Eur. J. Biochem. 213,461 -466 (1993) 0 FEBS 1993

Characteristics and specificity of the inhibition of liver glucose-6-phosphatase by arachidonic acid Lesser inhibitability of the enzyme of diabetic rats

Gilles MITHIEUX, Jean-Claude BORDETO, Carol MINASSIAN, Ahmed AJZANNAY, Isabelle MERCER and Jean-Paul RIOU Institut National de la Sand et de la Recherche MCdicale, Unit6 197 and 331", FacultC de MCdecine A. Carrel, Lyon, France

(Received October 5, 1992/January 8, 1993) - EJB 92 1399

The effect of arachidonic acid (A,Ach) on liver glucose-6-phosphatase (Glc6Pase) has been studied in v i m using untreated and detergent-treated microsomes prepared from fed and 48-h-fasted normal rats and from streptozotocin-induced diabetic rats. Glc6Pase of both untreated and detergent- treated microsomes (60 pg proteidml) is inhibited by d,Ach in a dose-dependent manner between 10-100 pM. The inhibition is very rapid and does not depend on preincubation of microsomes in the presence of A,Ach. It does depend on the concentration of microsomal membranes and on the concentration of glucose 6-phosphate: it is more pronounced at low Glc6P concentrations than at high. As a consequence, the enzyme displays sigmoidal kinetics in the presence of A,Ach. Hill coefficients (equal to 1 in the control experiments) of about 1.4 were determined in the presence of 50 pM d,Ach, indicating a clear positive cooperative dependency of the Glc6Pase upon its substrate in the presence of A,Ach. The A,Ach inhibition is fully reversible in the presence of bovine serum albumin. The inhibition does not depend on the metabolism of d,Ach through the prostaglandin synthase (cyclooxygenase) or arachidonate 12-lipoxygenase pathways since it is not affected by indomethacin and nordihydroguaiaretic acid. Several other unsaturated fatty acids are able to inhibit the enzyme within the same concentration range. In contrast, saturated fatty acids, the arachidonic acid methyl ester and numerous other lipid compounds containing esterified unsaturated fatty acids do not inhibit Glc6Pase within the same concentration range. The enzyme of fed rats was inhibited in the same manner as the enzyme of 48-h-fasted rats. However, Glc6Pase of untreated microsomes from diabetic rats was less inhibitable by d,Ach than the Glc6Pase of normal rats. This difference does not persist after solubilization of the membrane lipids by detergent treatment.

Liver glucose-6-phosphatase (Glc6Pase) is the last en- zyme common to both gluconeogenesis and glycogenolysis. Unlike the other important enzymes of both biochemical pathways, it is an integral membrane protein located in the endoplasmic reticulum. This particular location within the hepatocyte remains poorly understood to date. The marked alterations of specificity and of velocity of the enzyme upon membrane solubilization by the action of detergents gave rise to two types of interpretation (see [l] for a review). One of them refers to a compartmentation hypothesis based on the assumption that Glc6Pase is located at the luminal surface of the endoplasmic membrane and has free access to its sub- strate only upon detergent-induced removal of the membrane barrier [2, 31. The different features of the enzyme within the intact membrane are, in this case, believed to be attributable

Correspondence to G. Mithieux, INSERM U. 197, Facult6 de Mkdecine A. Carrel, rue G. Paradin, F-69372 Lyon C6dex 08, France

Fax: +33 78 77 86 12. Abbreviations. A,Ach, arachidonic acid; IC,,, 50% inhibiting

concentration ; GlcdPase, glucose-6-phosphatase ; Man6Pase, man- nose-6-phosphatase; PG, prostaglandins.

Enzymes. Glucose-6-phosphatase (EC 3.1.3.9) ; cyclooxygenase, prostaglandin synthase (EC 1.14.99.1); lipoxygenase, arachidonate 12-lipoxygenase (EC 1.1 3.11.31).

to a rate-limiting element responsible for the specific trans- port of glucose 6-phosphate (Glc6P) inside the lumen. The alternative hypothesis that the alterations of the Glc6Pase activity upon detergent treatment are only due to confor- mational changes of the structure of a unique protein recently received biochemical supports [4-71. A consequence of the latter idea is that Glc6Pase activity could be dynamically regulated by its membranous environment. That the Glc6Pase activity of untreated microsomes can be influenced by direct modulation of the membrane microviscosity [8- 111 seems to be in agreement with this hypothesis but could be interpreted in terms of alteration of the activity of a Glc6P translocator as well [8].

The importance of the membrane lipids on the Glc6Pase activity has been pointed out by numerous studies related to the phospholipid dependency of the enzyme activity. This problem has generally been adressed by the mean of cycles of removalkupplementation of membrane lipids [12, 131, through the action of various phospholipases [14-191, or using phospholipid transfer proteins [17, 201. A direct effect of the fatty acid products of phospholipase A on the Glc6Pase activity could be inferred from some of those stud- ies since the presence of bovine serum albumin appeared able to protect Glc6Pase against the inhibitory effects induced by

462

- 0 4 r g r t 0.3 .-

iE - c Y i? Q u) 3. "': p" 0.1

8 5

0

phospholipase A action [14, 151. However, studies reporting effects of some fatty acids on Glc6Pase activity are scarce and somewhat contradictory [21, 221.

In order to understand better the modulating role of the membrane lipids and of their constituents on the activity of Glc6Pase, we studied the effect of the presence of free fatty acids on the enzyme activity. We attempted to elucidate the mechanism of inhibition highlighted using arachidonic acid (d,Ach) as a model fatty acid. In addition, we compared the d,Ach-induced inhibition of the enzyme obtained from fed, fasted and diabetic animals.

u I1 ' I ' - * . . ' '

I fi: !\ P

4,q+ 4 -

\>-. . f f ' I . . . MATERIALS AND METHODS Materials

All chemicals were obtained from Sigma (La Verpilliere, France). Unsaturated fatty acids, prostaglandins (PG), plate- let-activating factor, lysophospholipids, phospholipids, ara- chidonic methyl ester, indomethacin and nordihydroguaiar- etic acid were stored in ethanol at -80°C at a concentration of 10 mM. Saturated fatty acids and diacylglycerols were stored in dimethylsulfoxide at the same concentration. Bo- vine serum albumin (fatty-acid-free) was obtained from Boehringer (Meylan, France).

Animals Male Sprague-Dawley (OFA) rats weighing 200-250 g

were used. Rats were fed laboratory chow or fasted for 48 h before preparation of microsomes. Rats became diabetic by a single intraperitoneal injection of streptozotocin (75 mgkg) and were used three days later for processing of the liver (mean glycemia * SD: 28.3 -C 5.2 mM, n = 7).

Preparation of microsomes Microsomes were prepared essentially as previously de-

scribed [23, 241. The abdomen of rats killed by decapitation was immediately incised and the liver rinsed with cold 0.15 M NaCI. The liver was removed, homogenized in 0.25 M sucrose, 10 mM Hepes pH 7.4 (9 voUg fresh mass) using a Teflon/glass homogenizer and centrifuged at 30000Xg for 10 min at 4°C. The supernatant was centri- fuged at lOOOOOXg for 1 h at 4°C. Pelleted microsomes were suspended in the same buffer and washed once under the same centrifugation conditions. The term 'untreated microsomes' refers to microsomes suspended in the homog- enization buffer at a final concentration of 15 -20 mg pro- teidml without further treatment. The term 'detergent-treated microsomes' refers to microsomes which were further treated in the presence of 0.5% (mass/vol.) sodium cholate for 20 min at 4°C.

Glc6Pase assay The GlcdPase activity was assayed in the presence of

20mM Tris/HC1 pH7.3 for 10min at 37°C. P, was deter- mined by a complexometric method [25] after protein pre- cipitation by addition of ascorbic acid/trichloroacetic acid (2%/10%; mass/vol.). Most of the experiments (with the ex- ception of those reported in Table 2) were performed with microsomes from 48-h-fasted rats. Unless otherwise indi- cated, Glc6P and lipid effectors were diluted in the reaction medium prior to the microsomal preparation. After 2 min of

temperature equilibration at 37 "C, microsomes were added to start the reaction. The kinetics of the enzyme of untreated and detergent-treated microsomes from fed, 48-h-fasted and diabetic rats were studied in the presence of 0.5-2OmM Glc6P. The latency of mannose-6 phosphatase (Man6Pase) activity of untreated microsomes was determined at a man- nose-6-phosphate concentration of 1 mM and calculated ac- cording to the formula : 100 - [ lOO(Man6Pase of untreated microsomes)/(Man6Pase of detergent-treated microsomes)]. Values of 92 5 2%, 94? 3 % and 93 t 1 % (mean ? SD) were found for microsomes of fed (n = 5) , 48-h-fasted ( n = 6) and diabetic (n = 7) rats, respectively.

RESULTS Characteristics of the inhibition of Glc6Pase by arachidonic acid

The presence of d4Ach (10-100 pM) in the incubation medium markedly altered the activity of Glc6Pase of un- treated microsomes (60 pg proteidml, 1 mM Glc6P) in a dose-dependent manner (Fig. 1). Inhibition did not take place at d,Ach concentrations equal to or lower than 10pM. A 50 % inhibition effect was observed at d4Ach concentrations ranging between 20-30 pM. Less than 10% of the Glc6Pase activity remained measurable at concentrations of A,Ach equal to and above 50 pM. Upon detergent treatment, the Glc6Pase activity was increased by 150% at a Glc6P con- centration of 1 mM. The enzyme activity of detergent-treated microsomes was inhibited by &Ach in the same concen- tration range with similar features (Fig. 1).

The inhibition was not dependent upon preincubation of microsomes with A,Ach. The inhibitory effect (studied under the conditions above) was the same whether microsomes had been preincubated in the presence of d,Ach for 15 min at 20°C or 5 min at 37°C or not at all. The preincubation of microsomes with a non-inhibiting concentration of d,Ach (10 pM) for a longer time (30 min) at 20°C did not give rise

463

I I ' I

6ot O 1

A Q

0

0 I I I

0 0.05 0.10 0.15

Microsomal Protein (mg/ml)

Fig. 2. Dependence of the inhibition of Glc6Pase by arachidonic acid upon the microsome concentration. The IC,, was determined from dose/response curves similar to those shown in Fig. 1 carried out with varying concentrations of untreated microsomes. The re- lationship between the two variables is described by the linear equation (1Csn = 232.9 X microsomal protein + 10.4, R = 0.893, P < The symbols 0, 0, ., 0, A and A correspond to differ- ent microsomal preparations from 48-h-fasted rats.

to a measurable inhibition of Glc6Pase. The inhibition was very rapid which took place immediately and which was not further dependent on time. The Glc6Pase velocity of un- treated microsomes was a linear function of time within a 15-min period at all A,Ach concentrations up to 50 pM (vir- tually no activity was measurable in the presence of 100 pM A,Ach). This was also true for the enzyme of detergent- treated microsomes at A,Ach concentrations lower than 30 pM. However, a slight deviation (about 20%) in linearity of the velocity/time plot occurred after 5 min in the presence of 50 pM A,Ach; this was probably related to the decreased thermostability of the enzyme in the presence of detergent

Since biological membranes possess several classes of binding sites for free fatty acids [27, 281 so we studied the dependency of the inhibition on the amount of membranes present in the incubation medium. The concentration of A,Ach inducing a 50% inhibitory effect (1C5,J was deter- mined from dose/response studies (analogous to that shown in Fig. 1) performed with different microsomal preparations, each of them being present at various concentrations. The higher the membrane concentration, the higher the IC,, deter- mined, i.e. the higher the A,Ach concentration required to obtain the same inhibitory effect of Glc6Pase (Fig. 2). This result was expected because of the probable trapping of A,Ach by the membranous binding sites which is likely to decrease the actual concentration of A,Ach available for inhi- bition of Glc6Pase. Consequently, to allow valuable com- parisons, all further experiments involved the same micro- some concentration as that used in the experiments shown in Fig. 1.

The inhibitory effect of A,Ach was more pronounced at low concentrations of substrates than at high. As a conse- quence, the kinetics of Glc6Pase of untreated microsomes (velocity/substrate concentration), which displayed a hyper- bolic shape in the absence of A,Ach, tended to exhibit a sigmoidal shape in the presence of d,Ach (Fig. 3A). This could not be due to non-linearity of the Glc6Paseltime at low concentrations of Glc6P (see paragraph above). The inhi- bition mechanism was thus characterized by to the Hill rep- resentation. Hill plots were linear for both the control ki-

[261.

A

5 10 15 20

Glc-6 P (mM)

I 1 4

- -6 0 1 2 3 Ln ([Glc-6 PIIM)

Fig. 3. Kinetics of Glc6Pase in the presence of arachidonic acid. The kinetics of Glc6Pase of untreated microsomes (60 pg/ml) from fasted rats were determined in the absence (0) and in the presence of 50 pM (0) A,Ach, as described in Materials and Methods. (A) Direct-coordinate plot (v versus [Glc6P]); (B) Hill plot, (log[(u/ V,,-v) versus log ([Glc6P]/M), of the same data. The slope of the lines, representing the Hill coefficients, were (0) 0.998 (control experiment) and (0) 1.369 (50 pM d4Ach). The results shown are representative of three experiments carried out under the same con- ditions.

netics obtained in the absence of A,Ach and the sigmoidal kinetics obtained in the presence of 50 pM A,Ach (Fig. 3B). The slope of the Hill plots of control kinetics was very close to 1, whereas that of the kinetics obtained in the presence of 50 pM d,Ach was very close to 1.4 (1.36-1.47), showing that Glc6Pase exhibited a clear positive cooperativity for its substrate in the presence of A,Ach. The kinetics of Glc6Pase of detergent-treated microsomes also exhibited a sigmoidal shape in the presence of 50 pM A,Ach (data not shown).

We have taken advantage of the capacity of bovine serum albumin to bind long-chain fatty acids with high affinity and capacity [29] to study the reversibility of the inhibition of Glc6Pase by A,Ach. The following experiments involved a preincubation period of microsomes for 15 min at 20°C un- der various conditions and the initiation of the enzymatic reaction upon addition of the substrate. As mentioned above, after a 15-min preincubation in the presence of A,Ach, Glc6Pase was inhibited in the same manner as when it was not preincubated (compare Fig. 1 and Fig. 4). When preincu- bated in the presence of bovine serum albumin (fatty-acid- free) only, the activity of Glc6Pase was slightly enhanced. The presence of albumin during preincubation completely prevented the inhibitory effect of the ensuing added A,Ach. The addition of bovine serum albumin at the end of a 15-min preincubation in the presence of A,Ach also allowed the full activity of Glc6Pase to be recovered (Fig. 4A). A slight activation effect was even observed at the highest A,Ach concentration tested (50 pM). To test the possibility that this activation effect was due to a partial release of the enzyme latency, we tested, in parallel, the Man6Pase activity of microsomes incubated under the same conditions. Man& Pase was inhibited by A,Ach in the same manner as Glc6Pase. The presence of bovine serum albumin also in- duced a slight activation effect of Man6Pase and totally can- celled the inhibitory action of the ensuing added A,Ach. When the albumin was added at the end of the 15-min prein- cubation in the presence of A,Ach, the inhibition was can- celled. An activation effect, which accounted for a loss of about 30% of the Man6Pase latency, was indeed observed at the highest A,Ach concentration tested (50 pM; Fig. 4B). Man6Pase and Glc6Pase of detergent-treated microsomes

464

Arachidonlc acid (@M)

Fig.4. Reversibility study of the inhibition of Glc6Pase by ara- chidonic acid. Glc6Pase (A) and Man6Pase (B) of untreated micro- somes were assayed as described in Fig. 1 in the presence of 1 mM substrate in each case, involving various preincubation conditions described in detail in Results. (0) 15-min preincubation with A,Ach without bovine serum albumin; (0) 15-min preincubation in the presence of bovine serum albumin (1 mg/ml) prior to addition of A,Ach; (H) 15-min preincubation with A,Ach prior to addition of bovine serum albumin. The Glc6Pase activity and the Man6Pase activity of detergent-treated microsomes assayed in the presence of bovine serum albumin were 0.31 and 0.24 pmol . min-’ . mg pro- tein-’, respectively. The results are the mean ? SD of two experi- ments performed under the same conditions (when not given, the SD are within the size of the symbols).

were inhibited in the same manner by d,Ach. Bovine serum albumin was able to reverse the d,Ach-induced inhibition of both (data not shown).

Specificity of the inhibition of Glc6Pase by arachidonic acid

The prior metabolization of d,Ach through the prosta- glandin synthase (cyclooxygenase) or the arachidonate 12- lipoxygenase pathways was not involved in the inhibition of Glc6Pase since the d,Ach-induced inhibition of Glc6Pase was not altered in the presence of indomethacin (10 pM) and nordihydroguaiaretic acid (10 pM), respective inhibitors of both pathways. We tested the effect of the presence of numer- ous compounds (at the same concentrations as for A,Ach up to a concentration of 100 pM) on the Glc6Pase activity measured under the same conditions as with d,Ach (con- ditions of Fig. 1). Several other unsaturated fatty acids were able to inhibit the Glc6Pase activity in the same concen- tration range as d,Ach: they belong to the n-3, n-6 or n-9 series (Table 1). In contrast, saturated fatty acids (G6 o, C,, o, C,, ,,), arachidonic acid methyl ester, l-a-phosphatidyl- cholines [ 1,2-dilinoleoyl glycerophosphocholine and p-ace- tyl-y-0-alkylglycerophosphocholine (platelet-activating fac- tor)], diacylglycerols (1,2-dipalmitoyl-sn-glycerol, 1,2-di- oleoyl-sn-glycerol, 1-stearoyl-2-arachidonoyl-sn-glycerol) prostaglandins [PGE,, PGE,, PGF,, and (11,9)-epoxymeth- anol-PGF,, (U46619, a stable analog of thromboxane] were without effect on the Glc6Pase activity. Lysophospholipids (2-stearoylglycerophosphocholine, 2-oleoyl-glycerophospho- choline) had no effect on the Glc6Pase activity at a concen- tration of 10 pM and exerted a slight activation effect (about 50%) which peaked at 20-30 pM and which was lowered (20%) at 50 and 100 pM (data not shown).

Influence of nutrition and of diabetes on the inhibition of Glc6Pase by arachidonic acid

We also determined the IC,, of d,Ach for the Glc6Pase of untreated and detergent-treated microsomes from fed rats

Table 1. Inhibition of Glc6Pase by unsaturated fatty acids. Glc6Pase of untreated microsomes was assayed as described in Fig. 1 in the presence of various unsaturated fatty acids (10- 100 pM). IC,, was determined from dosehesponse curves analogous to that of Fig. 1. The results are expressed as the mean ? SD of the number of experiments indicated in brackets.

Unsaturated Ic50 fatty acid

PM c,,:3. “-3

C Z Z : , “-3

cZ0:3, n--6

CZ,:, n-6

C Z Z : , “-6

31.5 ? 2.1 (3) 15.8 I l . 1 (3) 20.3 2 1.8 (3) 18.5 ? 2.1 (3) 26.5 ? 4.9 (3) 25.7 % 3.6 (10) 19.8 5 1.1 (3) 24.2 ? 1.1 (3)

Czo5, “ - 3

c18:2, rr--6

c,*:1, “-9

Table 2. Effect of fasting and diabetes on the inhibition of Glc6Pase by arachidonic acid. IC,, was obtained from dosehe- sponse curves obtained as described in Fig. 1. The results are ex- pressed as the mean ? SD (n in brackets). The data were analyzed for statistical significance of differences using the I test for unpaired data. The symbol * means that the value is significantly different from the values obtained for untreated microsomes of fed rats (P<O.OOl) and fasted rats (P<O.O01) and for detergent-treated microsomes of diabetic rats ( P < 0.001).

Rats Icm

untreated detergent-treated microsomes microsomes

Fed 25 -t 1.4 (5) 28.5 2 3 ( 5 ) Fasted 25.7 2 3.6 (10) 25 % 2 (6) Diabetic 38 2 2.7* (7) 29.7 ? 3.5 (7)

and from streptozotocin-diabetic rats. The enzyme from fed rats was inhibited by d,Ach in the same manner as the en- zyme of fasted rats, whether microsomes had been treated by detergent or not (Table 2). In contrast, the Glc6Pase of un- treated microsomes from diabetic rats was significantly less inhibitable by d,Ach (K5, of d,Ach close to 40 pM), whereas the enzyme of detergent-treated microsomes from the same rats exhibited the same inhibitability as the enzyme of normal fed and fasted rats (Table 2). The kinetics of the enzyme of normal fed and diabetic rats in the presence of 50 pM d,Ach exhibited a sigmoidal shape of the same type as that shown in Fig. 3 (data not shown).

DISCUSSION In this paper, we have reported that Glc6Pase is inhibited

in vitro in the presence of low concentrations of A,Ach and of other unsaturated fatty acids. The mechanism of inhibition has been more thoroughly investigated with d,Ach as a mo- del because of its quantitative importance in the liver micro- soma1 membranes [30, 311. In addition, we have shown that Glc6Pase of diabetic animals is less sensitive to this inhi- bition than the enzyme of normal animals.

465

To our knowledge, the inhibition of Glc6Pase by A,Ach had not been reported hitherto. The inhibition of Glc6Pase by some free fatty acids has been reported in previous studies [21,22]. However, some differences exist between the results previously reported and ours. Palmitate was found to inhibit the enzyme of untreated microsomes as did oleate, linoleate and linolenate [21], whereas we did not observe any effect of saturated fatty acids (including palmitate) in the range of concentrations studied. Linoleate and linolenate, but not satu- rated fatty acids, were reported to inhibit Glc6Pase of un- treated microsomes by 40% (maximal effect) in the range of concentrations studied here, albeit linolenate exhibited a biphasic effect (maximal effect at 15 pM, no effect at 20 pM and above). In addition, both unsaturated fatty acids were reported not to inhibit the enzyme of detergent-treated micro- somes [22]. Some of these differences could be due to the assay conditions. This could be especially important since we have shown here that the inhibition is dependent on both the microsome and the substrate concentrations used (in [22], microsomes and Glc6P were present at concentrations largely higher than in our study: 1-1.5 mg proteidml and 40 mM, respectively).

A,Ach is able to inhibit Glc6Pase activity in a rapid, dose-dependent and reversible manner. The reversibility has been addressed by using bovine serum albumin (fatty-acid- free) which exhibits several high-affinity binding sites for free fatty acids [29]. It is worth noting that bovine serum albumin is able, when added alone, to activate the activity of the enzyme. This could be due to the removal of inhibitory unsaturated fatty acids previously bound to Glc6Pase.

The interesting question which arises is whether A,Ach actually is the inhibitory molecule or if its prior metabo- lization through specific oxidative pathways is required to induce the inhibition of Glc6Pase. The hypothesis that the processing of A,Ach through either the cyclooxygenase or the lipoxygenase pathways is required for inhibition of Glc6Pase is unlikely since the presence of indomethacin and nordihydroguaiaretic acid, respective inhibitors of both path- ways, does not hinder nor diminish the A,Ach effect. In ad- dition, the preincubation of A,Ach in the presence of micro- somes has no effect on the subsequent inhibition (or absence of inhibition) of Glc6Pase by A,Ach. At least, some terminal products resulting from oxygenation of A,Ach (PG, U46619) do not have any effect on Glc6Pase activity.

An important question is whether the inhibition of Glc6Pase by A,Ach (or unsaturated fatty acids in general) involves the direct binding of A,Ach to the enzyme or to a tightly bound lipid moiety or is mediated through some modifications of the membrane structural organization con- secutive to the insertion at random of these fatty acids within the membrane. The latter possibility appears rather unlikely since stearic acid, a saturated fatty acid which has no effect on GlcGPase, binds efficiently to the microsomal membrane, as does linoleic acid, an unsaturated fatty acid which is one of the most powerful inhibitors of the enzyme [22]. In ad- dition, it is known that the higher the proportion of unsatu- rated fatty acids within a membrane, the lower the membrane microviscosity [ 111. Lowering the membrane microviscosity enhances the Glc6Pase activity [8]. As a consequence, the insertion of an unsaturated fatty acid within the microsomal membrane should promote the enhancement and not the inhi- bition of the Glc6Pase activity. Finally, the strongest argu- ment that the inhibition of microsomal Glc6Pase by d,Ach is not mediated through an indirect effect linked to an alter- ation of the membrane taken as a whole is that the enzyme

extracted from the membrane by detergent treatment is also inhibited by A,Ach. The mechanism of inhibition could in- volve either the direct binding of A,Ach to the enzyme or the binding of A,Ach to a lipid molecule which could remain tightly bound to the enzyme even after detergent treatment. A possibility according to the latter mechanism could be the displacement by A,Ach of a tightly bound lipid essential for Glc6Pase function from its site.

The minimal structural features required for the inhi- bition of Glc6Pase can be estimated by comparing the re- spective effects of the numerous compounds we tested. The presence of at least one double bond in the molecule is re- quired: C,, n-9 is able to inhibit Glc6Pase fully. Another feature required for the inhibitory effect to take place is that the terminal carboxylic function is present as a free acid. The esterification of the carboxylic function by a methyl group hinders the inhibition: methyl arachidonate does not inhibit Glc6Pase. Numerous other compounds are not able to inhibit the enzyme activity although they contain an esterified un- saturated fatty acid (see Results).

The estimation of the apparent inhibition constant of A,Ach for Glc6Pase seems difficult from our data. The con- centrations studied here are probably very different from the concentrations of A,Ach which are actually available for the inhibition of the enzyme because of the conjunction of two phenomena: (a) the trapping of fatty acids by the membra- nous binding sites [27, 281 which might account for the re- sults reported in Fig. 2; (b) the natural tendency of free fatty acids to assemble partially into micellar structures in aqueous phase. The actual concentration of free A,Ach inducing 50% inhibition of the enzyme is thus probably lower than the IC,, determined here.

Another important question concerns the possible site of inhibition of A,Ach with respect to the structural organi- zation of Glc6Pase. It is noteworthy that the IC,, of A,Ach for Glc6Pase of untreated microsomes is very close to that of the detergent-treated microsomes (with the exception of the Glc6Pase of diabetic animals). If we consider that the Glc6Pase activity results from the combined action of a rate- limiting Glc6P translocator and a distinct lumenal phos- phohydrolase referring to the compartmentation hypothesis [2, 31, we should consider that A,Ach inhibits the activity of the phosphohydrolase since d,Ach inhibits the Glc6Pase activity after detergent treatment of microsomes, conditions where the Glc6P translocator is shunted. To explain the A,Ach-induced inhibition of Glc6Pase in the intact membra- nous system, we have to postulate (a) that A,Ach is able to insert inside the membrane deeply enough to reach and in- hibit the phosphohydrolase and that the resulting inhibitory action of A,Ach on the whole functionning system in the intact membrane is very similar to that induced on the sole phosphohydrolase after membrane disruption by detergent ; alternatively (b) that A,Ach is able to inhibit both the rate- limiting Glc6P translocator (accounting for the inhibition of Glc6Pase in the intact system) and the phosphohydrolase (ac- counting for the inhibition of the enzyme in the detergent- disrupted system) in an analogous manner. Referring to the conformational hypothesis [4 -71, we should consider that, in spite of the conformational alteration following removal of the membrane lipids by detergent, Glc6Pase retains the same A,Ach-induced inhibition characteristics.

The last point which we would like to emphasize refers to the observed differences between the inhibition by A,Ach of the enzyme from diabetic animals and that from normal animals. It is worth noting that Glc6Pase of untreated micro-

466

somes of diabetic animals, but not that of detergent-treated microsomes, is less inhibitable by d,Ach than the enzyme of normal animals. That this difference could be due to post- translational modifications of the enzyme seems unlikely, since one might expect in this case that it would be retained after removal of the membrane by detergent. The other hy- pothesis is that it is caused by altered enzyme/membrane in- teractions which could be due, for example, to modifications of the lipid content of liver microsomes known to be associ- ated with the diabetic onset [32-361. This hypothesis seems in good agreement with the fact that this different in- hibitability is not retained after removal of membrane lipids by detergent treatment.

REFERENCES 1.

2.

3. 4.

5.

6.

7.

8.

9.

10. 11.

12.

13.

14.

Sukalsky, K. A. & Nordlie, R. C. (1988) Adv. Enzymol. 62,93-

Arion, W. J., Wallin, B. K., Lange, A. J. & Ballas, A. M. (1 975)

Burchell, A. (1990) FASEB J. 4, 2978-2988. Speth, M. & Schulze, H. U. (1988) Eul: J. Biochem. 174, 111 -

117. Ness, G. C., Sukalski, K. A., Sample, C. E., Pendleton, L. C.,

McCreery, M. J. & Nordlie, R. C. (1989) J. Biol. Chem. 264,

Zakim, D. & Dannenberg, A. (1990) J. Biol. Chem. 265, 201-

Bertheloot, A., Vidal, H. & Van de Werve, G. (1991) J. Biol.

Garda, H. A. & Brenner, R. R. (1984) Biochim. Biophys. Actu

Garda, H. A. & Brenner, R. R. (1985) Biochim. Biophys. Actu

Brenner, R. R. (1984) Prog. Lipid Res. 23, 69-96. Brenner, R. R., Castuma, C. E. & Garda, H. A. (1986) Prog.

Garland, R. C. & Cori, C. F. (1972) Biochemistry 11, 4712-

Gumbhir, K., Sanyal, S. N., Minocha, R., Wali, A. & Majumdar,

Duttera, S. M., Byme, W. L. & Ganoza, M. C. (1968) J. Biol.

117.

Mol. Cell. Biochem. 6, 75-83.

7111-7114.

208.

Chem. 266, 5497-5507.

769, 160-170.

819,45-54.

Lipid Res. 25, 47-52.

4718.

S . (1989) Biochim. Biophys. Actu 981, 77-84.

Chem. 243, 2216-2228.

15. 16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

26. 27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Zakim, D. (1970) J. Biol. Chem. 245,4953-4961. Snoke, R. E. & Nordlie, R. C. (1972) Biochim. Biophys. Actu

Crain, R. C. & Zilversmit, D. B. (1981) Biochemistry 20,5320-

Sawaki, K., Taguchi, R. & Ikesawa, H. (1983) J. Biochem. (To-

Sawaki, K., Taguchi, R. & Ikesawa, H. (1983) J. Biochem. (To-

Dyatlovitskaya, E. V., Lemenovskaya, A. F. & Bergelson, L. D.

Cater, B. R. & Hallinan, T. (1971) Biochem. J. 124, 60P. Hill, D. J., Dawidowicz, E. A., Andrews, M. L. & Karnovsky,

M. J. (1983) J. Cell. Physiol. 115, 1-8. Mithieux, G., Vega, F., Beylot, M. & Riou, J. P. (1990) J. Biol.

Chem. 265, 7257-7259. Mithieux, G., Vega, F. & Riou, J. P. (1990) J. Biol. Chem. 265,

20364- 20368. Baginsky, E. S., Foa, P. F. & Zak, B. (1974) in Methods of

enzymatic analysis (Bergmeyer, H. U., ed.) vol. 2, pp. 876- 880, Verlag Chemie Int., Deerfield Beach.

258, 188-205.

5326.

kyo) 93, 525-535.

kyo) 93, 537-546.

(1979) Eul: J. Biochem. 99, 605-612.

Speth, M. & Schulze, H. U. (1986) FEBS Lett. 202, 32-36. Spector, A. A,, Ashbrook, J. D., Santos, E. C. & Fletcher, J. E.

Cooper, R. B., Noy, N. & Zakim, D. (1989) J. Lipid Res. 30,

Spector, A. A., John, K. & Fletcher, J. E. (1969) J. Lipid Res.

Gibson, R. A., McMurchie, E. J., Charnock, J. S. & Kneebone, G. M. (1984) Lipids 19, 942-951.

Venkatraman, J. T., Pehowich, D., Singh, B., Rajotte, R. V., Thomson, A. B. R. & Clandinin, M. T. (1991) Lipids 26,

Hollavay, C. T. & Garfield, S. A. (1981) Lipids 16, 525-

Maddaiah, V. T., Stemmer, C. L., Clejan, S . & Collipp, P. J.

Mimouni, V. & Poisson, J. P. (1992) Biochim. Biophys. Acta

Mimouni, V., Narce, M. & Poisson, J. P. (1992) Biochim. Bio-

Eck, M. G., Wynn, J. O., Carter, W. J. & Haas, F. H. (1979)

(1972) J. Lipid Res. 13, 445-451.

1719-1726.

10, 56-67.

441 -444.

532.

(1981) Biochim. Biophys. Actu 657, 106-121.

1123, 296-302.

phys. Acts 1133, 187-192.

Diabetes 28, 479-485.