.0 mM) was measured over 30 min incubation in gr~,ups of 8 islets placed in 100.u.l of a medium deprived of exogenous labelled nutrient [13]. L-[U-~4C]glutamine, L-[1-14C]leucine, L-[I-'4C]valine and L-[U-~C] valine were obtained from New England Nuclear (Boston, Massachussetts, USA). O,L-[1-14C]norleucine, prepared as described elsewhere [14], was a gift from Drs. S. Lenzen, H. Formanek and U. Panten (University of G6ttingen~ G6ttingen, F.R.G.). In the presence of the latter but not other radioactive amino acids, 3-phenylpyruvate augmented the control value (no islet) for both ~4CO2 output and the production of labelled acidic metabolites, even when the incubation medium contained antibiotics (pen;ciHin 50 l~g/ml and streptomycin 100 ttg/mi). In the presence of 3-phenylpyruvate, the blank value for ~4CO2 output wa~ increased four-fold and tha,t for labelled acidic metabolites about 28 % higher than in the absence of the pyruvate analogue. All results are expressed as the mean value (_+ SEM) together with the number of individual observations (in parentheses). The statistical significance of differences between mean values was assessed by use of Student's t-test.
methods Results
All experiments were performed with pancreatic islets derived from fed albino rats. The methods used to measure insulin release [10], the production of 14CO2 [11] and labelled acidic metabolites [12] from e~ogenous nutrients, and the output of "CO2 from islets prelabelled with L-[U-'4C]glutamine [13] are described in detail elsewhere. Briefly, for measurement of insulin release, groups of 8 islets were incubated for 90 rain in
Insulin release L-valine (P < 0.005) or ~-norleucine (P < 0.001) inhibited insulin release induced by 3-phenylpyruvate (Table I). In the presence of 3-phenylpyruvate, aminooxyacetate exerted an even more marked inhibitory effect, the rate of
TABLE I Effect of 3-phenylpyruvate and other agems u?on insulin release 3-phenylpyruvate (raM) . 10.0 10.0 10.0 10.0 I 0.0
L-valine (raM) .
__ 10.0 _ _ _
.
L-norleucine (raM)
Aminooxyacetate (raM)
. --. 10.0
----
_ 10.0
5.0 5.0
Insulin release (~tU/90 rain per islet) 10.4 38.4 26.0 22.1
4- 2.0 + 2.I _+ 4.0 + 2.3
(68)
(67) (24)
(43) 10.8 + 2.8 (43) 15.0 5:: 2.7 (44)
Mean values (± SEM) are shown together with the number of individual observations (in parentheses).
3-phenylpyr~vate-induced insulin release insulin release becoming comparable to that found in the absence of exogenous nutrient. In the presence of l:oth L-norleucine and aminooxyacetate, the secretory rate induced by 3-phenylpyruvate was significantly lower (P < 0.05) than in the presence of Lonorleucine, but not significantly different (P > 0.2) from that recorded in the presence of aminooxyacetate.
Metabolism of leucine, norleucine and valine As judged from the production of ;4CO2 and labelled acidic metabolites by islets exposed to L-[I-~4C] leucine, D,L-[I-~4C] norleucine or L-El24C]valine, 3-phenylpyruvate increased significantly (P < 0.00I in all cases) the rate of conversion of these amino acids to their corresponding 2-keto acids (Table II). The relative magnitude of the enhancing action of 3-phenylpyruvate was somewhat lower with O,L-norleucine ( + 58.7 :i: 12.4%) than with either L-leucine ( + 90.8 _+ 11.9 %) or L-valine ( + 94.3 + 10.4%). The production of 2-[1-'4Cl keto acids from either L-[I-m4C]leucine, O,L-[l-14C]norleucine or L-[l-~4C]-valine was also markedly increased by 3-phenylpyruvate. The pyruvate analogue, however, affected in a variable manner the production of ~4CO~ from these [1-'4C] amino acids. Thus, whereas 3-phenylpyruvate increased by 40.8 _+ 9.0 % (P < 0.001) the production of ~4CO: from L-[I-~4C]valine, the pyruvate analogue barely increased the generation of ~4CO2 from
355
L-[I-I~C] leucine ( + 29.5 +_ 16.4 %; P < 0.1) and markedly decreased ( - 49.3 + 7.5 %; P < 0.001) the output of ~4CO2 by islets exposed to Dd.-[I-14C] norleucine. We had previously shown that 3-phenylpyruvate fails to affect significantly the production of ~4CO2 from islets exposed to/.-[U-~4C] leucine [2]. No [U-~4C] norleucine was available to study the fate of this amino acid. We were able, however, to compare the effect of 3-phenylpyruvate upon the metabolism of L-[I-~4C]valine and /.-[U-~4C]valine, respectively. In the absence of 3-phenylpyruvate, the yield of ~CO2 was 2.71 + 0.17 times higher in islets exposed to L-[U-~4C]~Jaline than in those incubated in the rJresence of L-[I-14C] valine (Table II). Relative to the control value found in the absence of 3-phenylpyJuvate, the rate of L-valine conversion to 2-kctoisovalerate was increased by the pyruvate analogue to 194.3 + 10.4 and 226.9 +_ 16.8 %, as judged from the data obtained with L-[i-I4C] valine and L-[UJ4C] valine, respectively (P < 0.001 in both cases). These estimates take into account the incorporation of 14C into both CO2 and acidic metabolites derived from L-[~C] valine. In the absence and presence of 3-phenylpyruvate, respectively, the recovery of ~4C-atoms in acidic metabolites and CO2 produced by islets exposed to L-[U-~4C]valine averaged, when e~pressed as valine residues, 73.6 _+ 10.0 % and 85.8 + 6.4 % of the rate of L-valine conver-
TABLE II Effect of 3-phenylpyruvate upon the metabolism of amino acids 3-phenylpyruvate
Nutrient (mM) '4CO, output (pmol/120 min per islet) t.-[I-~4Clleucine (10.0) O.L-[l-:4C]norleucine (20.0) L-[|-14C]valine (10.0) L-[U-m4C]valine (10.0)
Nil
71.1 64.1 26.0 70.4
_ _ + _
3.9 (32) 2.7 (19) 1.3 (39) 4.3 (16)
'4C-keto acid production (pg-atoms of 14C/120 min per islet) L-[I-'4C] leucine (10.0) 44.4 _+ 5.7 (22) O.L-[l-~4C]norleucine (20.0) 60.7 _ 4.1 (19) L-[I-~4CIvaline (10.0) 7.6 +_ 1.2 (39) L-[U-~4C]valine (t00) 53.2 + 13.2 (16)
I0.0 mM
92.1 32.5 36.6 66.6
_ + +_ +
7.7 (19) 4.0(19) 2.1(27) 4.1 (16)
128.3 + 12.0 (10) 165.6 _ 14.3 (21) 28.7 +_ 2.8 (24) 213.8 - 20.4 (16)
Mean values (+ SEM) are shown together with the number of individual observations (in parentheses).
A. Sener and W.J. Malaisse
356
L-valin. ~-133.6:k1.8]--~2-K IV~
,sobutyryl-CoA~~r1"1±~9~-~C02
(7"6+1"2)
~--~3.8 +3.i~--acidic
3-phenylpyruvate L-phenylalanine //~~C L-valine --~65.3+3.5~2~2-KIY~isobutyryl-CoA--~ f _ • (28"7+2"8)
metabolites
02
~1Z6 + 5.8~-~acidic L
B
rnetabolites
FIG. 1. -- Schematic riew of the metabolism of L-valine (IO mM) in pancreatic islets incubated in the absence or presence of 3-phenylpyru rate (10 m M). Flow raies (rectangles) or values for the accumulation of 2-ketoisovalerate (2-Kiv; parentheses) are expressed as pmol/120 min per islet. The data in the right part of the figure are expressed as 4-C residues (i.e. as isobutyr~te equivalents).
sion to 2-ketoisovalerate as judged from the data collected in the presence of L-[IJ4C] valine (Table IS). As judged from the data obtained in the presence of t.-[UJ4C] valine and L-[IJ4C] valine, respectively, the absolute value for the ~4C-atoms not recovered as either ~4CO2 or labelled acid metabolites in islets exposed to L-[UJ4C] valine was virtually identical in the absence (44.4 +_ 17.7 pg-atoms of ~4C/120 min per islet) and presence (46.1 + 27.6 pg-atoms of ~4C/120 min per islet) of 3-phenylpyruvate. The pyruvate analogue not only increased the rate of L-valine conversion to 2-ketoisovalerate, but also affected the fate of 2-ketoisovalerate formed from exogenous L-valine. Thus, whereas 3-phenylpyruvate increased about fourfold the amount of 2-[124C] ketoisovalerate recovered in the islets and incubation medium, the pyruvate analogue increased by no more than 40 % the rate ,,..,¢
~
fl
141"~'I l r , ~ a . , , ~ g , , , , . ~ , , . . ~ l , ~ . , ~ t , ~
,-la,,'..-J~-I~,e~w-,,iat;nr.
with an increased accumulation (P < 0.05) of acidic me~abolites derived from isobutyryl-CoA (+ 13.8 +_. 6.1 pmol/120 min per islet) but failed to cause any increase in the production of ~4CO2 attributable to the catabolism of this CoA ester (Fig. 1). On the contrary, the production of ~4CO2 from isobutyryl-CoA was slightly but significantly (P < 0.02) decreased by 3-phenylpyruvate. These findings, which are compatible with the accumumr:tion of 3-hydroxyisobutyric acid, indicate that, relative to the rate of generation of isobutyrylCoA, its further oxidative catabolism was inhibited by 3-phenylpyruvate from a control value of 42.7 :!: 3.3 to 20.5 _+ 3.0 % (P < 0.001).
Production o f 14C0z by islets prelabelled with L-[UJ4 C] glutamine In the last series of experiments, we examined tI,hlaa~.,
liQia
!). Moreover, the 3-phenylpyruvate-induced increase in the rate of generation of isobutyryl-CoA (4- 10.6 __ 2.5 pmol/120 min per islet) coincided
;a ana afh,,~n,-.,a a La~.,aav~
~v a "
"l_r~h,an*d n.,,-~,.t o t ~ a J a l . I J l *,a v ~aL~w ,
-.I-paa~..,a
~_~ol;n~
L,- v ¢~aaal~.~
L-norleucine and aminooxyacetate upon the productioh of '4CO2 by islets prelabelled with L-[U-~4C] glutamine (Table III). During preincu-
TABLE I I I
Effect of 3-phenylpyruvate and other agents upon " C02 outputfrom isletsprelabeiled with L-[U-"C]glutamine Agent(s) Nil 3-phenylpyruvate (I0 m M ) 3-phenylpyruvate (I0 m M ) + L-valine(I0 m M ) 3-phenylpyruvate (I0 m M ) + L-norleucine(I0 m M ) 3-phenylpyruvate (I0 m M ) + aminooxyacetate (5 m M ) Aminooxyacetate (5 raM) Antimycin A (0.01m M )
Z4CO2output 14.45 26.63 21.33 23.28 19.96 14.40 5.65
_ ± ± +
0.22 (49) 0.34 (22) 0.99 (12) 0.45 (I0)
± 0.68 (10)
± 0.98 (lO) ± 0.22 (44)
Mean values (_+ SEM) for "CO2 output are expressed as percent of the "C content of the islets and are shown together with the number of individual observations (in parentheses).
3-phenylpyruvate-induced insulin release bation, the incorporation of radioactivity into islets corresponded to 9.53 + 0.15 pmol of L-[U-~4Clglutamine/30 min per islet (n = 102). The output of 14CO2 during incubation was markedly increased by 3-phenylpyruvate (P < 0.001). Thus, after correction for the data found in the presence of antimycin A, the output of ~4CO2 in the presence of 3-phenylpyruvate averaged 238 + 4 % of the mean basal value. We had previously show that neither /.-valine nor L-norleucine affected significantly ~4CO2 output from islets prelabelled with L-[U-~4C]glutamine [15]. However, in the presence of 3-phenylpyruvate, both amino acids significantly decreased the output of ~4CO2 (P < 0.001). After correction for the basal value, the stimulant action of 3-phenylpyruvate was decreased by 43.5 _+ 8.1 and 27.5 + 3.7 % in the presence of L-valine and ~-norleucine, respectively. Aminooxyacetate failed to affect ~4CO2 output in the absence of 3-phenylpyruvate, but significantly decreased ~4CO2 output (P < 0.001) in the presence of the pyruvate analogue. Nevertheless, in the presence of both 3-phenylpyruvate and aminooxyacetate, the output of ~4COz remained significantly higher than its basal value (P < 0.001)..in ,:,,:her words, aminooxyacetate impaired but failed to abolish the enhancing action of 3-phenylpyruvate upon m4CO2output by islets prelabelled with L-[U-;4C]glutamine.
Discussion The pyruvate analogue, 3-phenylpyruvate, is thought to stimulate insulin release by facilitating the catabolism of endogenous amino acids in the islet cells [2, 3]. It could be expected, therefore, that exogenous amino acids should, as a rule, enhance 3-phenylpyruvate-stimulated insulin release [16]. However, the present results reveal that L-valine or L-norleucine inhibit insulin release induced by 3-phenyipyruvate. Any biochemical explanation for such a situation should take into account three closely related aspects of the relationship between insulin release and fuel metabolism in the islet cells, namely (i) the difference in behaviour of distinct amino acids in the absence of 3-phenylpyruvate, (ii) the reciprocal influence of amino acids and 3-phenylpyruvate upon their respective metabolism, and (iii) the alteration by aminooxyacetate of the metabolic and secretory response to the pyruvate analogue. These three aspects will be discussed here in succession.
3 57
Metabolism of distinct amino acids From the data in Table II, it would appear that the rate of conversion of exogenous amino acids to their respective 2-keto acids displayed the following hierarchy : L-leucine = t.-norleucine > L-valine, at least if the assumption is made that only the t-isomer of norleucine is metabolized in the islet cells [14l. The validity of the latter assumption may well be questioned since unlabelled o-nodeucine was previously found to inhibit 2-ketoisocaproate-stimulated insulin release in a manner comparable to that seen with either L-nodeucine or the natural branched chain amino acids L-valine and L-isoleucine [12]. Several factors could account for differences in the rate of transamination of distinct amino acids. First, the activity of amino acid aminotransferase in islet homogenates is known to differ greatly with distinct amino acids [14, 17]. Second, the ratio between extracellular, cytosolic and mitochondrial concentration does not need to be identical with all amino acids. Third, L-ieucine, as distinct from L-norleucine or L-valine, activates glutamate dehydrogenase in the islets [18] and, hence, facilitates the regeneration of 2-ketoglutarate from glutamate [19]. Last, the catabolism of L-norleucine or L-valine, as distinct from that of L-leucine, may generate 2-keto acids (e.g. oxalacetate or pyruvate) susceptible to act as transamination partners in the initial conversion of these exogenous amino acids to 2-ketocaproate or 2-ketoisovalerate. Thus, by analogy with the metabolism of L-leucine, L-norleucine could be converted to 2-ketocaproate and then to valeryl-CoA, which itself could yield by I~-oxidation acetyl-CoA and proprionyl-CoA, the latter being then converted in a stepwise manner to succinyl-CoA, succinate, fumarate, malate and either pyruvate or oxalacerate. Likewise, L-valine, which is also converted to succinyl-CoA [20], could as well yield pyruvate or oxalacetate as transamination partners. However, L-valine may be less efficient than L-norleucine in this respect to the extent that part of the 2-ketoisovalerate flux throul~h the branched chain 2-ketoacid dehydrogenasc may be lost as [t-hydroxyisobutyrate (see below). The finding that the rate of conversion of D./.-[l-14C]-norleucine (20 raM) to 2-ketocaproate and ~4CO2 was not greatly different from the rate of /.-[1-14C]leucine (10 mM) conversion to 2-ketoisocaproate and 14CO2 does not necessarily detract from the view that the secretory response to these amino acids tightly depends on their capacity to augment oxidative fluxes in the islet
358
A. Sener and W.J. Malaisse
cells [14]. Indeed, L-leucine, at variance from ~-norleucine, by activating glutamate dehydrogenase, facilitates the catabolism of endogenous amino acids [19, 21]. Such a facilitating action is probably responsible for the fact that 3-phenylpyruvate fails to enhance L-leucine-stimulated insulin output [2, 16]. The much lower rate of transamination and oxidation of L-valine, relative to that of L-leucine or L-norleucine, is in good agreement with the fact that L-valine fails to stimulate insulin release, even in the presence of such nutrients as o-glucose or L-leucine [22].
Reciprocal effects of 3-phenylpyruvate and amino acids upon their respective metabolism The most simple explanation for the inhibitory action of L-valine or L-norleucine upon 3-phenylpyruvate-stimulated insulin release would be that these amino acids, by facilitating the extramitochondrial conversion of 3-phenylpyruvate to L-phenylalanine, decrease the access of 3-phenylpyruvate to the mitochondria, in which the pyruvate analogue would otherwise be optimally located for facilitating the transamination and further oxidative catabolism of endogenous amino acids [16]. Two independent series of observations support such a hypothesis. First, 3-phenylpyruvate indeed increased the rate of conversion of exogenous amino acids to their corresponding 2-keto acids. Second, L-vallne and L-leucine indeed impaired the stimulant action of 3-phenylpyruvate upon 14CO2 output from islets pre!abeUed with L-[U-m4C]g!utamine. In both cases, the metabolic changes appeared somewhat more marked with L-valine than L-leucine. Thus, in good agreement with the relative rate of transamination of these two amino acids in islet homogenates incubated with either 3-phenylpyrovate [3] or 2-ketoglutarate [17], L-valine was more efficient than L-norleucine in either its susceptibility to undergo increased transamination in the islets exposed to 3-phenylpyruvate or its capacity to decrease 3-phenylpyruvate-stimulated 14CO2 output from islets prelabelled L-[U-14C]glutamine. It could be argued that, at variance with these metabolic data, the rate of insulin release tended to be somewhat lower in the presence of both 3-phenylpyruvate and t.-norleucine than in the simultaneous presence of 3-phenylpyruvate and L-valine, as already observed in a prior series of experiments [16]. However, under these conditions, the rate of insulin release does not depend solely on the fate of 3-phenylpyruvate, but also on that of L-valine or L-norleucine.
Our data clearly indicate that 3-phenylpyruvate dramatically affected the metabolism of these amino acids. In the case of D,L-[I-14C] norleucine, the production of x4CO2 was impaired by 3-phenylpyruvate, despite a sizeable increase in the rate of conversion of norleucine to 2-ketocaproate. By analogy with the effect of 3-phenylpyruvate upon the oxidation of 2-[U-~4C]-ketoisocaproate [2], this supports the view that the direct or indirect transamination between L-norleucine and 3-phenylpyruvate occurs mainly in the cytosol and, hence, coincides with a severe inhibition by 3-phenylpyruvate of the transport of 2-ketocaproate, derived from norleucine, into the mitochondria. In the absence of 3-phenylpyruvate, L-norleucine fails to stimulate insulin release [22], suggesting that the catabolism of this amino acid and, if any, its influence upon the catabolism of endogenous nutrients arc not sufficient to increase 02 consumption above the threshold value required for stimulation of insulin release [23]. Since 3-phenylpyruvate further decreased the oxidation of norleucine, it is not surprizing that the catabolism of this amino acid was not sufficient to compensate for its inhibitory action upon the metabolic response to the pyruvate analogue. In
contrast
to
its
inhibitory
effect
on
O,L-[I-14C]norleucine oxidation, 3-phenylpyruvate increased the production of m4CO2 from L-[l-14C] valine. It is conceivable, therefore, that an increased generation of reducing equivalents at the level of the reaction catalyzed by the branched chain 2-keto acid dehydrogenase compensated, to a limited extent, for the vaiine-induced alteration of the metabolic response to 3-phenylpyruvate. Incidentally, the relative magnitude of the 3-phenylpyruvate-induced increment in ~4COe production from L-[l-~4C]valine was much less marked than would be expected from the relative magnitude of the increase in the rate of conversion of L-[l-14C] valine to 2-ketoisovalerate (Fig. 1). This again suggests that 3-phenylpyruvate inhibits the transport of 2-keto acids into the mitochondria. The comparison of the data collected with L-[1-14C] valine and L-[U-m4C]valine, respectively, also revealed that 3-phenylpyruvate impairs the further oxidative metabolism of isobutyryl-CoA, relative to its rate of generation. This could be due, inter alia, to both the increased catabolism of endogenous amino acids and the induction of a more reduced redox state in the islets exposed to 3-phenylpyruvate [3]. Our data indicate that, especially in the presence of 3-phenylpyruvate, a significant amount of acidic metabolites generated from
3 59
3-phenylpyruvate-induced insulin release
isobutyrvl-Coa escaped full oxidation. By analogy with the situation recently identified in isolated hepatocytes, this would be compatible with the accumulation of 3-hydroxyisobutyrate [20].
that this drug failed to reverse the inhibition of 3-phenylpyruvate-stimulated insulin release by L-norleucine. Conclusions
Inhibitory effect o f a m i n o o x y a c e t a t e
In the islets, aminooxyacetate inhibits extramitochondrial transamination reactions [8]. At first glance, the fact that aminooxyacetate abolished 3-phenylpyruvate-induced insulin release could suggest, therefore, that such extramitochondrial transamination reactions represent an essential determinant of the secretory response to the pyruvate analogue. This interpretation, however, is invalidated by two series of observations. First, in the absence of exogenous nutrients, aminooxyacetate fails to affect the basal output of NH~ [24], indicating that the transamination of endogenous amino acids takes place in the mitochondria. Likewise, aminooxyacetate failed to affect ~4CO2 output from islets preincubated with I.-[U-~4C] glutamine and incubated in the absence of exogenous nutrient. Second, in the islets prelabelled with L-[U-~4C]glutamine, aminooxyacetate impaired, but failed to abolish, the enhancing action of 3-phenylpyruvate upon ~4CO2 output. This again indicates that the facilitating action of 3-phenylpyruvate upon the metabolism of endogenous amino acids is not solely a cytosolic process. Moreover, even the partial inhibition by aminooxyacetate of 3-phenylpyruvatestimulated ~4CO2 output from the prelabelled islets does not necessarily indicate that the pyruW I L I I glutamate ....... •,,~,~ is engaged, together __..,L (derived from exogenous !-[U-]4C] glutamine, [see ref. 25]), in a cytosolic transamination process. Indeed, the full oxidation of I-[U-~4C] glutamine is dependent on a transfer of malate from the mitochondria to the cytosol [25, 26] and the latter process may itself be impaired in islets exposed to aminooxyacetate [8]. As an alternative explanation, we believe that the inhibitory effect of aminooxyacetate upon insulin release induced by 3-phenylpyruvate could be due to impairment of the malate-aspartate shuttle involved in the intracellular transfer of reducing equivalents between the mitochondrial and cytosolic compartments. This view is consistent with the fact that aminooxyacetate impairs the ionic and secretory response to such nutrient secretagogues as 2-ketoisocaproate, ew:n though the oxidative metabolism of 2-ketoisocaproate is unaffected by the transamination inhibitor [27, 28]. The proposed mode of action of aminooxyacetate would also account for the fact ~,,'~t~
,~...
n
!,-.
. . . .
"
In conclusion, the present data provide further support to the view that the mitochondrial oxidative catabolism of endogenous amino acids and the transfer of reducing equivalents to the cyto~oi represent two essential biochemical determiuants of the secretory response of pancreatic islets to 3-phenylpyruvate. The present work also iUustrates the fact that two distinct nutrients (e.g. 3-phenylpyruvate and L-norleucine), which are each able to stimulate insulin release under suitable experimental conditions [2, 22], may act antagonistically to one another when used in combination. In this respect, the situation encountered in islets exposed to both 3-phenylpyruvate and Lnorleucine (or L-valine) is reminiscent of that previously characterized in islets exposed to both 2-ketoisocaproate and a branched chain amino acid [12]. Such converging observations do not solely emphasize the concept that the regulation of oxidative events represents the fundamental mechanism by which the B-cell identifies circulating nutrients as insulin secretagogues [29, 30]. These observations also indicate that due attention should now be paid to the metabolic interaction between distinct substrates (glucose, amino acids, fatty acids, ketone bodies, lactate and pyruvate) in order to gain a better understanding of the regulation of insulin release by the heterogenous constellation of circulating nutrients.
Acknowledgements This work was supported by grants from the Belgian Foundation for Scientific Medical Research and Belgian Ministry of Scientific Policy. The authors wish to thank J. Schoonheydt, A. Tinant and M. Urbain for technical assistance and C. Demesmaeker for secretarial help. We are most grateful to Drs. S. Lenzen, H. Formanek and U. Panten for the generous gift Of D.L-[I-14C]norleucine.
REFERENCES I. Panten, U. & Langer, J. (1981) Biochem. J., 198, 353-~56. 2. Sener, A., Welsh, M., Lebrun, P., Garcia-Morales, P., Saceda, M., Malaisse-Lagae, F., Herchuelz, A., Valverde, I., Hellerstr6m, C. & Malaisse, W.J. (1983) Biochem. J., 210, 913-919.
360
A. Sener and W.J. Malaisse
3. Malaisse, W.J., Sener, A., Welsh, M., MalaisseLagae, F., Hellerstr6m, C. & Christophe, J. (1983) Biochem. J., 210, 921-927. 4. Hutton, J.C., Sener, A., Herchuelz, A., Atwater, I., Kawazu, S., Boschero, A.C., Somers, G., Devis, G. & Malaisse, W.J. (1980) Endocrinology. 106, 203-219. 5. Sener, A., Hutton, J.C. & Malaisse, W.J. (1981) Biochim. Biophys. Acta, 677, 32-38. 6. Hutton, J.C., Atwater, I. & Malaisse, W.J. (1980) Horm. Metab. Res., suppl, series 10, 31-37. 7. Sener, A. & Malaisse, W.J. (1979) Eur. J. Biochem., 98, 141-147. 8. Malaisse, W.J., Malaisse-Lagae, F. & Sener, A. (! 982) Endocrinology. 111, 392-397. 9. Corneil, N.W., Crow, K.E. & Witefoot, R.P. (1981) Biochem. J., 198, 219-223. 10. Malaisse, W.J., Brisson, G. & Malaisse-Lagae, F. (1970) J. Lab. Clin. Med.o 76, 895-902. 11. Carpinelli, A.R., Sener, A., Herchuelz, A. & Malaisse, W.J. (1980) Metabolism, 29, 540-545. 12. Hutton, J.C., Sener, A. & Malaisse, W.J. (1980) J. Biol. Chem., 255, 7340-7346. 13. Malaisse, W.J., Sener, A., Malaisse-Lagae, F., Hutton, J.C. & Christophe, J. (1981) Biochim. Biophys. Acta. 677, 39-49. 14. Lenzen, S., Formanek, H. & Panten, U. (1982) J. Biol. Chem., 257, 6631-6633. 15. Sener, A., Malaisse-Lagae, & Malaisse, W.J. (1981) Proc. Natl. Acad. Sci. USA, 78, 5460-5464. 16. Sener, A., Malaisse-Lagae, F. & Malaisse, W.J. (1983) Acta diabetol, lat., 20, 205-209.
17. Sener, A., Owen, A., Malaisse-Lagae, F. & Malaisse, W.J. (1982) Horm. Metab. Res~, !4, 405 -409. 18. Sener, A. & Malaisse, W.J. (1980) Nature, 288, 187-189. 19. Malaisse, W.J., Sener, A., Malaisse-Lagae, F., Welsh, M., Matthews, D.E., Bier, D.M. & Hdlerstr/~m, C. (1982) J. Biol. Chem., 257, 8731-8737. 20. Corkey, B.E., Martin-Requero, A., Walajtys-Rode, E., Williams, R.J. & Williamson, J.R. (1982) J. Biol. Chem., 257, 9668-9676. 21. Malaisse, W.J., Malaisse-Lagae, F. & Sener, A. (1984) Biochim. Biophys. Acta, 797, 194-202. 22. Sener, A., Somers, G., Devis, G. & Malaisse, W.J. (1981) Diabetologia, 21, 135-142. 23. Hutton, J.C. & Malaisse, W.J. (1980) Diabetologia, 18, 395-405. 24. Sener, A., Best, L., Malaisse-Lagae, F. & Malaisse, W.J. (1984) Arch. Biochem. Biophys., 229, 155-169. 25. Malaisse, W.J., Sener, A., Carpinelli, A.R., Anjaneyulu, K., Lebrun, P., Herchuelz, A. & Christophe, J. (1980) Mol. Cell. Endocrinol., 20, 171 - 189. 26. Sener, A., Malaisse-Lagae, F. & Malaisse, W.J. (! 982) Biochem. J., 202, 309-3 ! 6. 27. Hutton, J.C., Sener, A. & Malaisse, W.J. (1979) Biochem. J.. 184, 291-301. 28. Lebrun, P., Malaisse, W.J. & Herchuelz, A. (1983) Am. J. Physiol., 245, E38-E46. 29. Malaisse, W.J., Sener, A., Herchuelz, A. & Hutton, J.C. (1979) Metabolism, 28, 373-386. 30. Malaisse, W.J. (I 983) Diab. M~tab., 9, 313-320.