The orientation of phosphate activated glutaminase in the inner mitochondrial membrane of synaptic and non-synaptic rat brain mitochondria

~

Pergamon

Neurochem. Int. Vol. 27, No. 4/5, pp. 367-376, 1995 Elsevier ScienceLtd. Printed in Great Britain

0197-0186(95)00018--6

THE ORIENTATION OF PHOSPHATE ACTIVATED G L U T A M I N A S E IN THE INNER MITOCHONDRIAL M E M B R A N E OF SYNAPTIC A N D NON-SYNAPTIC RAT BRAIN MITOCHONDRIA B J O R G R O B E R G , I N G E B O R G Aa. T O R G N E R and E L L I N G K V A M M E * Neurochemical Laboratory, University of Oslo, P.O. Box 1115, Blindern, 0317 Oslo, Norway (Received 8 November 1994 ; accepted 26 January 1995)

A~tract--When rat brain synaptic and non-synaptic mitochondria were incubated with [~4C]glutamine, [~4C]glutamate was rapidly released to the incubation medium, and the release was stimulated by phosphate, whereas [~4C]glutamate accumulated very slowly in the mitochondria and independently of the addition of phosphate. The specific activity of [14C]glutamate (dpm-nmol glutamate ~) in the incubation medium quickly reached the level of added [~4C]glutamine, but the specific activity of [~4C]glutamate in the mitochondria was found to be only 10-15% of that level. This indicates that glutamine-derived glutamate was released directly to the incubation medium, without being mixed with a general pool of endogenous glutamate in the mitochondria. Furthermore, there was no correlation between rate of glutamine hydrolysis and the uptake of glutamine into the mitochondria, as measured by the uptake of [3H]glutamine and glutamine induced mitochondrial swelling when calcium plus phosphate or asparagine were added. Glutamine hydrolysis was also not stimulated by partial disruption of the mitochondria following sonication, which should be expected if the rate of glutamine hydrolysis were limited by glutamine uptake. In addition, glutamine hydrolysis was strongly inhibited by mersalyl which is known to be impermeable to the inner mitochondrial membrane. Moreover, it is indicated that the enzyme was not an integral membrane protein. Thus, following fractionation of a Triton X-114 extract of brain synaptosomes, a major fraction of both the protein, as measured by immunoblot technique, and the enzyme activity were detected in the water phase. Our results therefore indicate that the whole molecule of phosphate activated glutaminase is externally localized in the inner mitochondrial membrane.

P A G is a mitochondrial enzyme, converting glutamine to glutamate and ammonia. The enzyme is involved in the formation of metabolic glutamate and the nitrogen metabolism in brain. P A G is in addition assumed to be engaged in the formation of transmitter glutamate, although other enzymes such as aminotransferases and glutamate dehydrogenase may also be involved. Furthermore, an interaction between P A G , the ketodicarboxylate carrier and aminotransferases has been suggested (Palaiologos et al., 1988). To evaluate the

role of P A G in transmitter glutamate formation, important knowledge is lacking about a possible mitochondriat glutamine carrier and where the glutamate is delivered, to the matrix region or to the surrounding synaptosol. In the former case, transmitter glutamate would first have to be transported from the matrix through the inner membrane by an active process, before entering the cytosolic transmitter pool, which is most likely located in synaptic vesicles. The functional and/or physical position of renal P A G relative to the mitochondrial compartments * Author to whom all correspondence should be addressed. have been subject to several studies over the past years, Abbreviations: ACHE, Acetyl choline esterase (3,1,1,7); and it has been a c o m m o n belief that P A G is localized CCO, Cytochrome c oxidase (EC 1,9,3,1) ; GGT, ),-Gluto the inner face of the inner membrane or in the tamyl transferase (EC 2.3.2.1); LDH, lactate dehydrogenase (EC 1,1,1,27); Mers, Mersalyl; OPA, o- matrix region (Crompton et al., 1973 ; Kalra and Brosphthaldialdehyde; PAG, phosphate activated glu- nan, 1974; Kovacevic, 1975; Shapiro et al., 1985). taminase (EC 3.5.1.2, glutamine amido hydrolase). Renal P A G shows latency (Kalra and Brosnan, 1974 ; Key Words: Glutaminase, Phosphate activated glutaminase, Mitochondria, Inner mitochondrial membrane, Brain, Kovacevic et al., 1980; Strzelecki and Schooiwerth, 1984), antibodies against P A G which inhibit the puriSynaptic mitochondria. 367

368

Bjorg Roberg et al.

fled enzyme do not inhibit P A G in renal mitoplasts (Shapiro et al., 1985), a n d avenaciolide inhibits oxidation o f externally added glutamate but not that o f glutamine-derived glutamate (Kovacevic, 1975 ; Brosn a n and Hall, 1977). We have obtained evidence t h a t pig renal P A G is functionally localized on the outer face of the inner m i t o c h o n d r i a l m e m b r a n e ( K v a m m e et al., 1991b). Several lines of evidence led to this conclusion: (1) glutamine-derived glutamate does not mix with a m i t o c h o n d r i a l pool of endogeneous glutamate, (2) differential inhibition pattern o f g l u t a m i n e uptake a n d hydrolysis indicating that P A G in pig renal mitoc h o n d r i a is not rate-limited by glutamine t r a n s p o r t into the m i t o c h o n d r i a , (3) the effects of n o n p e n e t r a b l e SH reagents on glutamine kinetics (only effect on

Vm.x). In brain mitochondria, experiments specifically aimed at investigating the orientation o f P A G in the inner m i t o c h o n d r i a l m e m b r a n e have not previously been published. However, we have shown that P A G in pig renal a n d b r a i n m i t o c h o n d r i a have similar general properties ( K v a m m e et al., 1988), which is in accordance with the results o b t a i n e d by Shapiro et al. (1991). In this paper we have m a d e further investigations of brain m i t o c h o n d r i a , a n d to obtain representative samples, we have examined b o t h synaptic as well as the unspecified, heterogenous p o p u l a t i o n o f non-synaptic or "free" mitochondria. The non-synaptic mitoc h o n d r i a are assumed to be derived mainly from glial cells and p r o b a b l y also from n e u r o n cell bodies, whereas the synaptic m i t o c h o n d r i a are mainly neuronal. Preliminary results have been published ( K v a m m e , 1983).

EXPERIMENTAL PROCEDURES

Chemicals

L-[3,4-3H(N)]glutamine, L-[~4C(U)]glutamine, and D-[I3H]mannitol were obtained from DuPont NEN, and L-glutamine as well as the other reagents from Sigma Chemical Co. (St Louis, Mo. U.S.A.). The decapeptide corresponding to the C-terminal end of brain PAG was synthesized in the Biotechnology Center, University of Oslo. Preparation o f mitoehondria and other subeellular particles

The S 1 fraction was prepared at 0 4 " C from homogenates of rat brain of various sources as described by Whittaker and Barker (1972), using an isotonic buffer containing 250 mM D-mannitol, 70 mM sucrose, and 10 mM Hepes, pH 7.4 (mitochondrial buffer). Synaptosomes and non-synaptic mitochondria were isolated essentially according to Dodd et al. (1981), using a Beckman Ti 60 fixed angle rotor for the sucrose cushion separations, at a centrifugation speed of

40,000 rpm for 25 rain. Synaptic mitochondria were further isolated essentially by the Method 2 of Lai and Clark (1979), by hypotonic lysis of the isolated synaptosomes in 6 mM Tris. HCI, pH 8.1. From the combined supernatant including the fluffy material on top of the mitochondrial pellet, "heavy synaptic membranes" were obtained by centrifugation at 17,000 g for 10 rain. Furthermore, microsomal and soluble fractions of brain homogenate were obtained from the supernatant following separation of crude mitochondrial fractions (prepared by centrifugation of the SI fraction at 18,000 g for l0 min), and synaptic microsomal and synaptosol of synaptosomes by centrifugation at 161,000 g for 38 min. The fractions were washed and resuspended in mitochondrial buffer. Sonicated mitochondria were prepared as in Kvamme et al. (1991b). To find the contamination of the brain mitochondria with constituents of other fractions, samples from the fractionation of the homogenate and the synaptosomes were examined with respect to marker enzymes (see below). The results from the fractionations are presented as histograms (Figs 1 and 2), where the mean relative specific activity (enzyme activity per mg protein in the fraction) is plotted against their mean relative protein content, according to de Duve (1975). In these diagrams, the percentage of activity of the enzyme recovered in each fraction is proportional to the area of the corresponding block. As can be seen from the figures, the non-synaptic (Fig. IC, d and b) and synaptic (Fig. 2C, d and a) mitochondria contained less than 1.2 and 3.5% of the cytosolic and synaptosolic LDH, respectively. The content of the plasma membrane marker AChE in nonsynaptic mitochondria was less than 2% of that in the microsomal fraction (Fig. 1A, d and c), whereas in synaptic mitochondria it was about 24% of that in the synaptic microsomes (Fig. 2A, d and b). These mitochondria were only washed once. However, when the synaptic mitochondrial pellets were washed five times with mitochondrial buffer (Gurd et al., 1974), the contamination by microsomes and lighter membranes was considerably reduced, to about 2 4 % of the microsomal AChE content (not shown). Furthermore, we can see from these figures that the mitochondrial marker CCO is enriched in the non-synaptic mitochondria (Fig. 1D), and also somewhat in the synaptic mitochondria, but a considerable part remains in the "heavy membrane" fraction (Fig. 2D). The distribution of PAG is quite similar to that of CCO (Figs 1 and 2, E), and that of GGT to that of the membrane marker AChE (Figs 1 and 2, B and A). Glutamate follows roughly the distribution of the cytosolic enzyme LDH (Figs 1 and 2, F and C). Glutamine hydrolysis

Glutamine hydrolysis was assayed by two assay methods, designated optimal assay and normal assay (Kvamme et al., 1991b). Briefly, the mitochondria (or fractions) were preincubated in assay buffer for 10 rain at 25°C before substrate addition. For the optimal assay, the incubation was performed in 150 mM potassium phosphate and 30 mM Lglutamine, pH 8.6, at 25°C, with a protein concentration of about 0.3 mg/ml. For the normal assay, which has been used here except in Figs 1, 2 and 5, the mitochondria (1-2 mg/ml) were incubated in 100 mM KC1, [~4C]glutamine, as indicated, 0.5 mM MgSO4, 25 mM mannitol, 7 mM sucrose, and 10 mM Hepes, pH 7.4, at 25°C, for 2 min when not indicated otherwise. Any other additions are as indicated in the tables and figures. Glutamate was monitored by the fluorimetric

369

Mitochondrial orientation of glutaminase

RSA

RSA

d.

D

2.0 A

6.0 4.0

1.0 2.0

a. a,

d,

E

B

4.0

1.0 0.5

ib

a.

a.

2.0

3.0

2.0

2.0 IF

1.0

a. 1.0

ao

2O

60

20

100 P r o t e i n , ~. of t o t a l

60

100

(cumulative)

Fig. 1. Distribution patterns of enzymes in rat brain homogenate. Ordinate : mean relative specific activity of fractions. Abscissa : fractions are represented by their mean relative protein content, and are from left to right : (a) nuclear ; (b) soluble ; (c) microsomes ; (d) non-synaptic mitochondria ; (e) synaptosomes. The capital letters are: (A) ACHE; (B) GGT; (C) LDH; (D) CCO; (E) PAG (optimal assay); and (F) glutamate. A typical experiment is shown.

method of Curthoys and Lowry (1973), or where indicated, by the column chromatographic method of Shapiro et al. (1982). The latter method was more suitable at low substrate concentrations. Since glutamine hydrolysis was activated by phosphate, it is assumed to be catalyzed by PAG. In this respect, we have found no evidence of a heat stable, maleate activatable, phosphate independent glutaminase, assayed following inactivation of PAG at 55°C for 10 min (Kvamme and Olsen, 1981). With normal assay and 0.5 mM [~4C]glutamine, PAG activity was found to be 13.5+0.7 and 8.3+0.3 nmol" min ~"mg protein- ~for synaptic (n = 15) and non-synaptic (n = 15) mitochondria, respectively. With optimal assay, PAG activity was found to be 330+98 n m o l ' m i n - l " m g protein ~ for frozen and thawed synaptosomal fractions (n = 3). Other assays

AChE was assayed essentially as described by Ellman et al. (1961), with the addition of 0.1% (w/w) Triton X-100 during the incubation. LDH was determined according to NichoUs (1978), CCO in the presence of 0.3% (w/v) deoxycholate according to Smith (1955) and Wojtczak et al. (1972), and GGT according to Orlowski and Meister (1965) and The Committee on Enzymes of the Scandinavian Society for

Clinical Chemistry and Clinical Physiology (1976). Units of enzyme activity are usually nmol-min- ~- mg protein except for CCO where the velocity constant in rain ~ was used. Protein was determined by the method of Lowry et al. (1951). Statistical significance was determined by unpaired and two-tailed t-tests, either by Student's test, or by Welch's alternate t-test for different SDs. Measurements o f specific activities o f labeled 9lutamate (dpm "nmol glutamate f) in the mitochondria and surroundiny incubation medium

The specific activities of [t4C]glutamate in mitochondria and the surrounding incubation medium were determined following incubation of the mitochondria with 2 mM [~4C]glutamine, essentially as described by Kvamme et al. (1991 b). Briefly, the mitochondria (2 mg/ml) were incubated for 0.5-5.0 min in 310 mM mannitol, pH 7.4, at 25°C, with 10 mM phosphate when indicated. To stop glutamine hydrolysis, 2 mM Mers was added, and the mitochondria were separated by centrifugation. To reduce the time for separation of the mitochondria from the incubation medium, they were also isolated by rapid filtration as described below for the uptake of [3H]glutamine. Glutamate was isolated from glutamine by a column chromatographic method (as

370

Bjorg Roberg et al.

RSA

RSA

D

A

2.0

2,0

1.O

1.0 C.

2.0

d,

a.

I

b.

C.

I E

B

1.5

1.0

½

C.

b. ]--

a,

d•.

0,5

C.

C 3.0 2.0 1.0

a,"~7 20

d.

C.

] 60

a.

b.

1 i00

c,

--[ 20

1.0

d,

, 60

I00

Protein, % of t o t a l (cumulative) Fig. 2. Distribution patterns of enzymes in rat brain synaptosomes+ Ordinates and abscissa as in Fig. l+ but the fraction are, from left to right ; (a) synaptosol, soluble ; (b) synaptic microsomes ; (c) heavy synaptic membranes ; and (d) synaptic mitochondria. Capital letters : see Fig. 1. A typical experiment is shown.

above, for glutamine hydrolysis), and the radioactivity of the ['4C]glutamate was determined. The activities of [L4C]glutamate were corrected for []4C]glutamate measured at zero time and 0"C. The total glutamate (the sum of endogenous and glutamine-derived glutamate) both in the mitochondria and the filtrate was measured by the fluorimetric method, and the specific activities calculated. Control experiments (with synaptic mitochondria) were done by isolating the OPA-derivative of glutamate and glutamine using reverse phase H P L C (Spectraphysics) according to Jones et al. (1981). Mitochondria were incubated with [3H]glutamine under conditions as above for 2 min, and the amino acids were separated. The peaks corresponding to OPA-glutamate and OPA-glutamine were isolated, and their contents of the OPA-amino acid in question and the corresponding radioactivity were measured. After correction for the values obtained after incubation for 0 min at 0'C, the specific activities were calculated. G l u t a m i n e transport

Glutamine transport was studied by two methods, by measuring the swelling of mitochondria in iso-osmotic glutamine, essentially as described by Joseph and Meijer (1981), and modified by K v a m m e et al. (1991a), or by monitoring the uptake of [~H]glutamine into mitochondria collected by vacuum-filtration according to Sastrasinh and Sastrasinh (1989), and slightly modified. Briefly, for the former method, mitochondria (about 0.3

mg/ml) were incubated at 2 5 C and pH 7.4 in a medium containing 200 m M u-glutamine, 50 m M o-mannitol, 2 m M sucrose, and 32 m M KCI, if not indicated otherwise. The decrease in absorbance at 546 n m was monitored for 4 min. Corrections were made for blanks where glutamine was substituted with mitochondrial buffer. The mitochondrial volume has earlier been shown to be inversely proportional to the optical density at 546 n m ( K v a m m e et al., 1991a). This method has some limitations because it depends on a very high glutamine concentration, and it is semiquantitative. Typically, the corrected decrease in absorbance at 546 n m for swelling in glutamine alone: 0.31 _+0.01 and 0.22_+ 0.02 for synaptic (n = 7) and non-synaptic (n = 7) mitochondria, respectively. For the second method, following 5 min preincubation at 2 5 C , the mitochondria (about 1 mg/ml) were incubated for 15 s at 25'C in a total volume of 0.2 ml, in a buffer containing 0.05 m M or 0.5 m M [3H]glutamine, 130 m M KC1, 0.5 m M MgSO4, 20 m M sucrose, 20 m M Hepes and 6/~g/ml rotenone, pH 7.4. The uptake was corrected for the non-specific binding of radioactivity to the mitochondria and filters at 3°C, using an incubation-temperature of - 2°C. At the end of the incubation, the reaction was stopped by the addition of I ml of a solution at - 2 ° C , containing 130 m M KCI, 20 m M Hepes and 0.5 m M MgSO+, pH 7.4 (stop-solution), followed by rapid filtration through a filter (DAWP, Millipore, diameter 25 mm, 0.65 #m pore size). The filters were immediately washed with 3 x 6 ml stop-solution. The filters were dissolved

371

Mitochondrial orientation of glutaminase in 12 ml scintillation-fluid (Picofluor 40, Packard) following the addition of 1 ml 2% Triton X-100. The radioactivity on the filters was measured with a scintillation-counter (Tricarb 1500, Packard) and corrected for unspecific binding. Corrected uptake of 0.5 mM [3H]glutamine: 247+__22 and 236+21 p m o l ' m i n - ~ ' m g protein -~ for synaptic (n = 10) and non-synaptic (n = 10) mitochondria, respectively. A disadvantage of this method (compared to the swelling method) is that glutamine-derived glutamate might interfere with the results because of the low glutamine concentration used, and there are no known inhibitors of PAG to completely restrain this activity. Therefore we have used short incubation times to minimize glutamine hydrolysis, and we have also avoided addition of phosphate which is a potent activator of PAG (Table 3), except when the effect of phosphate was tested.

Fractionation of synaptosomal fraction by Triton X-114 and immunoblottin# Fractionation in Triton X-114 was performed according to the method of Bordier (1980). Frozen synaptosomal fractions were thawed and diluted to 12 mg protein/ml. To this preparation an equal volume of 4% (w/v) Triton X-114 in a buffer containing 0.05 M Tris.HCl, 0.001 M EDTA, at pH 8.0. Following 15 rain on ice, the preparation was centrifuged at 100,000 g for 1 h. The Triton X-114 extract was incubated at 30°C for 15 min, leading to clouding, and then centrifuged at room temperature 12,100 g for 15 min to separate the phases, the water phase on the top. PAG was assayed by the optimal assay. lmmunoblottiny. Following SDS-PAGE the proteins were i

J

i

RESULTS As s h o w n in Fig. 3A, w h e n synaptic m i t o c h o n d r i a were i n c u b a t e d with ['aC]glutamine, the glutaminederived g l u t a m a t e released to the i n c u b a t i o n m e d i u m increased rapidly with time f r o m a n initial zero value. The release was m o r e t h a n three fold stimulated by 10 m M of p h o s p h a t e which indicates t h a t the g l u t a m a t e was p r o d u c e d by the P A G reaction. Similar results were o b t a i n e d w h e t h e r glutamate was m e a s u r e d by the radiometric or by the fluorimetric method. O n the o t h e r h a n d , as seen in Fig. 3B, the m i t o c h o n d r i a l glutamine-derived g l u t a m a t e increased slowly from zero value, a n d the total non-labeled m i t o c h o n d r i a l g l u t a m a t e was little c h a n g e d from a relatively high value at zero time. T h e r e was n o significant effect o f

i

i

8O

B

I

c q~ 0

~" O.

electrotransferred to nitrocellulose and incubated with the rabbit antiserum, raised against a decapeptide made from the deduced C-terminal sequence of PAG (Shapiro et al., 1991), followed by a horseradish peroxidase-conjugated secondary antiserum. The antibody-antigen complexes were visualized by 0.05% (w/v) diaminobenzidine, 0.03% (w/v) NiCI2 and 0.03% (w/v) H202 in 50 mM N H 4 C O ~. A band corresponding to the molecular weight of PAG appeared. No other major immunopositive bands were observed.

o

60

D~

D~

-g

~

4O

c~ v a)

~

20

0

t

Ill

I

I

I

1

2

3

4

5

Incubation

time (min)

0

i Incubation

2

3

4

time (rain)

Fig. 3. (A) Glutamine-derived glutamate in the medium of synaptic mitochondria following incubation with [~4C]glutamine (2 mM). Circles: 10 mM P~ added; triangles: no phosphate added. Open symbols: radiometric analysis; solid symbols: ftuorimetric analysis. (B) Glutamine-derived glutamate and total endogenous glutamate in synaptic mitochondria following incubation with glutamine, with and without phosphate averaged. Open squares : glutamine derived glutamate ; sohd squares : endogenous glutamate. Glutamine-derived glutamate: timepoints 1 and 2 min: mean + SEM (n = 4) ; data at 0.5 min are mean of 0.18 and 0.37, and at 5 rain: 0.44 and 0.50 nmol.mg protein -~. Endogenous glutamate: timepoints 0, 1 and 2 min: m e a n + S E M (n = 4) ; data at 5 min are mean of 0.50 and 0.44 nmol.mg protein -t. When using the radiometric method, glutamine-derived glutamate (nmol) was calculated by dividing the glutamate dpm by the specific activity (dpm- nmol glutamine ') of added [t4C]glutamine. The glutamate content was calculated as nmol.mg mitochondrial protein -~. Experiments with and without phosphate are averaged (no significant difference).

5

372

Bjorg Roberg eta/.

phosphate on mitochondrial glutamine-derived glutamate production which suggests that this glutamate is not produced by PAG. Furthermore, the specific activity of [~4C]glutamate in the mitochondria was markedly lower than that of the incubation medium, and increased slightly from I to 5 rain of incubation, as shown in Fig. 4. The specific activity of [~"C]glutamate in the incubation medium rapidly reached a level of that of added glutamine, which demonstrates that no non-labeled glutamate has been released, The specific activities of glutamate released and mitochondrial glutamate, were unaffected by the addition of 10 mM of phosphate. The specific activities of [J4C]glutamate and glutamine were also measured at 2 min incubation time in synaptosomal mitochondria by using HPLC for separation and quantitative determination of the amino acids. As shown in Fig. 4, the specific activities of [~4C]glutamate were essentially similar to those determined by the fluorimetric method. The corresponding specific activities of [~4C]glutamine in the incubation medium and mitochondria were almost equal (not shown), and were of the same order of magnitude as that of supernatant [~4C]glutamate,

z~

"-~ 100 4~

*d 60

(1) U?

2O ~o-0/

0

o-~J~

n

n

1

2

j[

a

t

n

3

4

5

I n c u b a t i o n time (rain) Fig. 4. Specificactivity (dpm" nmol glutamate ~)of [~4C]glutamate or [3H]glutamatefollowingincubation with [~4C]glutamine or [3H]glutamine. Circles: [~4C]glutamineadded. fluorimetric assay of glutamate: triangles: [3H]glutamine added, HPLC assay of glutamate. Open symbols: incubation medium, solid symbols: synaptic mitochondria. Experiments with and without phosphate added are averaged (no significant difference).The specific activity is calculated in percent of that of added glutamine. Timepoints 1 and 2 rain : average_+SEM (n = 4), timepoints 0.5, 3, and 5 min are average of two experiments, where the deviation from the means are no larger than 6% of the specificactivity of added glutamine.

demonstrating that the labeled glutamine was little diluted with endogenous mitochondrial glutamine. Non-synaptic mitochondria appeared to have a similar difference between the incubation medium and mitochondria with regard to glutamine-derived glutamate and specific activities as the synaptic ones (Table 1). The results were also independent of whether the mitochondria were separated by centrifugation or by the filter method. Table 2 demonstrates that in synaptosomal as well as non-synaptosomal mitochondria, glutamine hydrolysis was strongly inhibited by Mers, and so was the [3H]glutamine uptake and glutamine swelling. Similarly, phosphate stimulated glutamine hydrolysis, and had little effect on the uptake and no effect on glutamine swelling (Table 3). When phosphate and calcium were added together, glutamine hydrolysis was two to four fold stimulated, whereas [3H]glutamine uptake and the glutamine swelling were markedly reduced. It should be noted that essentially the same results were obtained by the [3H]glutamine uptake technique and that of glutamine swelling which minimizes the effect of PAG on the mitochondrial glutamine concentration due to the excess glutamine added. In contrast to the effect of calcium plus phosphate excess L-asparagine stimulated glutamine hydrolysis, but inhibited the uptake (Table 4). These experiments clearly demonstrate that there is no correlation between glutamine hydrolysis and glutamine uptake into the mitochondria. Glutamine hydrolysis is significantly higher in the synaptic mitochondria than in the non-synaptic ones, and this is contrary to the [~H]glutamine uptake. Partial disruption of the inner mitochondrial membrane by sonication does not effect any increase in the rate of glutamine hydrolysis (not shown). Following the fractionation of brain synaptosomes with Triton X-114, the activity of PAG was measured in the water phase, the detergent phase and the precipitate. In addition, PAG protein was monitored by immunoblot. Figure 5 demonstrates that about 90% of the activity and most of the PAG protein were found in the water phase, DISCUSSION

Dennis et al. (1977) found differences between the synaptic and non-synaptic mitochondria with regard to the ability to oxidize glutamate and malate, as well as to the activities of oxaloacetate aminotransferase and glutamate dehydrogenase. Furthermore, Steib et al. (1986) reported that glutamine was preferentially taken up by brain synaptic mitochondria, as corn-

373

M i t o c h o n d r i a l o r i e n t a t i o n of glutarninase Table 1. Glutamine-derived glutamate, total glutamate, and specific activity, as measured by the centrifugation and filter methods Centrifugation method

Filter method

Medium

Mitochondria

Medium

Mitochondria

31, 31 30, 29 104, 105

0.5, 0.6 3.8 +0.1 (4) 0.6, 0.5

24 _+2 (4) 20_+2 (4) 105 _+4 (3)

1.0 _+0.8 (4) 2.8_+0.7 (8) 5.2 _+0.8 (4)

15, 17 22, 20 70, 87

0.2, 0.2 2.2, 2.2 10, 11

29_+ 2 (3) 25 _+4 (4) 118 _+10 (3)

0.8 _+0.6 (4) 3.3 _+0.4 (8) 2 _+2 (4)

Synaptic mitochondria Glutamine-derived glutamate Total glutamate Specific activity (%)

Non-synaptic mitochondria Glutamine-derived glutamate Total glutamate Specific activity (%)

The conditions were as described for the determination of specific activity (dpm" nmol glutamate 1) of [~4C]glutamate in Experimental Procedures, with 2 mM ['4C]glutamine as substrate, no phosphate added, and 2 rain incubation time. Mitochondria were separated from the medium either by centrifugation or by rapid filtration. Glutamine-derived glutamate and total glutamate are presented as nmol "rag mitochondrial protein ], either as means _+SEM with the number of experiments in parentheses or both results where only two experiments were performed. The specific activity is calculated in percent of that of added ['4C]glutamine. Table 2. The effects of Mers on glutamine hydrolysis and glutamine uptake Glutamine hydrolysis (%)

[3H]glutamine uptake (%)

Gin-swelling (%)

Synaptic mitochondria Control Mers

100-+ 11 (4) 19+_21" (4)

100-+ I 1 (4) I +0.41" (3)

100_+9 (3) 57-+ I t (3)

100-+ 17 (4) 5_+4:~ (4)

100_+3 (3) 55-+ 15~ (3)

Non-synaptic mitochondria Control Mers

100-+2 (4) 11 _+1" (4)

The conditions for glutamine hydrolysis (normal assay) and uptake were as described in Experimental Procedures, with 0.5 mM ['4C]glutamine or[3H]glutamine as substrates, respectively, and for glutamine-induced swelling with 200 mM glutamine. Mers, 0.25 mM was present during the preincubation. The results are presented in percentage of the respective controls, as means + SEM, with the number of experiments in parentheses. * P < 0.001, t P < 0.01, ~:P < 0.05 vs the corresponding controls. Table 3. The effects of phosphate and calcium on glutamine hydrolysis and glutamine uptake Glutamine hydrolysis (%)

[3H]glutamine uptake (%)

Glutamine-swelling (%)

Synaptic mitochondria Control Pj P~+Ca 2+

100_+3 (11) 167_+4" (11) 239+7* (11)

100_+5 (6) 137_+ 105 (5) 77_+ 10 (5)

100_+5 (4) 94_+6 (4) 61 _+81" (4)

100 _+6 (6) 96 _+15 (6) 0_+3* (6)

100 -+ 3 (4) 88 + 18 (5) 52_+ 12t (3)

Non-synaptic mitochondria Control P~ Pi+Ca 2+

100 _+3 (11) 256 _+5* (I 1) 438 _+16" (10)

The conditions were as in Table 2. The concentration of phosphate (Pi) was 10 mM, and of Ca 2+ 0.5 mM. * P < 0.001, t P < 0.01, :~P < 0.05 vs the corresponding controls.

pared to non-synaptic ones. However, we could not confirm this finding, but since the glutamine hydrolysis was significantly higher in the synaptic mitochondria than in the non-synaptic ones, P A G is likely to be important for the synthesis of transmitter glutamate. The great similarities in experimental results previously reported for renal and brain P A G (Kvamme

et al., 1988) indicate that the same mechanism for releasing glutamate is operating in both these organs. Our present findings which are also, in principle, similar to those reported for renal mitochondria (Kvamme et al., 1991b) have certain implications for the orientation of PAG in the inner mitochondrial membrane. If PAG were localized to the inner face of this membrane or to the matrix region, one should expect that

374

B.jorg Roberg et al. Table 4. The effects of excess L-asparagineof glutamine hydrolysisand glutamine uptake Glutaminehydrolysis ['H]glutamine uptake (nmol'min ~'mgprotein ~) (pmol'min t'mgprotein 5[vnaptic mitochondria

Control L-asparagine

2.13±0.08 (4) 2.96_+0.11~"(4)

71 + 10 (5) 21 ±5* (10)

1.43_-+0.03+ (3) 2.50_+0.00* (4)

53 ± 5 (9) 21 +6* (10)

Non-synaptic mitochondria

Control t.-asparagine

The conditions were as in Table 2, except that the substrate concentrations ([~4C]glutamine and [~H]glutamine) were 0.05 raM. Asparagine (10 m M ) was added together with the substrate. * P < 0.001, t P < 0.01 vs the corresponding controls. ~ P < 0.001 vs glutamine hydrolysis in synaptic mitochondrial control.

glutamate produced was first released to the inner compartment and then transferred to the incubation medium. In that case, and if this glutamine-derived glutamate mixed with a general pool of mitochondrial endogenous glutamate, the specific activity of endogenous [14C]glutamate should be equal to that of glutamate released to the incubation medium. However, we have demonstrated here that the specific activity of mitochondrial [~4C]glutamate formed after incubation with [HC]glutamine is much lower than that of the incubation medium (Fig. 4, Table 1). Certain artifacts must be kept in mind, such as contamination of our mitochondrial preparations with other unlabeled subcellular particles, release of PAG to the incubation medium, or a not phosphate-activated externally localized glutamine hydrolyzing enzyme. However, our mitochondria are little contaminated with other subcellular particles (see Experimental Procedures, Figs I, 2). Furthermore, after centrifugation of the mitochondria no P A G activity could be detected in the supernatant, which demonstrates that very little P A G has been released, and

Phase:

following the addition of phosphate, the specific activity of glutamate was unchanged in spite of highly stimulated glutamate formation (Fig. 4). This renders the possibility unlikely that another glutaminase which is not phosphate-activated is operating in the release of glutamate to the incubation medium. Since substrates for aminotransferase reactions which participate in the transfer of glutamate across the inner mitochondrial membrane (Palaiologos et al., 1988), are present in very limited quantities in isolated mitochondria, the transfer of glutamate is restricted. Thus, by the action of PAG, glutamine-derived glutamate has in some way to be released directly to the incubation medium. Moreover. if PAG were internally localized in the mitochondria, glutamine from the outside would have to be transported through the inner membrane to be hydrolyzed. We have previously obtained evidence that the glutamine uptake in pig renal mitochondria is regulated by a glutamine carrier (Kvamme et al., 1991b: Roberg et al., 1994), which is similar to that of liver mitochondria (Joseph and Meijer, 1981;

Precipitate

Detergent

Water

3.5 ± 0.9

5.2 ± 2.3

91.5 _+ 3.0

Immunoreactivity PAG activity (% o f total)

Fig. 5. Distribution of PAG following fractionation with Triton X-114 of brain synaptosomal fraction. Triton X-114 fractionation was performed as described in Experimental Procedures. The three fractions were diluted to the same volume, and the PAG content determined by activity (optimal assay) and expressed as mean_+ SEM (n = 3), or immunoblots (one representative experiment shown). PAG activity in precipitate and detergent phases were significantly lower than in the water phase, P < 0.0001.

Mitochondrial orientation of glutaminase Soboll et al., 1991a) and rat renal mitochondria (Adam and Simpson, 1974; Goldstein and Boylan, 1978 ; Kovacevic et al., 1980). Minn (1982) and Steib et al. (1986) also provided evidence for an active, saturable uptake mechanism for glutamine in brain mitochondria. This transport is dependent on a highly regulated transporter (Roberg et al., to be published). Whether the activity of PAG is limited by the supply of glutamine, has been controversial. Based on measurements of free glutamine in the matrix space, opposite views are published (Curthoys et al., 1984; Simpson, 1988). As shown in Tables 3 and 4, there is no correlation between glutamine hydrolysis and uptake, similar to that previously described for renal mitochondria (Kvamme et al., 1991b). In addition, the rate of glutamine hydrolysis does not increase following partial disruption of the inner mitochondrial membrane by sonication (not shown), which should be expected if the rate of glutamine hydrolysis were limited by glutamine uptake into the mitochondria. The notion that PAG is localized externally in the inner mitochondrial membrane is supported by the finding that the enzyme is inhibited by Mers (Table 2) which is known to be impermeable to the inner membrane (Tietze, 1969 ; Kvamme, 1983). In that case it is possible that glutamine-derived glutamate is in part actively channeled to glutamate dehydrogenase in the matrix region for oxidation by a similar process as described by Cohen et al. (1987) for ornithine. On the other hand, previous results (Kvamme and Olsen, 1981), indicate that there may be both an internal and external species of glutaminase. The former species that is not a PAG, may provide glutamate for the glutamate dehydrogenase reaction. It is also a possibility that PAG penetrates the inner membrane and that the molecule is localized both on the outer and the inner face of the membrane, but with the active sites on the outer face. However, as a major fraction of the PAG protein and activity is found in the water phase following phase separation of a Triton X-114 extract of brain synaptosomes, it is indicated that PAG is not an integral membrane protein (Fig. 5). Recent sequence analysis also suggests that the PAG protein does not have any membrane spanning domains (Shapiro et al., 1991). Since it is widely accepted that PAG is attached to the inner mitochondrial membrane, and the enzyme thus appears not to penetrate the membrane, these results together with the other experimental findings, support the view that the PAG protein is predominantly externally localized in this membrane and mainly responsible for glutamate formation in isolated mitochondria.

375

These results most likely have implications for the possible role of PAG in the production of transmitter glutamate. Acknowledgement--This work was supported in part by

grants from the funds of Jahre, Inger Haldorsen and Nansen.

REFERENCES

Adam W. and Simpson D. P. (1974) Glutamine transport in rat kidney mitochondria in metabolic acidosis. J. clin. Invest. 54, 165-174. Bordier C. (1980) Phase separation of integral membrane proteins in Triton X- 114 solution. J. biol. Chem. 256, 16041607. Brosnan J. T. and Hall B. (1977) The transport and metabolism of glutamine by kidney-cortex mitochondria from normal and acidotic rats. Biochem. J. 164, 331 337. Cohen N. S., Cheung C.-W. and Raijman L. (1987) Channeling of extramitochondrial ornithine to matrix ornithine transcarbamylase. J. biol. Chem. 262, 203-208. Crompton M., McGivan J. D. and Chappell J. B. (1973) The intramitochondrial location of glutaminase isoenzymesof pig kidney. Biochem. J. 132, 27-34. Curthoys N. P. and Lowry O. H. (1973) The distribution of glutaminase isoenzymes in the various structures of the nephron in normal, acidotic, and alkalotic rat kidney. J. biol. Chem. 248, 162-168. Curthoys N. P., Shapiro R. A. and Haser W. G. (1984) Enzymes of renal glutamine metabolism. In: Glutamine Metabolism in Mammalian Tissues (H/iussinger D. and Sies H., eds), pp. 16-31. Springer, Berlin. de Duve C. (1975) Exploring cells with a centrifuge. Science 189, 186-194. Dennis S. C., Lai J. C. K. and Clark J. B. (1977) Comparative studies on glutamate metabolism in synaptic and nonsynaptic rat brain mitochondria. Biochem. J. 164, 727736. Dodd P. R., Hardy J. A., Oakley A. E., Edwardson J. A., Perry E. K. and Delaunoy J.-P. (1981) A rapid method for preparing synaptosomes : comparison, with alternative procedures. Brain Res. 226, 107-118. Ellman G. L., Courtney K. D., Andres Jr V. and Featherstone R. M. (1961) A new and rapid colorimetric determination and acetylcholinesteraseactivity. Biochem. Pharmac. 7, 88-95. Goldstein L. and Boylan J. M. (1978) Renal mitochondrial glutamine transport and metabolism : studies with a rapidmixing, rapid filtration technique. Am. J. Physiol. 234, F514-F521. Gurd J. W., Jones L. R., Mahler H. R. and Moore W. J. (1974) Isolation and partial characterization of rat brain synaptic plasma membranes. J. Neurochem. 22, 281 290. Jones B. N., P~i/iboS. and Stein S. (1981) Amino acid analysis and enzymatic sequence determination of peptides by an improved o-phthaldialdehyde precolumn labeling procedure. J. liq. Chrom. 4, 565-586. Joseph S. K. and Meijer A. J~ (1981) The inhibitory effects of sulfhydryl reagents on the transport and hydrolysis of glutamine in rat-liver mitochondria. Eur. J. Biochem. 119, 523 529. Kalra J. and Brosnan J. T. (1974) The subcellular local-

376

Bjorg Roberg et al.

ization of glutaminase isoenzymes in rat kidney cortex. ,I. biol. Chem. 249, 3255-3260.

Kovacevic Z. (1975) Possible mechanisms of the efflux of glutamate from kidney mitochondria generated by the activity of mitochondrial glutaminase. Biochim. Biophvs. Acta 396, 325-334. Kovacevic Z., Bajin K. and Pavlovic M. (1980) Volume changes of rat kidney mitochondria, transport of glutamine and its inhibition of mersalyl, lnt. J. Biochem. 12, 139-143. Kvamme E. (1983) Glutaminase (PAG). In: Glutamine, Glutamate and GABA in the Central Nervous System (Hertz L., Kvamme E., McGeer E. G. and Schousboe A., eds), pp. 51-67. Alan R. Liss Inc., New York. Kvamme E. and Olsen B. E. (1981) Evidence for compartmentation of synaptosomal phosphate-activated glutaminase. J. Neurochem. 36, 191(~1923. Kvamme E., Svenneby G. and Torgner I. Aa. (1988) Glutaminases. In: Glutamine and Glutamate in Mammals (Kvamme E., ed.), Vol. 1, pp. 53-67. CRC Press, Boca Raton, Fla. Kvamme E., Roberg B. and Torgner 1. Aa. (1991a) Effects of mitochondrial swelling and calcium on phosphate-activated glutaminase in pig renal mitochondria. Eur. J. Biochem. 197, 675 680. Kvamme E., Torgner I. Aa. and Roberg B. (1991 b) Evidence indicating that pig renal phosphate activated glutaminase has a functionally predominant external localization in the inner mitochondrial membrane. J. biol. Chem. 266, 13185 13192. Lai J. C. K. and Clark J. B. (1979) Preparation of synaptic and non-synaptic mitochondria from mammalian brain. Meth. En~vm. 55, 51-60. Lowry O. H., Rosebrough N. J.. Farr A. L. and Randall R. J. (1951) Protein measurement with the folin phenol reagent. J. biol. Chem. 193, 265 275. Minn A. (1982) Glutamine uptake by isolated rat brain mitochondria. Neuroscience 7, 2859-2865. Nicholls D. (1978) Calcium transport and proton electrochemical potential gradient in mitochondria from guineapig cerebral cortex and rat heart. Biochem. J. 170, 511 522. Orlowski M. and Meister A. (1965) Isolation of y-glutamyl transpeptidase from hog kidney. J. biol. Chem. 240, 338 347. Palaiologos G., Hertz L. and Schousboe A. (1988) Evidence that aspartate aminotransferase activity and ketodicarboxylate carrier function are essential for biosynthesis of transmitter glutamate. J. Neurochem. 51, 317 320. Roberg B., Torgner I. Aa. and Kvamme E. (1994) Properties of the pig renal mitochondrial transporter. In: Renal

Ammonia,qenesis and lnterorgan Cooperation in Acid-base Homeostasis (Tizianello A., Baverel G., Endou H., Schoolwerth A. C. and O'Donovan D. J., eds) Contr. Nephrol. (Berlyne G. M. and Giovannetti S., eds), Vol.

110, pp. 7-12. Karger, Basel. Sastrasinh H. and Sastrasinh M. (1989) Glutamine transport in submitochondrial particles. Am. J. Physiol. 257, F1050F1058. Shapiro R. A., Morehouse R. F. and Curthoys N. P. (1982) Inhibition by glutamate of phosphate-dependent glutaminase of rat kidney. Biochem. J. 207, 561-566. Shapiro R. A., Haser W. G. and Curthoys N. P. (1985) The orientation of phosphate-dependent glutaminase on the inner membrane of rat renal mitochondria. A rchs Biochem. Biophys. 243, 1-7. Shapiro E. A., Farrell L., Srinivasan M. and Curthoys N. P. (1991) Isolation, characterization, and in vitro expression of a cDNA that encodes the kidney isoenzyme of the mitochondrial glutaminase. J. biol. Chem. 266, 18,79218,796. Simpson D. P. (1988) Renal glutamine metabolism. In : Glutamine and Glutamate in Mammals (Kvamme E., ed.), Vol. 1, pp. 203-219. CRC Press, Boca Raton, Fla. Smith L. (1955) Spectrophotometric assay of cytochrome c oxidase. In : Methods o f Biochemical Analysis (Glick D., ed.), Vo[. II, pp. 427434. Interscience, New York. Soboll S., Lenzen C., Rettich D., Gr~ndel S. and Ziegler B. (1991) Characterisation of glutamine uptake in rat liver mitochondria. Eur. J. Biochem. 197, 113 117. Steib A., Rendon A., Mark J. and Borg J. (1986) Preferential glutamine uptake in rat brain synaptic mitochondria. FEBS Lett. 207, 63-68. Strzelecki T. and Schoolwerth A. C. (1984) The significance of the attachment rat kidney glutaminase to the inner mitochondrial membrane. Biochim. Biophys. Acta 801, 334 341. Tietze F. (1969) Enzymic method tbr quantitative determination of nanogram amounts of total and oxidized glutathione: applications of mammalian blood and other tissues. Analyt. Biochem. 27, 502--522. The Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology (1976) Recommended method for the determination of 7-glutamyltransferase in blood. Scand. J. clin. Lab. Invest. 36, 119 125. Whittaker V. P. and Barker L. A. (1972) The subcellular fractionation of brain tissues with special reference to the preparation of synaptosomes and their component organelles. In: Methods of Neurochemistry (Fried R., ed.), Vol. 2, pp. 1 52. Marcel Dekker, New York. Wojtczak L., Zaluska H., Wroniszewska A. and Wojtczak A. B. (1972) Assay for the intactness of the outer membrane in isolated mitochondria. Acta Biochim. Polon. 19, 227 234.