THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 250, No. 20, Issue of October 25, pp. 8148-8158, 1975

Printed in IJ.S.A.

Glutamine Transport and Metabolism by Mitochondria from Dog Renal Cortex GENERAL PROPERTIES AND RESPONSE TO ACIDOSIS AND ALKALOSIS*

(Received for publication, February 6. 1975)

DAVID P. SIMPSON+ AND WILLIAM ADAM§

From the Departments of Medicine and of Physiology and Biophysics, University of Washington School of Medicine, and the United States Public Health Service Hospital, Seattle, Washington 98114

Mitochondria from dog renal cortex were incubated with L- [‘“C jglutamine. Glutamate metabolism was prevented by inhibitors so that glutamate accumulated either in the mitochondrial matrix space or in the medium. The formation and accumulation of glutamate formed from glutamine and the distribution of glutamine in the mitochondrial fluid spaces were studied. In the matrix space glutamate rapidly reaches levels over 5 times that of glutamine in the medium. A more gradual accumulation occurs in the medium as glutamate is transported out of the mitochondria. Addition of an energy source such as succinate to the medium accelerates glutamate formation. A K, of 0.6 mM appears to govern the reaction at low concentrations of glutamine; at about 4 mM an abrupt change in kinetics occurs with a K, of 5 mM

above that level. Both NH,+ and glutamate inhibit glutamine metabolism, and phosphate stimulates it, but little effect of glutamate or phosphate occurs at low levels of these substances. The pH optimum of the reaction is between 7.4 and 7.8. Mersalyl and p-chloromercuribenzoate strongly inhibit glutamate formation; N-ethylmaleimide and bromcresol green have weaker inhibitory actions, and borate increases the reaction rate. In the presence of mersalyl, glutamine is strictly confined to the outer space of mitochondria and none is detectable in the matrix space. Similarly at 0” glutamine is confined to the simultaneously determined sucrose or mannitol spaces.

Glutamate formation from L-glutamine was also compared in mitochondria and submitochondrial preparations from renal cortex of litter-mate dogs with chronic metabolic acidosis or alkalosis. Total glutamate formation and concentration of glutamate in the matrix space are about twice as great in mitochondria from acidotic animals. Ammonia production is also significantly greater by the acidotic preparations. However, there is no difference in the levels of phosphate-dependent glutaminase in the two groups of mitochondria. Mitochondrial water spaces, ADP:O ratios, and oxygen consumption are similar in the acidotic and alkalotic mitochondria. Compared to kidney, liver mitochondria show only a limited ability to metabolize glutamine, and there is no increase in glutamate formation by liver mito- condria from acidotic dogs. The increased formation of glutamate in acidosis by dog renal cortex mito- condria is retained by submitochondrial fractions containing the inner membrane prepared by digitonin and Lubrol treatment.

The results indicate the presence of an energy-dependent element in the inner membrane which delivers glutamine from outside the mitochondria to phosphate-dependent glutaminase. This element may be either a distinct carrier molecule or an integral part of phosphate-dependent glutaminase which is not detected by the assay for the extracted enzyme. The carrier does not appear to release glutamine directly into the matrix space; rather, by the combined action of the carrier and phosphate-dependent glutaminase, glutamine is taken up from the medium, and after removal of ammonia glutamate is released into the matrix space. Chronic metabolic acidosis causes an activation of this element or carrier which increases delivery of substrate to glutaminase without change in the activity of extracted phosphate-dependent glutaminase itself. This process accounts for the increases in urinary ammonium excretion and in glutamine utilization which occur in chronic metabolic acidosis.

*This research was supported by Grants AM-09822 and AM-06741 3706 from the National Institutes of Health. 5 Recipient of United States Public Health Servire International

$ To whom correspondence regarding this paper should be directed. Fellowship (FO 5TW1816). Present address, University of Melbourne. Present address, Department of Medicine, The University of Wisconsin Department of Medicine, Austin Hospital, Heidelberg, 3084, Victoria, Center for Health Sciences, 1300 University Avenue. Madison, Wis. Australia.

8148

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8149

Hydrogen ions produced by metabolism in the form of noncarbonic acid are eliminated from the body in the urine principally as NH,+. Most of t.he NH, which combines with hydrogen ions to form urinary NH,+ is produced in the renal cortex of the kidney from the amide, and to a lesser extent from the amine, nitrogen of L-glutamine (1). The glutamine metabo.. lized in the kidney is delivered to the kidney in the renal arterial blood. In dog kidney, unlike the rat, glutamine synthetase is lacking so that formation of glutamine is not a significant process in this species (2, 3). Metabolism of L-glutamine by cells of renal cortex occurs principally through oxidative pathways in which mitochondrial reactions play a predominant role. Although other reactions may make some contribution to its metabolism, most of the glutamine used by the kidney is deamidated to glutamate by phosphate-depend- ent glut,aminase (L-glutamine amidohydrolase, EC 3.5.1.2), an intramitochondrial enzyme (4-7). Regulation of ammonia formation from glutamine in the kidney provides a means by which the kidney can adjust NH,+, and hence hydrogen ion, excretion to compensate for changes in the rate of hydrogen ion production (8, 9).

Of the various conditions which influence renal ammonia production from L-glutamine in the intact animal, chronic metabolic acidosis has the most significant role in homeostasis and has received the most attention by investigators. When metabolic acidosis is produced experimentally a gradual in- crease in ammonium excretion occurs, reaching a maximum in man in 4 to 6 days at which t,ime ammonium excretion may exceed control levels by 5-fold or more (8, 9). The peak rate of ammonium excretion does not coincide with the maximum degree of acidosis but follows it by several days. The change in ammonium excretion reflects a marked alteration in renal ammonia production from glutamine, an alteration which is preserved in tissue slices from acidotic animals. Slices from acidotic dog or rat renal cortex deamidate and oxidize more glutamine than slices from control kidneys when incubated in a medium of identical composition and pH (10-12). Thus stimulation of ammonia production from glutamine is not the consequence of a direct action of acidosis on some step in renal metabolism but rather represents a change in the intrinsic ability of cells of renal cortex to metabolize glutamine. The nature of this adaptive response to acidosis and the site at which it occurs have been extensively investigated, but no clear understanding of this phenomenon has resulted (9, 13).

A number of studies of the properties of phosphate-depend- ent glutaminase have been carried out after extraction of the enzyme from renal tissue to obtain a soluble form, and a few studies have been made of glutamine metabolism in intact kidney mitochondria (14-16). Recently we described some properties of glutamine metabolism in rat kidney mitochondria and obtained evidence suggesting that acidosis increases the delivery of glutamine to glutaminase by an effect on the inner mitochondrial membrane (17). The present study extends these results to dog kidney. The properties of glutamate formation from glutamine are examined in an intact mitochon- drial system, and characteristics are described which indicate that glutamine is delivered to glutaminase by an inner membrane element or carrier with properties distinct from those of extracted phosphate-dependent glutaminase. Acidosis activates this carrier leading to increased uptake of gluta- mine and delivery to glutaminase; the action of this enzyme on the increased amount of substrate available to it results in increased ammoniagenesis.

METHODS

Preparation and Incubation of Mitochondria-Dogs were killed by injecting a bolus of potassium chloride intravenously and were then exsanguinated. The kidneys were removed and placed in an ice-cold solution of 0.14 M NaC110.01 M KCl. The renal cortex was separat,ed by slicing from the rest of the kidney. Three gram portions of cortex were placed in 15 ml of ice-cold 0.3 M sucrose/5 rnrvr Hepes’/l mM EDTA and minced finely with scissors. Ilsually six to eight such portions were prepared from a pair of kidneys. Homogenization was carried out with six strokes of a Potter-Elvehjem homogenizer modified to provide a clearance of 0.25 mm and held in a motor-driven stirrer operated at approximately 300 rpm. The homogenizer was surrounded by chipped ice throughout the procedure. The homogenate was centri- fuged for 5 min at 700 x g. The supernatant was centrifuged for 10 min at 8000 x g (max) in an 870 head of an IEC B-20 centrifuge with slow acceleration. The supernatant was decanted until only a few milliliters remained. The fluffy layer was suspended by gentle swirling of the contents of the tube and removed with a capillary pipet. The pellet, was resuspended in 0.3 M sucrose/5 mM Hepes by the “cold-fin- ger” technique (18) and centrifuged in the same manner. The pellet from this centrifugation was again resuspended and centrifuged. The final pellet was resuspended in a small volume of 0.3 M sucrose/5 mM

Hepes. All steps in the isolation procedure were carried out at 0”. Unless otherwise noted, incubation was performed for 2 min at 37”

in sealed Erlenmeyer flasks containing 0.1 ml of’ mitochondrial suspension (2 to 5 mg of protein) and 0.9 ml of medium equilibrated with 100% 0,. A typical medium contained 115 mM KCI. 20 mM sucrose (from the mitochondrial suspension), 18 mM Hepes (pH 7.4). 0.5 mM

MgSD,, 1 mM L-[U-“Cjglutamine, 1 mM sodium arsenite, 1 pg/ml of rotenone, and a tritium source, either [3H]water, [3H]sucrose, or [3H]mannitol. When changes in the composition of the medium were made, the concentration of KC1 was adjusted to keep the osmolality constant.

Separation and Treatment of Mitochondria and Medium-Mito- chondria and medium were separated by the technique of van Dam and Harris (19). One hundred microliters of 1 N HClO, were placed in a microcentrifuge tube. A layer of silicone oil (Versilube F-50, Gen- eral Electric Co.) 3- to 5-mm-thick was placed on top of the perchloric acid. After incubation 200 ~1 of the incubation mixture were layered on top of the silicone oil, and the tube was centrifuged in a Beckman model 152 Microfuge for 1 min. Another sample was simultaneously prepared by adding 200 ~1 of incubation mixture directly to 100 ~1 of 1 N perchloric acid. Perchloric acid extracts of mitochondria or medium were immediately neutralized with a solution of equal parts 0.1 N potassium phosphate, pH 6.2, and 1 N KOH. After neutraliza- tion a sample was taken for measurement of radioactivity. In the remainder of the solution [“C]glutamine and [“C]glutamate were separated by ion exchange chromatography on small Dowex 1, formate, columns as previously described (17). This technique quantitatively separates glutamine and glutamate and yields fract,ions without significant contamination by other labeled compounds under the con- ditions of these experiments. Samples were counted by double iso- tope measurement in a Tri-Carb liquid scintillation spectrometer (Packard Instrument Co.). Efficiency was determined by re-counting after addition of an internal standard, first of [“C]toluene and then of [3H]toluene.

Space Measurements and Calculations--In most experiments the outer and total water spaces were measured in a separate set of flasks containing [“Clsucrose and [*HIwater. The volume of a space was determined by dividing the total disintegrations per min of a labeled compound present in the mitochondrial sample by the disintegrations per min per ~1 of that substance in the medium. The matrix space was calculated by subtracting the volume of the outer (sucrose) space from the total water space. The concentration of glutamate in the matrix space was determined hy first dividing the disintegrations per min of glutamate in mitochondrial sample by the disintegrations per min per nmol of glutamine present in the original medium and then by dividing this quotient by the volume of the matrix space. The concentration of glutamate outside the mitochondria was determined after correcting the total disintegrations per min in glutamate in the medium for glutamate disintegrations per min in the matrix space.

Each sample provided data on (a) the concentration of glutamate

‘The abbreviations used are: Hepes, N-2-hydroxyethylpiperazine- N’-2.ethanesulfonic acid; TMPD, tetramethyl-p-phenylenediamine; p-CMB, p-chloromercuribenzoate.

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8150

present in the matrix space, (b) the concentration of glutamate in the medium outside the mitochondria, (c) the volume of distrihution of glutamine, and (d) the volume of the outer or total water space of’ the mitochondria in the sample. Each experimental point is the average of two determinations performed under identical conditions. Statistical results are reported as mean + S.E.

Effects of Chronic Acidosis and Alkalosis-Each of a pair of litter mate dogs was fed ammonium chloride (8 meqikglday) or sodium bicarbonate (18 meq/kg/day) mixed with standard dog ration for 10 to 14 days in order to produce states of chronic metabolic acidosis or alkalosis. Plasma bicarbonate levels were checked at intervals during the preparation period to ensure adequacy of the alteration in acid-base state. At the time of sacrifice plasma bicarbonate levels were between 12 and 18 meq/liter in the acidotic group and between 26 and 35 meq/liter in the alkalotic animals. In some cases as noted the ammonium chloride was supplemented with potassium chloride, 4 meq/kg/day. On the day of an experiment the two animals were killed at the same time, and mitochondria from each dog were prepared simultaneously. Before incubation the mitochondria from the two animals were diluted with sucrose-Hepes to obtain preparations of approximately equal protein concentration.

Submitochondrial Preparations-Submitochondrial fractions of mi- tochondria from acidotic and alkalotic dogs were prepared by the method of Chan et al. (20). Mitochondria were incubated for 15 min with freshly prepared digitonin solution (0.12 mg/mg of protein) to remove outer membrane and outer space activity. Matrix activity was reduced by further incubation with Luhrol (0.16 mg/mg of protein) for 15 min. The degree of fractionation was determined by measuring activity of marker enzymes for the outer membrane (monoamine oxidase), inner membrane (cytochrome oxidase), and matrix (malate dehydrogenase).

Assays-Oxygen consumption, ADP:O ratios, and respiratory con- trol ratios were measured with an oxygen electrode (Yellow Springs Instrument Co.) (21). For these measurements 0.1 ml of mitochondrial suspension was added to 2.9 ml of medium containing 60 mM KCI, 10 mM potassium phosphate (pH 7.4), 5 mM MgSO,, 142 mM sucrose, 0.3 mM ADP, 5 mg/ml of bovine serum albumin, and 10 mM succinate or 10 mM malate plus 10 mM pyruvate.

Phosphate-dependent and phosphat,e-independent glutaminase were measured fluorometrically by the method of Curthoys and Lowry (22) in the single animal experiments and by measuring ammonia production from glutamine in borate extracts of lyophilized mitochon- dria (23) when litter mate animals were used. Ammonium was assayed by a modification of the alkaline phenate technique (24). Endogenous glutamine and glutamate were measured by the enzymatic cycling technique of Curthoys and Lowry (25). All enzymes in the membrane fractions were assayed after treatment with 0.3 mg of Lubrol/mg of protein (26). Monoamine oxidase activity was measured with [“Cltyramine for substrate (27). Cytochrome oxidase (28) and malate dehydrogenase (29) were determined spectrophotometrically. Protein was determined by the biuret technique (30).

For measuring potassium content of renal cortex a piece of cortex was placed in a tared weighing vessel and dried in a desiccator to a constant weight. Three milliliters of water were added, and the potassium was extracted by heating in a boiling water bath for 1 hour. After cooling, the extract was diluted to a volume of 5 ml. Potassium was determined on tissue extracts and plasma by flame photometry.

Sources-Radioactive compounds were products of New England Nuclear Corporation. L-Glutamine was obtained from Calbiochem (A grade).

RESULTS

General Properties

General Mitochondrial Characteristics-Mitochondria in these experiments had a total water space of 2.3 * 0.12 pl/mg of protein and a sucrose space of 1.6 * 0.5 pl/mg of protein. The ratio of the matrix to total water spaces averaged 29% and ranged from 17 to 44% in different preparations. Using 10 mM

succinate as substrate, respiratory control ratios were between 2.7 and 5.2 with a mean of 4.2. With 10 mM pyruvate plus 10 mM malate as substrate, the respiratory control ratios had a mean value of 3.6 and ranged from 2.4 to 4.7. ADP:O ratio was 1.9 f 0.06 with succinate and 2.5 f 0.07 with pyruvate and

malate. Phosphate-independent glutaminase activity, mea- sured in heated extracts (50”) to destroy phosphate-dependent glutaminase activity (22), was less than 1% of the activity of phosphate-dependent glutaminase.

Time Course-Fig. 1 shows the formation and accumulation of labeled glutamate produced from 1 mM [‘%]glutamine by mitochondria from dog renal cortex over a 5-min period. Inside the mitochondria a glutamate concentration of over 4 mM is

achieved within 30 s and a maximum matrix space concentra- tion of over 5 mM occurs at 1 min. The concentration in this space then declines gradually during the ensuing minutes. Simultaneously with its formation inside mitochondria, gluta- mate is transported out of the mitochondria into the medium resulting in a steadily increasing concentration of glutamate in the medium. Although the concentration of glutamate in the medium is only 1 to 10% that in the matrix space, the total amount of glutamate -in the medium quickly exceeds that present inside the mitochondria due to the much greater volume of the medium than of the matrix space. In this experiment, recovered glutamine plus glutamate accounts for over 92% of the total counts present in the mitochondrial pellet, demonstrating the effectiveness of the inhibitors in the medium in preventing significant metabolism of glutamate after it is formed.

Glutamate Formation in Presence of Energy-yielding Substrates--In experiments such as that shown in Fig. 1, glutamate formation from glutamine proceeds without addi- tion of an exogenous source of energy. Table I shows that, addition of succinate to the medium results in an increase of 30% or more in total glutamate formation accompanied by a doubling of the concentration of glutamate in the matrix space. These changes are prevented by antimycin which reduces glutamate formation below the level found in the absence of additional substrate. When both succinate and phosphate are added to the medium further stimulation of glutamate forma- tion occurs resulting in about a 2-fold increase in glutamate formation above the basal level. When cytochrome c, TMPD, and ascorbate, which provide cytochrome oxidase substrates, are added, stimulation of glutamate formation similar to that found with succinate occurs, and cyanide abolishes this effect (Table II). As shown subsequently, high phosphate levels in the medium and in mitochondria can stimulate glutamate forma-

r 0.20

- 0.15

- 0.10

- 0.05

I I I

I ; 3 k 5 TIME - MINS

FIG. 1. Time course of the formation and accumulation of gluta- mate Curmed from L-[“Clglutamine. Medium contained 115 mM KCl, 30 IIIM sucrose, 5 m&i MgSO,, 18 mM Hepes (pH 7.4), 1 mM gluta- mine, 1 mM sodium arsenite, and 1 fig/ml of rotenone. Incubation temperature was 37”; gas phase, 100% 0,. O-O, concentration of glutamate in mitochondrial matrix space; 17- -0, concentration of glutamine in medium.

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8151

TABLE I

Effect of succinate and phosphate on glutamate formation from

glutamine

Standard medium contained 117 rn~ KCI, 18 mM Hepes, 30 mM

sucrose, 0.5 mM MgSO,, 1 mM glutamine, 1 rnM arsenite, and 1 pg/ml of rotenone.

Additions

Experiment 1 Experiment 2

[Gluta- Total

[Gluta- Total

make],” gluta-

mate],” gluta-

mate mate

rnM nmol mM nmol

None 6.9 12.3 7.2 14.4

0.5 mM Succinate 12.2 16.3 15.1 19.5

+ Antimycin” 4.0 a.7 5.4 11.3 0.5 mM Succinate, 5 mh4 14.3 24.5 12.2 26.5

phosphate

+ Antimycin” 4.3 14.6

5 mM Phosphate 5.4 14.8 3.8 16.6

’ [glutamatelr, concentration of glutamate in matrix space. “Mitochondria were preincubated for 1 to 3 min with 0.8 nmol of

antimycin/mg protein.

TABLE II

Effect of cytochrome onidase substrates on glutamate formation from glutamine

Standard medium was the same as in Table I.

Additions

Experiment 1

[Gluta- Total

mate],’ gluta- mate

Experiment 2

[Cluta- Total

mate],a gluta- mate

None 0.03 mM Cytochrome c,

0.3 mM TMPD, 3.7 mM ascorbate

+ ImMCN-

rnM nmol nm nmol

6.1 11.7 6.6 20.3 11.5 14.9 10.9 34.8

4.6 10.4 6.1 19.5

a [glutamate],, concentration of glutamate in matrix space.

tion from glutamine. However, the effects of succinate or of

cytochrome c, TMPD, and ascorbate are not due to modulation

of intramitochondrial phosphate concentration since endoge-

nous levels of phosphate in mitochondria are below those which

cause significant change in glutamate formation. In addition these substrates increase both glutamate formation and ma- trix-glutamate levels, whereas high phosphate increases the former but reduces the latter. Rather, the results obtained by

the addition of these energy-yielding substrates indicate that at least one of the steps in the formation of glutamate from

glutamine in intact mitochondria is an energy-dependent process.

Variation in Glutamine Concentration-Measurement of glutamate formation with concentrations of glutamine between 0.1 and 10 EIM in the medium produces a curve such as that of Fig. 2. This result does not conform to a typical Michaelis-

Menten pattern. When analyzed by the method of Hofstee and Woolf (31) (Fig. 3) the points appear to fall onto two straight lines, with a sharp change in slope at about 4 mM. At high concentrations of glutamine a line with a K, of 5 mM occurs, which approximates the K, obtained with extracted phos- phate-dependent glutaminase. At low concentrations a line with a much lower K, is found. In three experiments least squares analysis of the points obtained with glutamine concen- trations between 0.1 and 4 mM give K, values of0.61,0.56, and

~glutomineI] mM

FIG. 2. Formation and accumulation of glutamate formed from varying concentrations of L-glutamine in medium. Conditions and composition of media were the same as in Fig. 1 except that the concentration of KC1 was reduced as concentration of glutamine increased to keep osmolality constant. l - --O, concentration of glutamate in mitochondrial matrix space; q - - -0, concentration of glutamate in medium.

FIG. 3. Hofstee-Woolf plot of data from Fig. 2 showing relationship between rate of glutamate formation (VI and rate of formation divided by substrate concentration (V/S). K, values are obtained from the slopes of the lines.

0.54 with correlation coefficients of 0.98 or more in each case. The closeness with which the points fall on the derived lines suggests that the kinetics governing glutamate formation in intact mitochondria differ at low and high concentrations of glutamine. Since plasma and renal cortical concentrations of glutamine lie between 0.25 and 2 mM (32), it would appear that the K, obtained with the lower concentrations of substrate more accurately describes the kinetics of glutamine metabo- lism under physiological conditions.

pH Optimum-When the pH of the medium is varied from

6.6 to 9.0 a gradual increase in the concentration of glutamate inside the mitochondria occurs (Fig. 4). In three experiments of this type, total glutamate formation showed a maximum between pH 7.4 and 7.8. In contrast the pH optimum for ammonia production by extracted phosphate-dependent gluta- minase from dog kidney was reported to be over 8 by Sayre and Roberts (33), and later was determined to be 9.2 by Mattenheimer and DeBruin (34).

Influence of Glutamate, Ammonium and Phosphate-The

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8152

/* ./ /

O/ PH

FIG. 4. Effect of medium pH on the formation and accumulation of glutamate formed from 1 mM [“C]glutamine. Conditions and composi- tion of media were the same as in Fig. .1 except that the media contained 10 mM Hepes/lO mM Tris buffer adjusted to the pH shown. l - -0. concentration of glutamate in matrix space; 04, concen- tration of glutamate in medium.

products of the glutaminase reaction, glutamate and ammonia, are inhibitors of extracted phosphate-dependent glutaminase (33, 35). In the intact mitochondrial system glutamine metabo- lism is progressively inhibited as the concentration of ammo- nium in the medium is raised from 0 to 10 mM. Both total glutamate formation and the concentration in the matrix are reduced by increasing ammonium concentration (Fig. 5) with a slight effect being present even at the lowest concentration tested (0.05 mM).

Fig. 6 shows a similar experiment in which the concentration of glutamate in the medium is varied. In this case [3H]gluta- mate was used so that the total glutamate in the matrix space could be separated into glutamate derived from glutamine ([‘“C]glutamate) and that derived from medium glutamate transported into the matrix ( 13H]glutamate). Increasing gluta- mate in the medium causes a progressive rise in [3H]glutamate concentration in the matrix, the level in this compartment being 5 to 10 times that outside the mitochondria. The level of [“C]glutamate in the matrix progressively declines but total matrix space glutamate concentration shows only a small change until the glutamate in the medium exceeds 2 mM.

When glutamine and glutamate concentrations in the medium are equal at 1 mM, matrix space glutamate is composed of approximately equal amounts from each source. Total [“Clglutamate formation is inhibited by increasing glutamate in the medium. Fifty per cent inhibition of glutamate forma- tion occurs at, a medium glutamate concentration of about 4 mM with an accompanying level in the matrix space of 20 mM.

In early studies of extracted glutaminase, phosphate was shown to be helpful in improving stability of the enzyme, and the rate of activity of the enzyme was greatly enhanced by high levels of phosphate in the medium (33, 35, 36) leading to the designation “phosphate-dependent glutaminase”. This stimu- latory effect is also exhibited in the intact mitochondrial system by high concentrations of phosphate in the medium, but at low concentrations, only a minor influence is exerted. Fig. 7 shows the effect of varying concentrations of [33P]phos- phate on glutamate formation from glutamine. Levels of phosphate in the matrix space of 4 to 20 times those in the medium occur. As phosphate concentration rises, matrix space glutamate levels decline, suggesting stimulation of glutamate exit by phosphate. Despite the marked fall in matrix space glutamate, glutamate formation is little altered until phos-

0 25 .50 .75 1.00 5.00 10.00

[NH:] mM

FIG. 5. Effect of varying concentrations of NH,+ on glutamate formation and accumulation with 1 mM glutamine as substrate. O-- -0, concentration of glutamate in matrix space; O-O, concentration of glutamate in medium.

[3H-glutomote]MEDIUM mM

FIG. 6. Effect of varying concentrations of glutamate on glutamate formation from glutamine. Media contained (3H]glutamate at concen- trations shown and 1 mM [‘*C]glutamine. O---O, concentration of [“C]glutamate in medium; l - -0, concentration of [I’C jgluta- mate in matrix space; A---A, concentration of [3H]glutamate in matrix space. 0pen circles (not connected) represent total glutamate concentration in matrix.

phate concentration exceeds 5 mM, when a marked increase occurs.

Influence of Other Agents-Phosphate-dependent glutamin- ase is inhibited by a number of sulfhydryl agents including mersalyl, p-CMB, and N-ethylmaleimide (33). Bromcresol green and similar dyes comprise another group of inhibitors. In the intact mitochondrial system mersalyl and p-CMB, the effects of which are described in the next section, are also potent inhibitors of glutamate formation. One millimolar N-ethylmaleimide completely blocks extracted phosphate- dependent glutaminase activity (33) but has only moderate inhibitory action in mitochondria, as shown in Fig. 8. One millimolar bromcresol green also has a powerful action on the soluble enzyme (33) and is only partly inhibitory in mitochon- dria. Borate has the property of stabilizing phosphate-depend- ent glutaminase (37) but competitively inhibits its action on glutamine (38). In the absence of phosphate, 10 mM borate has been reported to inactivate phosphate-dependent glutaminase (35). However, in mitochondria borate increases the rate of glutamate formation (Fig. 8).

Volume of Distribution of G&amine-In experiments like

that of Fig. 1 and others described earlier in this paper, the volume of distribution of glutamine is less than that of the

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

FIG. 7. Effect of varying concentrations of phosphate on formation and accumulation of glutamate formed from glutamine. Media con- tained 33P at concentrations shown and 1 mM [“Clglutamine. O---O, concentration of glutamate in medium, x-x, concentration of glutamate in matrix space; O---O, concentration of total 31P uptake in matrix space.

FIG. 8. Influence of N-ethylmaleimide (NEW, bromcresol green mM NEM I mM KG 10 mM BORATE

(BCG), and borate on formation and accumulation of glutamate formed from glutamine. Media contained 1, 2, or 10 mM glutamine and 1 mM N-ethylmaleimide, bromcresol green, or 10 mM sodium borate. [Glutamate],, concentration of glutamate in matrix space.

outer space of the mitochondria. The faster the rate of glutamate formation the greater this discrepancy. For example in experiment 1 of Table I, in mitochondria incubated without additions to the medium, the volume of distribution of glutamine is 67% of the simultaneously determined mannitol space. In the presence of succinate, glutamate formation is significantly increased, and the volume of distribution of glutamine in this experiment equals 61% of the mannitol space. These low volumes of distribution are the result of metabolism of glutamine during passage of the mitochondria through the silicone oil at which time the mitochondria are isolated from the medium but continue to metabolize actively. Although the transit time is only a few seconds, by the time the mitochon-

8153

dria reach the perchloric acid layer, conversion of glutamine to glutamate has been sufficient to reduce the amount of gluta- mine remaining to less than the amount present in the outer space before centrifugation. These findings differ from those reported in pig kidney mitochondria by Crompton and Chap- pell who describe an experiment in which labeled glutamine appeared to be present in the matrix (16). In our hands using incubation conditions identical to those of these authors we could not demonstrate glutamine in the matrix space of pig renal cortex mitochondria.’

In order to study the true volume of distribution of gluta- mine, conditions were sought in which glutamine conversion to glutamate is greatly reduced so that metabolism during passage through the silicone oil layer does not alter the amount of glutamine in the mitochondria. We found that in the presence of mersalyl or p-CMB or when incubated at 0” only a small amount of glutamate is formed. Table III shows the results of eight experiments in which glutamine metabolism was compared in the presence or absence of 1 mM mersalyl. The latter results in over 90% inhibition of glutamate formation. A slight increase in the total water space occurs in the presence of mersalyl but there is no significant change in the sucrose space. The volume of distribution of glutamine is 93% that of sucrose indicating that no glutamine is present in the matrix space.

Similar distribution of glutamine in the presence of almost complete inhibition of glutamate formation occurs when p- CMB is present in the medium. However, the volumes of spaces of dog kidney mitochondria are usually expanded by this agent making interpretation of the results more compli- cated (see below).

Fig. 9 shows an experiment at 0” in which the volume of distribution of glutamine is compared with that of water, sucrose, and mannitol. In this experiment glutamate formation is too small to measure. The volumes of distribution of glutamine, sucrose, and mannitol are identical and do not change appreciably with time. Thus when glutamate formation is prevented by low temperature, glutamine does not enter the matrix space.

In addition to providing information on the penetration of glutamine these studies on various volumes of distribution in mitochondria also demonstrate the need for great care in performing and interpreting such experiments. In some in- stances at 0” or with concentrations of p-CMB of 0.5 mM or

more (but in none of the mersalyl experiments) the volume of distribution of glutamine exceeds that of sucrose. From these findings alone one might conclude that glutamine enters the matrix space. However, in these circumstances the mannitol space also exceeds the sucrose space indicating that change in the inner membrane occurred prior to or during incubation allowing partial penetration of mannitol (M,, 182) and gluta- mine (M,, 146) but not of sucrose (M,, 342). Occasionally mitochondria incubated at 37” without p-CMB or mersalyl have mannitol spaces equal in volume to the total water space although the sucrose space is normal in size. In other re- spects these mitochondria show no evidence of altered integrity as judged by respiratory control ratios, ADP:O ratios, permea- bility to NADH, or ability to accumulate glutamate in the matrix space. Thus subtle alteration in the inner membrane can occur which is not reflected by most measurements of mitochondrial performance but which results in increased permeability to small molecules. This problem arises com-

z W. Adam and D. P. Simpson, unpublished results.

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8154

TABLE III 5.07

Effect of mersalyl on glutamine distribution and glutamate formation”

No mersalyl 1 mM mersalyl

V, of’glutamine (% of sucrose

space) Glutamate formed (nmol/mg pro-

tein/2 min) Sucrose (rllmg protein) space Total water (&mg protein) space

93 zt 1.6

32 i 5 2.7 + 0.6

1.76 + 0.12 1.81 * 0.09

2.56 + 0.14 2.99 + 0.16

On = 8.

Volume of Dlstrrbulwn I” Dog Kidney Mitochrondrla ot 0’

5 15

‘0 &

HZ0

Q 1.0 Monnltoi F

5

*

I::_;

S”crose

Q F

0.5 &--o/------4

? 52

.5 1 1.5 2 Minutes

FIG. 9. Volumes of distribution of water, sucrose, mannitol, and L-glutamine in mitochondria from dog renal cortex incubated for varying periods of time at 0”. O- -0, the ratio of the volume ot distribution of glutamine (S,,) to that of mannitol (S,), shown separately to avoid superimposing several lines. O-0, total water space: O-O, sucrose space; A-A, mannitoi space.

monly with dog but rarely with rat kidney mitochondria. Homogenization of dog renal cortex requires considerably more force and higher speed than for rat kidney, factors which may be important in producing this change in permeability. How- ever, attempts to make homogenization gentler or otherwise to alter the preparative procedure or composition of the medium were unsuccessful in preventing this change; regard- less of the technique used the mannitol space varied from one preparation to another, in some cases equaling the sucrose space and in other cases exceeding it to a variable extent. Since preparations can be obtained in which the sucrose, mannitol, and glutamine spaces are all identical at O”, as shown in Fig. 9, we conclude that the cases in which these spaces do not coincide so that glutamine appears to be present in the matrix represent artifacts of preparation.

Effects of Acidosis and Alkalosis

Glutamate Formation from Glutamine-Mitochondrial glu- tamate formation from glutamine was measured in four experiments with acidotic and alkalotic litter mate dogs. Fig. 10 shows the concentration of glutamate in the matrix space and in the medium in these experiments. Total gluta- mate formation and concentration inside the mitochondria is about twice as great in the acidotic as in the alkalotic group. Thus mitochondria from chronically acidotic dogs have an intrinsically greater ability to form glutamate from glutamine. In contrast to the greater amount of glutamate present inside the acidotic mitochondria, the amount of glutamine in the mitochondrial extracts after separation from the medium is less in acidosis. The average glutamine in this fraction in these four experiments is 1.22 * 0.06 nmol/mg of protein in the acidotic animals and 1.39 h 0.06 in the alkalotic ones. As noted in the preceding section, this glutamine is present in the outer space of the mitochondria. The differ-

3.0

% 2.0

I

1 .o -’

i

0.051

[glutamate] [glutornate]

Inside n:4 I” medium mrtochondra

FIG. 10. Comparison of the effect of chronic metabolic acidosis and alkalosis on the mitochondrial formation and accumulation of gluta- mate formed from glutamine. Medium contained 97 mM KCI, 30 mM sucrose, 18 mM Hepes (pH 7.4), 20 mM KHCO,, 0.5 mM MgSO,, 1 mM (“C]glutamine, 1 rnM sodium arsenite, and 1 &g/ml of rotenone. Incubation time was 2 min; gas phase, 96% O,, 4% CO,; temperature, 37”.

ence between the amounts found in the two groups reflects the more rapid rate of glutamate formation in the acidotic mitochondria which consumes more glutamine during passage of the mitochondria through silicone oil,

In seven similar experiments the accumulation of glutamate in mitochondria was measured with various alterations in the composition of the medium. The total amount of glutamate present in the medium was determined in only a few of this group of experiments but when measured the changes paral- leled those present in the concentration of glutamate in the matrix space. Fig. 11 shows the results of these experiments. With medium containing glutamine as the only substrate, whether buffered with Hepes or HCO,-, the concentration of glutamate in the matrix is significantly greater in the acidotic group. When 2 mM pyruvate, 2 mM malate, and 1 mM citrate are also present a similar magnitude of glutamate accumula- tion occurs, and the concentration is about twice as great in the aridotic as in the alkalotic preparations. When 0.5 mM succinate is available as a source of energy the accumulation of glutamate is markedly increased, but a significant differ- ence is still present between uptake by mitochondria from acidotic and alkalotic dogs.

The experiments shown in Fig. 11 also provide information on certain other general properties of mitochondria from acidotic and alkalotic dogs. Mitochondrial water spaces and oxygen electrode data from these experiments are given in Table IV. The average outer (sucrose) spaces, total water spaces, and the calculated matrix space volumes are similar in size in the two different acid-base states. The respiratory control ratios found with 10 mM pyruvate plus 10 mM malate are significantly greater in the acidotic mitochondria (p<O.O5). This difference is derived from a greater rate of oxygen con- sumption in the acidotic mitochondria in the presence of ADP (State 3) (82.3 * 7.9 J/mg of protein/min in acidosis versus 62.0 * 5.3 pl/mg of protein in alkalosis). The rates of oxygen consumption with pyruvate and malate after the ADP was consumed (State 4) are not significantly different (16.8 *

1.2 and 16.7 * 0.6 pl/mg of protein/min, respectively). The similarity of these general properties in the two groups of mitochondria indicate that the differences in glutamate forma- tion and accumulation shown in Figs. 10 and 11 cannot be accounted for by intrinsic differences in the quality of the

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8155

0 NOIE t 20 mM HCO; Pyruvate, Succlnote

Malate. citrate;

Additions to medium (1mM glutamlne in all)

FIG. 11. Concentration of glutamate in matrix space of mitochon- dria from chronically acidotic and alkalotic animals incubated with 1 mM glutamine and different substrates or buffers. Composition of the control medium was similar to that given in Fig. 10 except that KHCO, was replaced by KCl; gas phase was 100% 0,. All other media contained 20 mM KHCO, and were gassed with 96% O,, 4% CO,. Where noted 2 mM pyruvate, 2 mM malate, and 1 mM citrate or 0.5 mM succinate were present in the medium in addition to glutamine. n = 7.

TABLE IV

General properties of mitochondrial preparations from acidotic and

alkalotic animals”

Acidosis Alkalosis

Outer wl/mg protein space, Total water pl/mg protein space,

Matrix hl/mg protein space, Respiratory control ratio (citrate)

Respiratory control ratio (malate and pyruvate)

ADP:O (citrate) ADP:O (malate and pyruvate)

1.65 i 0.06 1.65 + 0.11 2.53 i 0.12 2.49 * 0.16

0.90 * 0.07 0.84 + 0.05 4.30 * 0.40 4.00 + 0.60

5.00 * 0.70 3.70 + 0.32’

2.50 A 0.10 2.62 i 0.12 2.33 zt 0.08 2.47 + 0.10

an = 7.

bp <0.05.

mitochondria, their size, or their general biochemical behavior. Endogenous glutamine was less than 4 nmol/mg of protein in

isolated mitochondria. Glutamate levels varied between 2 and 9 nmol/mg of protein. No consistent differences in endogenous glutamine or glutamate content were noted between prepara- tions from acidotic and alkalotic dogs.

Ammonia Production-As a corollary to the above studies, production of ammonia by mitochondria from acidotic and alkalotic animals was measured in five experiments. In these the medium contained 1 mM oxalacetate in addition to glutamine, and incubation times of 5 and 10 min were used. The oxalacetate provided substrate to remove glutamate by transamination in order to reduce inhibition of glutaminase activity. The results are similar to those obtained when glutamate formation and accumulation are measured. Fig. 12 (left) shows that the acidotic mitochondria produce 75% more ammonia per min than the alkalotic ones.

Phosphate-dependent Glutaminase Levels-Lyophilized mi- tochondria in 13 experiments were extracted with borate and assayed for phosphate-dependent glutaminase activity. Fig. 12

(right) shows that the content of extractable phosphate- dependent glutaminase in the acidotic and alkalotic prepara- tions is almost identical, 325 + 23 and 337 * 24 units/mg of protein in the two cases, respectively. Thus the difference in glutamate formation and accumulation by mitochondria from

NH3 formatlon nmoles/m~n/mg protein Glutammose actlwty

nmoles NH, /mln/mg protein

FIG. 12. Ammonia formation by intact mitochondria (left) or by extracted phosphate dependent glutaminace (right) from chronically acidotic or alkalotic animals. Assay methods are described in the text.

acidotic and alkalotic dogs does not arise from a change in the amount of extractable phosphate-dependent glutaminase.

Role of Potassium Depletion-Ammonium chloride adminis- tration causes a considerable K+ loss in the urine leading in some instances to K+ depletion (39). Since K+ depletion as well as acidosis stimulates ammonia production in the kidney (40, 41), it is possible that the alteration in glutamine metabolism found in these experiments is induced by changes in K+ balance. To investigate this tissue and plasma K+ levels were measured in three experiments without supplemental potassium chloride in the diet and in two experiments in which 4 meq/kg/day of potassium chloride were added to the diet of the acidotic animals. Without K+ supplementation the plasma K+ is lower in the acidotic animals by 0.56 to 0.68 meq/liter. After K+ supplementation the plasma levels were 0.3 and 0.1 meq higher in the acidotic dogs. Without supplementation tissue potassium levels are identical in both acidotic and alkalotic renal cortex, averaging 0.30 f 0.1 and 0.29 * 0.01 meq/g dry weight, respectively. In the two pairs of dogs given K+ supplements, the tissue K+ levels in acidosis and alkalosis were also identical, 0.30 and 0.29 meq/g in the first pair and 0.27 and 0.30 meq/g in the second. In these two experiments glutamate levels were 71 and 55% higher in the acidotic than in the alkalotic mitochondria, differences comparable to those found without added potassium in the diet. Thus supplement- ing the diet of the acidotic dogs with potassium chloride does not affect the ability of their kidney mitochondria to form more glutamate from glutamine. Hence the stimulation of glutamine metabolism in acidosis is not mediated by low potassium levels but rather appears to be induced by the change in acid-base state itself.

Glutamine Metabolism by Liver Mitochondria-Since kid- ney is the only organ known to increase ammonia production in response to acidosis, we examined mitochondria from liver to see if any difference in glutamine metabolism is present in acidosis compared to alkalosis. Table V compares glutamate formation and accumulation in mitochondria isolated from renal cortex and liver of three pairs of dogs. In contrast to kidney, liver mitochondria produce only small amounts of glutamate from glutamine; less than 5% as much glutamate is produced by liver mitochondria as by kidney mitochondria in the experiments shown. The rate of glutamate formation is slow enough in liver for the rate of exit of glutamate from the matrix space to remove glutamate as rapidly as it is produced so that no detectable glutamate is present inside the mitochondria. Table V also shows that while acidosis

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8156

TABLE V

Glutamate formation and accumulation by liverand kidney mitochondria from acidotic and alkalotic dogs

Values are results from individual flasks.

Total glutamate [glutamate]

in matrix

Liver Kidney Liver Kidney

Experiment I

Acidotic Alkalotic

Experiment 2

Acidotic

Alkalotic Experiment 3

Acidotic Alkalotic

nmolhg protein mhf

2.1, 1.7 58,58 0 2.8, 2.8

2.0, 1.4 33, 34 0 1.6, 1.5

0.8, 0.9 50,48 5.9, 5.3

0.7, 0.6 24, 22 0 2.6, 2.4

1.6, 1.3 36,34 0 3.3, 3.2

2.1, 1.5 22,22 0 2.2, 2.0

stimulates kidney glutamate formation, the acid-base state of the animal does not influence the rate of glutamine metab- olism by liver mitochondria.

ulated in the matrix space or was transported out of the mitochondria into the medium by the glutamate transporter. These conditions enable study of glutamate formation from glutamine under conditions which preserve the normal rela- tionship of phosphate-dependent glutaminase to other mito- chondrial components.

In five experiments on liver the volume of distribution of glutamine was found to approximate that of the simultaneously determined mannitol space. In acidotic mitochondria the ratio of the glutamine to mannitol spaces averaged 1.04 f 0.08; for alkalotic mitochondria this value was 1.01 + 0.03. Suc- rose, total water, and matrix space volumes for the acidotic liver mitochondria were 2.14 i 0.08, 3.25 G= 0.09, and 1.11 f 0.03 &mg of protein, respectively. For the alkalotic prepara- tions these values were 1.96 f 0.17, 2.94 h 0.20, and 0.98 f 0.04 &mg of protein.

Glutamate Formation by Submitochondrial Preparations-

In four experiments digitonin and Lubrol preparations of mito- chondria from acidotic and alkalotic animals were prepared. The digitonin preparations were essentially free of outer mem- brane as judged by monoamine oxidase activity. After treat- ment with Lubrol, the matrix activity, determined with malate dehydrogenase, was reduced by about 50%. Thus the Lubrol preparations represent inner membrane activity with some residual contamination by matrix. Lubrol treatment caused partial loss of phosphate-dependent glutaminase activity so that these preparations retained only about 50% of the original enzyme.

Table VI shows the formation of glutamate from glutamine by these submitochondrial preparations. In order to compen- sate for loss of phosphate-dependent glutaminase activity in the Lubrol fractions the results are expressed both as nano- moles per mg of protein and as nanomoles per unit of residual phosphate-dependent glutaminase activity in the preparation. In each type of preparation the difference in glutamate forma- tion between fractions from acidotic and alkalotic animals is still present so that the acidotic preparations produce about twice as much glutamate as the alkalotic ones. Thus, despite removal of the outer membrane and most of the matrix space, inner membrane preparations from acidotic dogs retain their increased ability to form glutamate.

The properties of glutamate formation in this intact mito- chondrial preparation differ significantly in many ways from those of soluble phosphate dependent glutaminase. One of the most striking of these differences is the dependence of glutamine metabolism in mitochondria on energy supply. While glutamate formation occurs in mitochondria without addition of an exogenous energy source, addition of energy- yielding substrates considerably enhances the reaction. Others have previously noted the stimulation of glutamate formation from glutamine by succinate (15, 16). This effect has been interpreted as the result of energy dependence of glutaminase. However, extracted phosphate-dependent glutaminase does not require a cofactor or high energy intermediate so there is no known property of the enzyme which clearly provides a means of coupling to an energy source. Another important characteristic of glutamate formation in mitochondria is the K, of about 0.6 mM observed with concentrations of gluta- mine below 4 mM, which includes the levels present in oiuo.

This value contrasts with the K, of 5 mM found in ex- tracted phosphate-dependent glutaminase and at high levels of glutamine in intact mitochondria.

DISCUSSION

Extracted phosphate-dependent glutaminase requires par- ticular conditions of extraction and assay in order to retain high activity (33-36). In contrast efficient glutamate formation in intact mitochondria occurs with low concentrations of substrate in the absence of added phosphate at pH 7.4, condi- tions resembling those present in uiuo but poorly suited for activity of soluble phosphate-dependent glutaminase. Phos- phate which is essential for high activity of extracted phos- phate-dependent glutaminase also stimulates glutamate for- mation in intact mitochondria when present at concentrations in excess of 5 mM, but has little effect below that concentra- tion. The effect of some inhibitors of phosphate-dependent glutaminase is also modified in the mitochondrial system. While mersalyl or p-CMB strongly inhibits glutamate for- mation in both situations, 1 mM N-ethylmaleimide or brom- cresol green has only a modest inhibitory effect in the intact preparation, but each is a potent inhibitor of extracted phos- phate-dependent glctaminase (33).

General Properties of Mitochondrial Glutamate Formation The products of the glutaminase reaction, ammonia and --In order to study glutamine metabolism in isolation from glutamate, inhibit both soluble phosphate-dependent gluta- subsequent metabolic steps, in the experiments described minase and the mitochondrial reaction (33, 35). When 1 mM here metabolism of glutamate was blocked by rotenone and glutamine is present in the medium without added gluta- arsenite. The glutamate formed from glutamine either accum- mate, as in most of the experiments reported here, the ac-

TABLE VI

Glutamate formation by mitochondria and submitochondrial

preparations from acidotic and alkalotic dogs”

Glutamate formed Preparation

Acidotic Alkalotic Acidotic Alkalotic

nmollmg protein nmollunit glutaminasea

Mitochondria 59.7 s 2.3 30.7 * 3.1 708 i 146 436 i 87

Digitonin 63.4 * 5.9 37.6 + 2.7 753 i 167 413 * 132

Lubrol 9.8 + 2.0 4.4 * 0.7 170 i 24 91* 21

“n = 4.

‘Phosphate-dependent glutaminase activity.

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8157

companying levels of glutamate in the medium and matrix are insufficient to exert much effect on glutamate formation so that under these conditions the results are independent of inhibitory influence of glutamate. Indeed, except when the influence of added phosphate was studied, whenever gluta- mate formation was increased so was the level of glutamate in the matrix. The possibility that glutamate exit is linked to glutamate formation so that an increase in the rate of gluta- mate exit leads to a corresponding increase in glutamate formation has been suggested by others (16). However, the present results do not support this hypothesis. In Fig. 5, between phosphate levels of 0 and 5 mM, a fall in matrix glutamate of 50% occurs, implying a significant increase in the rate of glutamate exit; almost no increase in the rate of gluta- mate formation accompanies this change. Consequently the rate of glutamate exit does not appear to be a rate-limiting factor in glutamine metabolism at physiological concentrations of glutamine and a close linkage between these processes is unlikely.

The contrasting properties of glutamate formation by intact mitochondria and by extracted phosphate-dependent gluta- minase indicate that there is a distinct component of gluta- mate formation in mitochondria which is lost when the enzyme is removed from its normal location. Recently, evidence has been present which localizes phosphate-dependent gluta- minase activity to the inner membrane in rat kidney mitochon- dria (7) rather than in the matrix space as others have concluded (6). Our evidence from inner membrane prepara- tions of dog kidney mitochondria suggests that phosphate- dependent glutaminase is loosely bound to the inner mem- brane, but conclusive evidence on this point is difficult to obtain because of the decreased recovery of glutaminase activity after Lubrol treatment. The location of phosphate- dependent glutaminase in the matrix space or in close associa- tion with the inner membrane raises the possibility that the inner membrane contains a glutamine transport system as it does for many substrates of mitochondrial enzymes (42).

Many of the properties demonstrated in the present study suggest the existence in the inner membrane of a specific glutamine transport mechanism which would also provide a reasonable explanation for the more outstanding differences between the properties of glutamate formation by mito- chondria and by phosphate-dependent glutaminase. Since glutamate formed from glutamine appears in the matrix and reaches a concentration of over 5 times that of medium glutamine within seconds after starting incubation, it is clear that glutamine or at least its carbon skeleton readily passes through the inner membrane. On the other hand, failure of glutamine to penetrate the matrix space at 0” or in the presence of mersalyl demonstrates that the inner membrane is not passively permeable to glutamine. Consequently the steps in formation of glutamate must include one in which glutamine is picked up on the outer surface of the inner membrane, transported across this membrane, and delivered to glutamin- ase. In addition, stimulation of glutamate formation by energy-producing substrates is a property common to inner membrane transport systems. The low K, at physiologic sub- strate concentrations and the pH optimum of 7.4 to 7.8 are also compatible with a carrier moiety involved in glutamate formation. Therefore we postulate that glutamate formation in intact mitochondria is a two-step process involving first the uptake of glutamine from the medium by an inner mem- brane component and second its conversion by glutaminase

to glutamate and ammonia. This inner membrane component or carrier for glutamine

apparently differs from most inner membrane transport mech- anisms in that it does not lead to accumulation of its substrate in the matrix space. Our failure to detect glutamine in the matrix when glutamate formation is inhibited suggests that glutamine as such does not enter the matrix space. However the experiments with intact mitochondria do not entirely exclude the possibility of direct entry of glutamine into the matrix; if glutamine were released into the matrix before deamidation both mersalyl and low temperature might inhibit glutamine transport as much as or more than phosphate- dependent glutaminase activity, preventing detectable accu- mulation of glutamine even when phosphate-dependent gluta- minase activity is reduced. However, as noted below, results obtained in purified inner membrane preparations from aci- dotic and alkalotic animals favor the interpretation that glutamine is not released into the matrix space before deami- dation.

The precise relationship in the inner membrane of the carrier and phosphate-dependent glutaminase is not defined by these studies. Both the transport and enzyme functions could be provided by a single molecular unit in the inner membrane, and the designation “carrier” as used here does not imply that this component is necessarily separate from glutaminase. Leaving aside this question of the exact relationship between its two components, the results show that the inner membrane of mitochondria from dog renal cortex contains a rate-limiting, energy-dependent, functional unit which is responsible for picking up glutamine on the outer surface of the inner membrane, transporting it across the membrane, deamidating it, and releasing glutamate into the matrix space.

Effects of Metabolic Acidosis and Alkalosis-The results of the experiments on mitochondria from chronically acidotic and alkalotic animals provide the first demonstration in dog kidney of a direct effect of acidosis on glutamate formation from glutamine. Previous studies have shown that in the dog (3, 43), unlike the rat (44, 45), phosphate-dependent glutaminase levels do not increase in chronic metabolic acidosis. Among the possibilities suggested previously for control of renal am- moniagenesis by acidosis, regulation of phosphate-dependent glutaminase activity by alteration of end product inhibition by glutamate has received considerable attention (9). The results of the present study show that glutamate formation by mitochondria in acidotic preparations is stimulated despite the presence of increased matrix space and medium levels of glutamate, which if anything should repress glutamate forma- tion. Consequently removal of end product inhibition is not necessary for the action of acidosis on glutamine metabolism. Since glutamate levels in renal cortex are known to fall in aci- dosis (46, 47), removal of end product inhibition may add some additional impetus to glutamate formation. But the results of the present study demonstrate that the primary stimulatory action of acidosis is on a step preceding formation of glutamate.

Taken in conjunction with the results discussed in the pre- ceding section the experiments described here indicate that the site of action of acidosis in the dog is on the mitochondrial carrier-glutaminase steps in glutamine metabolism. Since extracted phosphate-dependent glutaminase levels are un- changed, acidosis must activate the carrier step. We conclude that chronic metabolic acidosis increases the uptake of gluta- mine from the medium and its delivery to phosphate-depend- ent glutaminase; increased ammonia formation then results

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

8158

from the action of glutaminase on its substrate. Thus the physiologic phenomenon of increased excretion of ammonium in the urine during metabolic acidosis is a reflection of a unique biochemical regulatory mechanism whereby the rate of product formation by an enzyme is determined by gradual adaptation in the rate of delivery of substrate to the enzyme.

This conclusion is supported by the results obtained using submitochondrial preparations in which increased glutamate formation persists in material from acidotic animals. In Lubrol preparations the inner membrane is thought to be inverted so that the inner surface faces the medium (20). It is likely that in the process of this rearrangement the permeability of the inner membrane is altered so that both surfaces of the membrane are exposed to medium glutamine after Lubrol treatment. The fact that we can still show activation of glutamate formation in acidosis in these preparations suggests that the carrier-gluta- minase relationship is preserved after Lubrol treatment and that as long as glutaminase is attached to the inner membrane it can only act on glutamine delivered to it from the outer surface by the carrier. This same finding also suggests that glutamine is not released by the carrier into the matrix space in intact mitochondria but rather is delivered directly to gluta- minase.

Acidosis could increase the activity of the glutamine carrier either by increasing the number of carrier sites or by increasing the rate of transport by existing carriers. The gradual increase in ammonium excretion after acidosis commences is compati- ble with a process in which time is necessary to permit synthesis of new carrier molecules. However, it is also compati- ble with a mechanism in which increased energy is gradually made available to the glutamine carrier. Such a mechanism would require adaptation of a specific process channeling energy to this carrier since basal rates of oxygen consumption are the same in acidotic and alkalotic mitochondria, a finding which suggests that net energy production is the same in the two groups of mitochondria. We noted earlier in the discussion that the carrier and phosphate-dependent glutaminase could be distinct molecules or that a single molecule might serve both functions. In the event that only a single molecule is involved, another explanation of the mechanism of adaptation could be postulated. If phosphate-dependent glutaminase exists in active and inactive forms, as in the case of phosphoryl- ase, acidosis might increase the amount of the active form and result in both increased transport and deamidation of gluta- mine. Since extracted phosphate-dependent glutaminase lev- els do not change, this possibility would imply that the assay for phosphate-dependent glutaminase measures both the ac- tive and inactive forms of the enzyme. In the absence of any other evidence for active and inactive forms of phosphate- dependent glutaminase, and since increase in transport of the glutamine carbon skeleton is clearly involved in the response to acidosis, it seems more reasonable at present to consider the adaptation one of the carrier moiety.

Acknowledgments-We wish to thank Gloria Roton and Susan Brown for competent technical assistance.

REFERENCES

1. Pitts, R. F., Pilkington, L. A., and De Haas, J. C. M. (1965) J. Clin. Invest. 44.731-745

2. Richterich, R. W., and Goldstein, L. (1958) Am. J. Physiol. 195, 316320

3. 4.

Rector, F. C., Jr., and Orloff, J. (1959) J. Clin. Inuest. 38, 366-372 Crompton, M., McGivan, J. D., and Chappell, J. B. (1973) Bio-

them. J. 132, 27-34 5. KovaEeviC Z. (1974) Biochim. Biophys. Acta 334, 199-207 6. Kalra, J., and Brosnan, J. T. (1974) J. Biol. Chem. 249,3255-3260 7. Curthovs. N. P.. and Weiss. R. F. (1974) J. Biol. Chem. 249,

326113266 8. 9.

Simpson, D. P. (1971) Medicine 50, 503-539 Pitts, R. F. (1973) in Rena2 Physiology (Orloff, J., and Berliner, R.

W., eds) pp. 455-496, American Physiological Society, Washing- ton, D. C.

10. 11.

Alleyne, G. A. 0. (1970) J. Clin. Inuest. 49, 943-951 Kamm, D. E., and Strope, G. L. (1973) J. Clin. Inuest. 51,

1251-1263 12. 13. 14.

Simpson, D. P. (1972) J. Clin. Inuest. 51, 1969-1978 Orloff, J., and Burg, M. (1971) Annu. Reu. Physiol. 33, 110-121 Hird, F. J. R., and Marginson, M.. A. (1968) Arch. Biochem.

Biophys. 127, 718-724 15. KovaEeviE, Z., McGivan, J. D., and Chappell, J. B. (1970)

Biochem. J. 118, 265-274 16. 17. 18.

19. 20.

Crompton, M., and Chappell, ,J. B. (1973) Biochem. J. 132, 35-46 Adam. W.. and Simoson. D. P. (1974) J. Ctin. Invest. 54. 166-174 Chappell, J. B., and Hansford, R. G. (1972) in Subcellular Com-

ponents. Preparation and Fractionation (Birnie, G. D., ed) 2nd Ed., pp. 77-91, University Park Press, Baltimore

Harris, E. J., and Van Dam, K. (1968) Biochem. J. 106,759-766 Chan, T. L., Greenawalt, J. W., and Pederson, P. L. (1970) J. Cell

Biol. 45, 291-305 21. 22.

23. 24.

Estabrook, R. W. (1967) Methods Enrymol. 10, 41-47 Curthoys, N. P., and Lowry, 0. H. (1973) J. Biol. Chem. 248,

162-168 Goldstein, L. (1967) Am. J. Physiol. 213, 983-989 Sherrard, D. J., and Simpson, D. P. (1969) J. Lab. Clirz. Med. 73,

877-882 25.

26.

27.

28.

29. 30. 31.

32.

Curthoys, N. P., and Lowry, 0. H. (1973) Am. J. Physiol. 224, 884-889

Schnaitman, C., and Greenawait, J. W. (1968) J. Cell Biot. 38, 158-175

Allmann, D. W., Bachmann, E., Orme-Johnson, N., Tan, W. C., and Green. D. E. (1968) Arch. Biochem. Biophys. 125,981-1012

Wharton, D. C., and Tzagoloff, A. (1967) Methods Enzymol. 10, 245-250

Ochoa, S. (1955) Methods Enzymol. .l, 735-739 Layne, E. (1957) Methods Enzymol. 3, 447-454 Dixon, M., and Webb, E. C. (1964) Enzymes, 2nd Ed, Academic

Press, New York Shalhoub, R., Webber, W., Glabman, S.. Canessa-Fischer, M.,

Klein, J., De Haas, J., and Pitts, R. F. (1963) Am. J. Physiol. 204, 181-186

33. 34.

Sayre, F. W., and Roberts, E. (1958) J. Biol. Chem. 233, 1128-1134 Mattenheimer, H., and DeBruin, H. (1964) Enzymol. Biol. Clin. 4,

65-83 35.

36.

37.

38.

Klingman, J. D., and Handler, P. (1958) J. Biol. Chem. 232, 369-380

Carter, C. E., and Greenstein, J. P. (1947) J. Nat. Cancer Inst. 7, 433-436

Otey, M. C., Birnbaum, S. M., and Greenstein, J. P. (1954) Arch. Eiochem. Biophys. 49,245-247

Roberts, E. (1960) in The Enzymes (Bayer, P. D., Lardy, H., and MyrbHck, K., eds) 2nd Ed, Vol. 4, pp. 285-300, Academic Press, New York

39. Lennon, E. J., and Lemann, J., Jr. (1968) Clin. Sci. 34, 365-378 40. Adam, W., and Simpson, D. P. (1975) Kidney Int. 7,325-330 41. Kamm. D. E.. and Strone. G. L. (1973) Am. J. Physiol. 224.

1241-1248 42. 43.

Klinaenbere. M. (1970) Essavs Biochem. 6.119-159 Pollik, V. E., Mattenheimer, H., DeBruin, H., and Weinman,

K. J. (1965) J. Clin. Inuest. 44, 169-181 44. Davies, B. M. A., and Yudkin, J. (1952) Biochem. J. 52,4077412 45. Rector. F. C.. Jr.. Seldin. D. W.. and Conenhaver. J. H. (1955)

J. &in. Inuest. 34, 20-26’ 1 46. Goldstein, L. (1966) Am. J. Physiol. 210,661-666 47. Damian, A. C., and Pitts, R. F. (1970) Am. J. Physiol. 218,

1249-1255

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from

D P Simpson and W AdamGeneral properties and response to acidosis and alkalosis.

Glutamine transport and metabolism by mitochondria from dog renal cortex.

1975, 250:8148-8158.J. Biol. Chem. 

  http://www.jbc.org/content/250/20/8148Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/250/20/8148.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on February 21, 2020http://w

ww

.jbc.org/D

ownloaded from