Eur. J. Biochem. 72, 301 - 307 (1977)
Mitochondria1 and Cytosolic NADPH Systems and Isocitrate Dehydrogenase Indicator Metabolites during Ureogenesis from Ammonia in Isolated Rat Hepatocytes Helmut SIES, Theodorus P. M. AKERBOOM, and Joseph M. TAGER
Instirut fur Physiologische Cliemie. Pli) sikalische Biochemie und Zellbiologie der Universitat Munchen, and Laboratory of Biochemistry, B.C.P. Jansen Institute, University of Amsterdam (Received August 4 1 September 28, 1976)
1. Citrate, isocitrate and 2-oxoglutarate levels were determined in isolated rat hepatocytes and in particulate and soluble fractions thereof, obtained by the digitonin and silicone oil fractionation technique. 2. Calculated from isocitrate/2-oxoglutarate ratios (“indicator metabolite method”), the redox potential of mitochondrial free NADPH is -402 mV, whereas that of the extramitochondrial (cytosolic) space is about 10 mV more positive, -392 mV. 3. Addition of ammonia (either as ammonium chloride or from urea plus urease) to isolated hepatocytes causes preferential oxidation of mitochondrial NADPH, as demonstrated by spectrophotometry of the dihydro band and by the changes in the isocitrate/2-oxoglutarate ratios. The redox potential difference of free NADPH between mitochondria and cytosol is abolished or even reversed. 4. It is concluded that during ureogenesis from ammonia mitochondrial isocitrate oxidation is shifted largely in favor of the NADP-linked as opposed to the NAD-linked enzyme; isocitrate concentration under these conditions is less than 10 pM, below the K, (isocitrate) of the NADlinked enzyme but in the range of that for the NADP-linked enzyme. 5. Both in the absence and in the presence of ammonia there is a concentration gradient across the mitochondria1 inner membrane (from mitochondria to cytosol) for citrate, isocitrate, and also, to a smaller extent, for 2-oxoglutarate. 6. These results and data in the literature on enzyme activity are in agreement with the assumption of near-equilibrium of NADP-dependent isocitrate dehydrogenases in the mitochondrial matrix and cytosolic spaces in the absence of ammonia; accordingly, during urea formation from added ammonia the redox potential of mitochondrial free NADPH is increased to -391 mV or possibly even higher if there exists an indicator error under this condition. Isocitrate dehydrogenases occur both intramitochondrially and extramitochondrially in rat liver [l]; NADP-dependent enzymes are present in both spaces, whereas an NAD-dependent enzyme is present only intramitochondrially (see [2] for a review). The clear concept, put forward by Goebell and Klingenberg [3], of the dichotomy of mitochondrial isocitrate oxidation into a non-equilibrium, regulated (or regulatory) pathway catalyzed by the NAD-dependent enzyme and an equilibrium pathway catalyzed by the NADP-dependent enzyme implies that the compoEnzymes. Isocitrate dehydrogenase (NAD+) (EC 1.1.1.41); isocitrate dehydrogenase (NADP+) (EC 1.1.1.42).
nents of the mitochondrial isocitrate dehydrogenase system may be utilized as ‘indicator metabolites’ for the mitochondrial NADP system, provided that the activity is high enough (see below). Regardins the extramitochondrial (cytosolic) space, there are no major complications to utilize the isocitrate dehydrogenase system as an indicator for the redox potential of cytosolic free NADPH. Veech etal. [4] have, in fact, utilized the isocitrate dehydrogenase system for this purpose ; however, their analytical data were the total tissue levels rather than the cytosolic ones. Furthermore, attempts have been made by Williamson [5] and Greenbaum et al. [6] to calculate mitochondrial and cytosolicconcentrations from the total tissue levels
302
Mitochondria1 and Cytosolic NADPH Systems
Shaking frequency was 120 min-'. Samples were of metabolites, based on certain assumptions on the taken after 20 or 25min. In some incubations, subcellular distribution. NH4Cl (10 mM) was added after 20 min. Hoek and Ernster [7] have shown that, under energized conditions, isolated rat liver mitochondria maintain a ratio of ([NADPH]/[NADP'])i,/([NADPH]/ Hepu t ocyte Sump ling and Frac t ionat ion [NADP+]),,, = 2 in the presence of the components The digitonin fractionation procedure was essenof the isocitrate dehydrogenase reaction in the suspentially as described [l5]. At one time point, three sion medium. In a first approximation, this would samples of 1 ml each were taken from the incubation : correspond to a redox potential difference for free sample 1, direct deproteinization with perchloric NADPH of 10mV more negative inside the mitoacid, 0.94 M final concentration, to give the total chondrial matrix space than outside, and it implies contents of the metabolites; sample 2, mixing with that the isocitrate and 2-oxoglutarate carrier systems 5 ml of a separation medium (0.25 M sucrose, 20 mM are capable of bringing about equilibration of these 3-(N-morpholino)propane sulfonate, pH 7.0, 3 mM substrates between the two spaces [7,8]. The present EDTA) at -2 "C containing digitonin ( 2 mM, final investigation was carried out to determine, with the concentration) and, after 15 s, centrifugation through aid of the indicator metabolite system of NADP2 ml of a silicone oil mixture (Wacker Chemie AR 200/ linked isocitrate dehydrogenase (for a recent discusGeneral Electric SF 96/100, S / l , v/v, and in some exsion, see [9]), whether a similar redox potential difperiments, 6/1, v/v) into 2 ml of perchloric acid, ference exists within the intact hepatocyte. For this 0.94 M ; sample 3, same as sample 2, but without purpose, isolated rat hepatocytes were subfractionated digitonin. with the digitonin and silicone oil fractionation techPart of the supernatant fractions of samples 2 and 3 nique of Zuurendonk and Tager [lo] and metabolite were taken for assay of lactate and glutamate dehydropatterns determined. genases as marker enzymes of the cytosolic and miIn perfused liver, urea formation from ammonia tochondrial matrix spaces, respectively. Total enzyme is accompanied by a substantial oxidation of NADPH activities in the incubation were measured after treat[I 1,121 and, indeed, ureogenesis from ammonia has ment of a sample with 0.5% lubrol WX, and were been identified as a major NADPH-requiring process taken as reference ( = 100%). For metabolite deterof the mitochondrial space in isolated hepatocytes minations, 4 ml of the supernatant fractions were 1131. It was, therefore, of interest to assess the extent quickly deproteinized with perchloric acid. The perto which this metabolic transition is reflected in the chloric acid extracts were neutralized with KOH-tritwo compartments. It was found that the steady-state ethanolamine to a pH of about 6. oxidation in the isocitrate dehydrogenase indicator The pellet fractions of samples 2 and 3 were desystems during ureogenesis from ammonia is largely signated mitochondrial, and intracellular, respectively, confined to the mitochondrial matrix space. Thus, whereas the corresponding supernatant fractions were the possibility exists that during enhanced rates of designated cytosolic, and extracellular, respectively. NADPH utilization the equilibration across the mitochondrial inner membrane may be limited, e.g. by the tricarboxylate carrier. Monitoring of the Dihydro Band of Nicotinamide Nucleot ides MATERIALS AND METHODS Hepatocyte Prepuration and Incubation Livers from male Wistar rats (1 50 - 200 g), fed on Altromin stock diet, were perfused and used for isolation of hepatocytes as described [14]. For incubation, hepatocytes were transferred into 25 ml conical flasks containing an incubation medium at 37 "C, equilibrated with a gas mixture of O,/CO, (95/5, v/v). Final concentrations were: D-glucose, 10 mM ; L-lactate, 2.1 mM ;pyruvate, 0.3 mM; ~,~-3-hydroxybutyrate, 0.6 m M ; acetoacetate, 0.3 m M ; L-ornithine, 2.4 mM ; and the salt mixture of the standard medium [14]: NaCl, 115 m M ; KCl, 5.9 m M ; MgCl,, 1.2 mM; NaH,PO,, 1.2 m M ; CaC12, 2.5 m M ; Na2S04, 1.2 mM ; and NaHCO,, 25 mM. Hepatocytes were added to a final concentration of 30-40 mg dry wt per ml.
Hepatocytes were incubated as above in l-cm glass cuvettes and stirred magnetically. The absorbance difference between 350 and 380nm was measured with a special photometer designed and constructed in the Electronics Department of Sonderforschungsbereich 51, as described [12,14]. M e tub0 lite Assuys Concentrations of citrate, isocitrate, 2-oxoglutarate, glutamate, NADP', ATP, ADP, AMP, urea and ammonia in the neutralized extracts were determined by enzymatic optical tests, either fluorometrically or spectrophotometrically, based on the procedures described in [16]. Measurements were carried out on Eppendorf fluorometers, on a Perkin-Elmer dual wavelength spectrophotometer (model 156), and on a Hitachi spectrophotometer (model 181).
H. Sies, T. P. M. Akerboom, and J . M. Tager
303
Chemicals The silicone oils, AR 200 and SF 96/100, were generously provided by Wacker Chemie (Burghausen) and General Electric (Bergen op Zoom), respectively. Chemicals and Biochemicals were obtained from Boehringer (Mannheim), Merck (Darmstadt), Serva (Heidelberg), and Sigma (Miinchen).
RESULTS A N D DISCUSSION Oxidation of NADPH in Isolated Hepatocytes Isolated hepatocytes rapidly respond to addition of ammonia (either added as an ammonium salt, or by generation with urea plus urease (Fig.1)) with a substantial oxidation of whole cell NADPH. This is indicated by the decrease of the dihydro band absorbance of the nicotinamide nucleotides (Fig. 1) and by recent enzymatic analyses of the reduced nicotinamide nucleotides, NADPH and NADH [17]. The further addition of tert-butyl hydroperoxide in the
Urea
Urease
tert-Butyl hydroperoxide
(10rnM) (14bg/rnl) (0.75rnM)
experiment indicates a further oxidation of NADPH, catalyzed by glutathione reductase reacting on GSSG formed by glutathione peroxidase. The latter transition involves both the mitochondria1 and cytosolic compartments. A substantial oxidation of NADPH also occurs upon addition of ammonia to isolated mitochondria [18-20). Since about z/3 of the total NADPH + NADP' in the liver is localized in the mitochondria and 1 / 3 in the cytosol [21], the question arises of whether the oxidation of NADPH observed in the intact hepatocyte on addition of ammonia is confined to the mitochondrial compartment. Digitonin Fractionation of the Hepatoc\~tcs In order to investigate this problem, the experiments described below, which were essentially similar to that of Fig. 1, were carried out. Hepatocytes were transferred to an incubation medium and further processed as described in the Materials and Methods section. A separation of the mitochondrial and cytosolic compartments was achieved, as indicated by the marker enzymes (Tablel) and by the ATP/ADP ratios (Table2). In the digitonin-treated cells, 83 '%, of lactate dehydrogenase and 14% of glutamate dehydrogenase activity were found in the supernatant (cytosolic) fraction (Table 1). The ATP/ADP ratios in the supernatant (cytosolic) and pellet (mitochondria]) fractions of 7.7 and 2.2, respectively, are
Table 2. A TPIADP ratios in t o t d extracts of isolated Iiepatocytes and in pellel i~nirochonclsiuf J ~iro'.c.~t~c'l.tlcir[~i7f ic:1-to.rolic. t /rwriom of digitonin-trerrted heputocytes in ubsence und presrnw ( 5 nijn J of ammoniu Data are given as means +S.E.M. (6 observations)
T
r -
~~
0
T
~~~
10
1 20
Fiaction
Addition
ATPIADP
Total
none 10 mM NH,CI
5.07 -I 0.14 4.06 k 0.21
Mitochondria1
none 10 mM NH,CI
2.29 2 0.33 1.49 k 0.20
Cytosolic
none 10 mM NH,CI
7.71 F 0.58 9.36 2 1.11
Incubation time (rnin)
Fig. 1. Oxidution of N A DPH in isoluted heputocytes upon uddition of unzmoniu (generated from ureu plus ureuse) and of tert-butyl hvdroperosidr. Dual-wavelength recording at 350 - 380 nm
Table 1. Murkrr enzymes in the supernatunts of digztonin-treated und untreated heputocytes Data are given as means t_S.E.M. (6 observations) Enzyme
Total activity in sample
Activity in supernatant of untreated cells (extracellular fraction)
Lactate dehydrogenase Glutamate dehydrogenase
U/g dry wt
'x total
1200 -I 165 855 62
4.0 i 1.5
8.3 k 1.4
digitonin-treated cells (cytosolic fraction)
83 & 16 14 2 3.6
Mitochondria1 and Cytosolic NADPH Systems
304
Table3. Citrate, isocitrate and 2-oxoglutarate in isolated hepatocytes in the absence of ammonia and 5 min afier the addition of 10 m M ammonium chloride The total, mitochondrial and cytosolic contents of metabolites were obtained as described in Materials and Methods, except for the isocitrate contents in the cytosolic fractions which were calculated by difference, between total and mitochondrial data. Calculated cytosolic values as obtained by difference between intracellular and mitochondrial values, are also given. Redox potentials were calculated with = -433 mV, assuming constant CO, concentration of 1.2 mM. Urea production rate was 9.6 pmol rnin-' (g dry wt)-'. Data are given as means fS.E.M. (4-8 observations) Fraction
Addition
Citrate
Isocitrate
2-0x0glutarate
Citrate/ isocitrate
Isocitratel 2-oxoglutarate
nrnol/g dry wt Total
none 10mM NH4CI
1780 i: 110 770 ? 70
Mitochondrial, from pellet
none 10mMNH4CI
870 & 110 380 2 70
Cytosolic, from supernatant
none 10 mM NH4C1
1010 f I0 670 +_ 130
Cytosolic, intracellular minus mitochondrial
none 10 mM NH,C1
47 5 210
Redox potential of free NADPH mV
95 f 8 38 f 6
1990 k 376 489 -i- 58
30 k 3
345 i 45 96 k 23
33.8 k 7.0 0.097 kO.015 > 148 c 0.042
-402 > -391
65 f 6 > 34
1470 k 179 289 i 34
15.8 f 3.4 0.047 f0.005 < 18.2 > 0.130
-392 < -406
28 k 4 > 13
497 I1 122 -t 22
0.062 ? 0.024 > 0.079
-396 < -399
i3
comparable to those obtained previously with cells from fed rats [22]. The marker enzyme activities in the supernatant fractions were similar in the presence of ammonia (not shown). The ATP/ADP ratio in the hepatocytes (total fraction) decreased slightly (Table 2). Interestingly, the addition of ammonia appears to be accompanied by a relatively pronounced decrease of the mitochondrial ATP/ADP ratio concomitant with a slight increase of the cytosolic ATP/ADP ratio. This phenomenon, together with additional information on the adenine nucleotide systems, will be reported in a separate paper (in preparation). Levels of Tricarboxylates and 2-Oxoglutavate in Steady State without Added Ammonia
Data on citrate, isocitrate and 2-oxoglutarate levels in the hepatocytes of fed rats are presented in Table3. The relative proportions of these metabolites in the total extracts are similar to those obtained from freeze-quenched liver in situ [4,6]. Regarding the mitochondrial values, it is seen that the aconitase mass action ratio is about 34 and the isocitrate/2-oxoglutarate ratio about 0.1, compared to values in the total extract of 20 and 0.06, respectively. From the isocitrate/2-oxoglutarate ratio, the redox potential of mitochondrial free NADPH is calculated to be -402mV, which is equivalent to a [NADPH]/ [NADP'] ratio of 83.5 at pH 7.0. The corresponding values for the cytosolic space are less readily obtained. The problem is approached in two ways, resulting in similar values. First, the values for cytosolic fraction in Table3 are directly used for calculation of the isocitrate/2-oxoglutarate ratio and the citrate/isocitrate ratio. The former is
19.7 f 1.3 22.9 k 2.7
17 < 16.2
0.060 f 0.012 0.085 f0.020
about 0.05, i.e. considerably lower than that obtained for the mitochondrial space, corresponding to a redox potential of free NADPH of -392 mV, or a [NADPH]/[NADP+] ratio of 39.5. This indicates that the extramitochondrial NADPH system is operating at a redox potential 10 mV more positive than inside the mitochondria, in agreement with the findings obtained with isolated mitochondria and external isocitrate dehydrogenase [5]. However, the direct cytosolic values in Table 3 represent not only the cytosolic metaboIite contents but also the extracellular contribution. This is eliminated in the second way of calculation as follows. The cytosolic values are calculated by the difference of the intracellular minus mitochondrial data, as shown in the lower part of Table 3. The isocitrate/2-oxoglutarate ratio obtained is 0.06, corresponding to a redox potential of free NADPH of - 396 mV, again more positive than that inside the mitochondria. It cannot be decided at present which method of calculation is a better reflection of the situation within the cell. While the uncorrected supernatant values would require the assumption of equilibration of isocitrate and 2-oxoglutarate across the plasma membrane, the values calculated from intracellular minus mitochondrial data depend on the difference between two separate samples. However, the results obtained with the two methods are in fairly close agreement, and indicate a more positive extramitochondrial redox potential of NADPH. Therefore, it may be reasonable to interpret the data as being in support of the operation of the isocitrate dehydrogenase reaction in equilibrating intramitochondrial and extramitochondrial NADPH/NA DP' . Total activity of the enzyme is 22.4 U/g liver wet wt 141, of which about 80% is localized extramitochondrially 111. The more
H. Sies, T. P. M. Akerboom, and J. M. Tager
305
Table4. Calculation of tricarboxylate and 2-oxoglutarate concentrations in mitochondrial and cytosolic water spaces Data from Table3. The water contents of 0.21 ml/g dry wt and 2.0 mljg dry wt for the two spaces, respectively, were used according to Williamson’s data [5]. In the lower section, the resulting mitochondriaI/cytosolic concentration gradients are given Fraction
Data obtained Ti-oin
Addition
Citrate
lsocitrate
2-Oxoglutarate
mM Mitochondrial
pellet
-
10 mM NH,CI Cytosolic
intracellular minus mitochondrial supernatant
_________ Mitochondi-ial/cytosolic concentralion gi-adieiit
-
10 mM NH,CI
0.142 < 0.014
1.64 0.46
0.24 0.11
0.014 > 0.0065
0.25 0.06
0.032 > 0.017
0.14 0.14
17.2 16.4
10.1 2.2
6.6 7.5
8.1 5.3
4.4 0.8
2.2 3.3
0.51 10 mM NH,Cl 0.34 ________
~~lilochondrial/intracellular
-
minus mitochondria1
10 mM NH,CI
pellet/supernatant
4.14 1.81
-
10 mM NH,CI
negative redox potential within the mitochondria presumably results from the pH difference (more alkaline inside), favoring a relative accumulation of the tricarboxylate with respect to the dicarboxylate C23L Levels of Tricarboxylates and 2-Oxoglutarate in Steady State during Ureogenesisfrom Ammonia
The metabolite pattern observed in the presence of added ammonia exhibits some striking differences to the situation without added ammonia (Table 3). First, it is seen that all three metabolites are present in lower amounts, while the content of glutamate is increased (not shown). The total levels of citrate and isocitrate decrease by about 60% and that of 2-0x0glutarate by about 80 %, in agreement with the results obtained with freeze-quenched liver [24,25]. The major point of interest is the following: In the mitochondrial fraction, the isocitrate level decreases from 30 nmol/g dry wt to values smaller than are detectable with the employed analytical methods. This causes the mitochondrial citrate/isocitrate ratio to rise from 34 to > 148, and the isocitrate/2-oxoglutarate ratio to decrease from about 0.1 to < 0.042. Thus, the redox potential of mitochondrial NADPH increases from -402 mV to less than -391 mV. In contrast, the isocitrate/2-oxoglutarate ratio in the extramitochondrial space (either from the supernatant or from intracellular minus mitochondrial data) is increased concomitantly. This is taken to indicate that during the enhanced rate of urea production and the associated flux through the glutamate dehydrogenase reaction the difference of redox potentials of NADPH in mitochondrial and cytosol not only is diminished but actually is reversed in sign. N A D P H in the mitochondrial space becomes more oxidized than that in the cytosol.
Calculation of Metabolite Concentrations in Mitochondria1 and Cytosolic Spaces
For a further evaluation of the cell-physiological aspects, it is necessary to have information on the free cellular concentrations of the metabolites in the respective spaces. In a first approximation, this is obtained by referring the measured metabolite amounts to the respective water contents (Table4), although clearly a further distinction must be made between free and bound metabolites. Nevertheless, the figures in Table4 provide useful information. The concentration of isocitrate in the mitochondrial space, calculated to be 0.142 mM without added ammonium chloride, decreases to less than 0.014 mM in its presence. Thus, the isocitrate concentration decreases to values much lower than the K, (isocitrate) of the NAD-dependent dehydrogenase, which is of the order of 0.1 mM [26]. On the other hand, the K , (isocitrate) of the NADP-dependent dehydrogenase is of the order of 0.01 mM or lower [26]. Thus, the NADP-linked mitochondrial enzyme is not only favored because of its higher activity compared to the NAD-linked enzyme, but also for kinetic reasons. A second piece of information resulting from the calculation of concentrations is obtained from the mitochondrial/cytosolic gradients. As expected from the observations on isolated mitochondria [23], the gradient is higher for the tricarboxylates than for the dicarboxylate. The results for citrate are of the same order of magnitude as those calculated by Williamson [5] for perfused liver in presence of oleate and by Greenbaum et al. [6] for freeze-quenched liver from fed rats. The distribution of 2-oxoglutarate is also in favor of the mitochondrial space, in marked contrast to the value of less than 0.1 for the mitochondrial/ cytosolic gradients calculated for fed rat liver by the authors of [6],but roughly in agreement with the recent
306
data for fed rats obtained by Soboll et al. [27] with the non-aqueous fractionation procedure [28].
Concluding Remarks First, it should be mentioned that in the present experiments the particular emphasis was on determining the low levels of isocitrate, thus requiring a relatively high cell concentration in the incubation ; the latter obviously is unfavorable to a perfect separation by the digitonin procedure, and, therefore, a compromise was sought between cell concentration on the one hand and the digitonin concentration (plus reaction time) on the other. As indicated by the ATP/ADP ratios, the separation of the particulate fraction from the soluble fraction was comparable to results in the literature and was considered acceptable. The somewhat larger proportion of glutamate dehydrogenase in the supernatant fraction indicates that some back-mixing of the two fractions is not excluded. However, the extent of such back-mixing must be rather limited, since it allowed the isocitrate levels in the mitochondrial fraction to decrease to < 3 nmol/g dry wt (Table 3), for example, when ammonia was present . The following conclusions can clearly be drawn from the present experiments. Firstly, the isocitrate dehydrogenase system can be used as an indicator metabolite system for NADPH in the mitochondrial and cytosolic spaces. Secondly, the isocitrate concentration in the mitochondrial matrix space decreases during ureogenesis from ammonia by at least one order of magnitude, to less than approx. 10 pM. Isocitrate removal occurs at a high rate causing a deviation of the aconitase reaction from near-equilibrium, and also exceeding isocitrate import rates on the tricarboxylate carrier. Thirdly, although the mitochondrial/cytosolic concentration gradients can only be calculated with respect to the order of magnitude, due to possible pitfalls mentioned above, it is clear that the isocitrate gradient is substantially decreased in presence of ammonia. On the other hand, there is relatively little effect on +! ie citrate and 2-oxoglutarate concentration gradients. The first conclusion is in agreement with the conclusions obtained by Krebs and his colleagues regarding the cytosolic space [4] but not the mitochondrial space [29], and the following further comments are given for clarification that mitochondrial NADPH-dependent isocitrate dehydrogenase may be suitable as a redox indicator system (see also the recent discussion in [9,30]). In the absence of added ammonia, the activity of the enzyme is sufficiently high, about 5 U/g liver wet wt, to maintain near-equilibrium, in view of the approximate rate of metabolite flow of 0.8 pmol/min per g liver wet wt
Mitochondria1 and Cytosolic NADPH Systems
through this step 191 and the high concentration of isocitrate compared to the K, (Table4). During the increased rate of flow in the presence of ammonia, activity of the enzyme is high enough to lead to a substantial decrease in the mitochondrial isocitrate/2oxoglutarate ratio in spite of the rapid utilization of 2-oxoglutarate by glutamate dehydrogenase, occurring at a rate of about 2 pmol/min per g liver wet wt in the steady state. The latter can be estimated because under such condition the rate of urea formation from ammonia is equal to the rate of amination of 2-0x0glutarate (cf: Scheme 1 in [17]). It is possible that isocitrate dehydrogenase activity may no longer be in excess under such extreme condition with isocitrate concentration in the region of the K , or lower, leading to an indicator error. At present, the methodological problem of distinguishing isocitrate levels lower than 3 nmol/g liver dry wt from zero precludes an evaluation of the possible indicator error; in other words, it is difficult to estimate whether the redox potential of mitochondrial free NADPH is significantly more positive than -391 mV (Table3) in the presence of ammonia. The present experiments provide further insight into the compartnientation of NADPH redox potentials and the organisation of NAD-dependent and NADP-dependent isocitrate dehydrogenases and energy-linked transhydrogenase, an area of research that lo date has not been completely elucidated [2.9,3 I ] . It is evident that the increased NADPH demand during ureogenesis is accompanied by a corresponding shift in the isocitrate dehydrogenase system : whereas without added ammonia part of isocitrate oxidation proceeds via the NAD-dependent, non-equilibrium enzyme [3], and another part through the NADPdependent enzyme coupled to the energy-linked transhydrogenase, the increase in the NADPH redox potential (as initiated here by the addition of ammonia) causes a shift in favor of the NADP-dependent enzyme, in agreement with other lines of evidence obtained by previous investigators [3,20,32,33]. A new question arising from the present study is whether under the condition of enhanced mitochondrial NADPH utilization there is a transfer of NADPH equivalents from the cytosol into the mitochondria, i. e. in the opposite direction to that usually envisaged, for example in lipogenesis; and whether the mitochondria] oxidation of NADPH leads to an increased formation of NADPH by the pentose phosphate pathway. However, the tricarboxylate carrier may not be capable of equilibrating, since the citrate gradient between mitochondria and cytosol is still considerable in the presence of ammonia, and the cytosolic citrate concentration is decreased toward the K , value of 0.12 mM for the carrier [34]. This limitation may explain the higher degree of reduction of cytosolic NADPH compared to mitochondria1 NADPH as
H. Sies, T. P. M. Akerboom, and J. M. Tager
307
indicated by the isocitrate dehydrogenase system, in the presence of ammonia.
15. Zuurendonk, P. F.,Akerboom, T. P. M. & Tager, J. M . (1976) in Use of' Isolated Liver Ce1l.s and K i d n q Tubules it7 Meruholic Studies (J. M. Tager, H . D. Soling & J. R. Williamson, Expert technical assistance was provided by Ingrid Linke and eds) pp. 17-27, North Holland, Amsterdam. Annegret Marklstorfer. Thanks are due to Prof. Theodor Bucher 16. Bergmeyer, H. U., ed. (1 974) Merhoden der En:yyrnuiischen for his helpful comments and suggestions and to Dr Karl-Hein7 Anul,yse, 3rd edn, Verlag Chemie, Weinheim. Summer for his help in part of the assays. This investigation wii\ 17. Sies. H.. Summer, K . H . , Haussingel-. D & Uiichel-. T l i . (1076) supported by Deursche Forschungs~emeinschufi,SonderJOr.rchun~~.\.in U,se of lsokzted Liver Cells and K i r h q Tiihuli~.~ in Mrlubereich 51, Medizinisrhe Molekularbiologie und Biochemie (Grant holic Studies (J. M. Tager, H. D. Soling & J. R. Willlainson. D/8) and by a grant from the Netherlands Foundation for Pure eds) pp. 311 -316, North Holland, Amsterdam. Scientific Research (ZWO) under auspices of the Netherlands 18. Klingenberg, M. & Slenczka, W. (1959) Bioclim?. Z. 331,486Foundation for Chemical Research (SON). 517. 19. Tager, J . M. (1 966) in Regulution of' Metabolic Procrssra in Mitochondriu (Tager, J. M., Papa, S.. Quagliariello, E. & Slater, E. C., eds) pp. 202-216, Elsevier, Amsterdam. 20. Nicholls, D. G. &Garland, P. B. (1969) Biocheni. J . 114,215REFERENCES 225. 1. Hogebooni, G . H. & Schneider, W. C. (1950) J . Biol. Chem. Chrm. 11, pp. 21. Klingenberg, M . (1961) Colloy. Grs. Ph~~siol. 186,417-427. 82-114. 2. Plaut, G. W. E. (1970) Curr. Top. Cell Regul. 2, 1 -27. 22. Siess, E. & Wieland, 0. (1 975) FEBS Lett. 52, 226 - 230. 3. Goebell, H. & Kiingenberg, M. (1964) Biochem. 2.340, 441 23. Palmieri, F., Quagliariello, E. & Klingenberg, M. (1970) Eur. 464. J . Bioc,hern. 17, 230-238. 4. Veech. R. L., Eggleston, L. V. & Krebs, H. A. (1969) Biochem. 24. Nordmann. R., Petit, M . A . & Nordmann, J. (1972) Biochin7ir. 54, 1473- 1478. J . 115, 609-619. 5. Williamson, J . R. (1969) in The Energy Level and Metabolic 25. Brosnan, J. T . & Williamson, D. H. (1974) Biochem. J . 138, Control in Mitochondria (S. Papa, J. M. Tager, E. Quaglia453 - 462. riello & E. C. Slater, eds) pp. 385-400, Adriatica Editrice, 26. Colman, R. F. (1975) Adv. Enzymr Rcgul. 13, 413-433. Bari. 27. Soboll, S., Scholz, R., Freisl, M.. Elbers, K. & Heldt, H. W. 6. Grecnbaum, A . L., Gumaa, K. A. & McLean, P. (1971) Arch. (1 976) in Use of Isolrrted Liver Cells and Kidney Tubules it7 Bio~he177.B / O , O ~ I(43, ~ , S .617 - 663. Metub~licStudies (J. M. Tager, H.D. Soling & J. R. Williamson, eds) pp. 29 -40, North Holland, Amsterdam. 7. Hoek, J. B. & Ernster, L. (1974) in Alcohol und Aldehyde Mctubolizink. S.vsiems (R. G. Thul-man, T. Yonetani, J . R . Wil28. Elbers, R., Heldt, H. W., Schmucker, R., Soboll, S. & Wiese, H. (1 974) Hoppe-Sc,.vler'.s Z.Plf.vsiol. Chem. 355. 378 - 393. liamson & B. Chance, eds) pp. 351 -364, Academic Press. 29. Krebs, H. A. & Veech, R. L. (1969) in Thr Energ! Lruel ( r i d New Yoi-k. i ~ cuitl hli~rtrholic~ Coiitrol iu Metubolic Control in Mitochondria ( S . Papa, J. M. Tager. 8. Papa. S (1969) in The I k y ~ Loivl Mitochondrirr ( S . Papa, J. M. Tager, E. Quagliariello BL E. C. E. Quagliariello & E. C. Slater, edsj pp. 329-382, Adriatica Slate)', eds) pp. 401 -409, Adriatica Editrice, Bari. Editrice, Bari. Its Regulution in 9. Biicher. Th. & Sies, H. (1976) in Use of Isolated Liver Cells 30. Williamson. J. R . (1 976) in GI~lc~oneogene.sis. und Kidntj, Tubules in Metabolic Studies (J. M. Tager, Mammcrliun Species (R. W. Hanson & M. A. Mehlman, H. D. Soling & J. R. Williamson, eds) pp. 41 -64, North eds) pp. 165-220, J. Wiley & Sons, New York, London, Sydney, Toronto. Holland, Amsterdam. 10. Zuurendonk, P. F. & Tager, J. M. (1974) Biochrm. 5iophy.s. 31. Rydstrom, J., Hoek, J. B. & Ernster, L. (1976) in The EllActu, 333, 393 - 399. rymes (P. D. Boyer, ed.) 3rd edn. vol. 13, pp. 51 -88, Aca11. Chamalaun, R. A. F. M . & Tager, J. M. (1970) Biochini. Biodemic Press, New York. pllys. act^, 222. 1 19 - 134. 32. Stein, A. M., Stein, J . H. & Kirkman, S. K . (1967) 5iochnni.stry.6, 1370-1379. 12. Sies, H., Hiiussingcr, D. & Grosskopf, M . (1974) Hoppc,-S'.v33. Smith, C. M., Beach, R . L. & Plaut. G . W. E. (1975) Ahstr. Ier'x Z. Phy.siol. Cheru. 355, 305 -318. Corninun. 10th M w r . Fed. Eur. Biochm. Sot.., 717. 13. Sies, H., Summer, K. H. & Bucher, Th. (1975) FEBS Lett. 34 Palmieri, F., Stipani, I., Quagliariello, E. & Klingenberg. M. 54,274-278. (1972) Eur. J . Biochem. 26, 587 - 594. 14. Sies, H. & Grosskopf, M. (1975) ELIT.J . Bioc,hrrn.57, 513 -520.
H. Sies, Institut fur Physiologische Chemie, Physikalische Biochemie und Zellbiologie dei- Ludwig-Maximilians-Universitiit Munchen, GoethestraBe 33, D-8000 Munchen 2, Federal Republic of Germany T. P. M. Akerboom and J. M. Tager. Laboratorium voor Biochemie. Universiteit van Amsterdam, B.C.P. Jansen Instituut, Plantage Muidergracht 12, Amsterdam C, The Netherlands
Note Addrdin Pioof(December 28, 1976). The theoretical redox potential difference of the NADP-linked isocitrate dehydrogenase system between mitochondria and cytosol can be calculated from the tricarboxylate and dicarboxylate gradients as was indicated above. Assuming equilibration across the mitochondria] inner membrane according to the Donnan relationship, and with equal COz concentration on either side of the membrane, a pH difference
(more alkaline inside) of 0.41 results from the data in Table 4 for citrate3- and 2-oxoglutarate2-, and 0.33 for isocitrate3-. This corresponds t o a calculated redox potential difference of 0.41 x 30.7 = 12.6 mV (or 10.1 mV, respectively) more negative inside the mitochondria than outside. The analytical data obtained with isolated mitochondria [7] and with the hepatocytes (Table3), thereCore, arc within the range of theoretically expected values.
Mitochondria1 and Cytosolic NADPH Systems and Isocitrate Dehydrogenase Indicator Metabolites during Ureogenesis from Ammonia in Isolated Rat Hepatocytes Helmut SIES, Theodorus P. M. AKERBOOM, and Joseph M. TAGER
Instirut fur Physiologische Cliemie. Pli) sikalische Biochemie und Zellbiologie der Universitat Munchen, and Laboratory of Biochemistry, B.C.P. Jansen Institute, University of Amsterdam (Received August 4 1 September 28, 1976)
1. Citrate, isocitrate and 2-oxoglutarate levels were determined in isolated rat hepatocytes and in particulate and soluble fractions thereof, obtained by the digitonin and silicone oil fractionation technique. 2. Calculated from isocitrate/2-oxoglutarate ratios (“indicator metabolite method”), the redox potential of mitochondrial free NADPH is -402 mV, whereas that of the extramitochondrial (cytosolic) space is about 10 mV more positive, -392 mV. 3. Addition of ammonia (either as ammonium chloride or from urea plus urease) to isolated hepatocytes causes preferential oxidation of mitochondrial NADPH, as demonstrated by spectrophotometry of the dihydro band and by the changes in the isocitrate/2-oxoglutarate ratios. The redox potential difference of free NADPH between mitochondria and cytosol is abolished or even reversed. 4. It is concluded that during ureogenesis from ammonia mitochondrial isocitrate oxidation is shifted largely in favor of the NADP-linked as opposed to the NAD-linked enzyme; isocitrate concentration under these conditions is less than 10 pM, below the K, (isocitrate) of the NADlinked enzyme but in the range of that for the NADP-linked enzyme. 5. Both in the absence and in the presence of ammonia there is a concentration gradient across the mitochondria1 inner membrane (from mitochondria to cytosol) for citrate, isocitrate, and also, to a smaller extent, for 2-oxoglutarate. 6. These results and data in the literature on enzyme activity are in agreement with the assumption of near-equilibrium of NADP-dependent isocitrate dehydrogenases in the mitochondrial matrix and cytosolic spaces in the absence of ammonia; accordingly, during urea formation from added ammonia the redox potential of mitochondrial free NADPH is increased to -391 mV or possibly even higher if there exists an indicator error under this condition. Isocitrate dehydrogenases occur both intramitochondrially and extramitochondrially in rat liver [l]; NADP-dependent enzymes are present in both spaces, whereas an NAD-dependent enzyme is present only intramitochondrially (see [2] for a review). The clear concept, put forward by Goebell and Klingenberg [3], of the dichotomy of mitochondrial isocitrate oxidation into a non-equilibrium, regulated (or regulatory) pathway catalyzed by the NAD-dependent enzyme and an equilibrium pathway catalyzed by the NADP-dependent enzyme implies that the compoEnzymes. Isocitrate dehydrogenase (NAD+) (EC 1.1.1.41); isocitrate dehydrogenase (NADP+) (EC 1.1.1.42).
nents of the mitochondrial isocitrate dehydrogenase system may be utilized as ‘indicator metabolites’ for the mitochondrial NADP system, provided that the activity is high enough (see below). Regardins the extramitochondrial (cytosolic) space, there are no major complications to utilize the isocitrate dehydrogenase system as an indicator for the redox potential of cytosolic free NADPH. Veech etal. [4] have, in fact, utilized the isocitrate dehydrogenase system for this purpose ; however, their analytical data were the total tissue levels rather than the cytosolic ones. Furthermore, attempts have been made by Williamson [5] and Greenbaum et al. [6] to calculate mitochondrial and cytosolicconcentrations from the total tissue levels
302
Mitochondria1 and Cytosolic NADPH Systems
Shaking frequency was 120 min-'. Samples were of metabolites, based on certain assumptions on the taken after 20 or 25min. In some incubations, subcellular distribution. NH4Cl (10 mM) was added after 20 min. Hoek and Ernster [7] have shown that, under energized conditions, isolated rat liver mitochondria maintain a ratio of ([NADPH]/[NADP'])i,/([NADPH]/ Hepu t ocyte Sump ling and Frac t ionat ion [NADP+]),,, = 2 in the presence of the components The digitonin fractionation procedure was essenof the isocitrate dehydrogenase reaction in the suspentially as described [l5]. At one time point, three sion medium. In a first approximation, this would samples of 1 ml each were taken from the incubation : correspond to a redox potential difference for free sample 1, direct deproteinization with perchloric NADPH of 10mV more negative inside the mitoacid, 0.94 M final concentration, to give the total chondrial matrix space than outside, and it implies contents of the metabolites; sample 2, mixing with that the isocitrate and 2-oxoglutarate carrier systems 5 ml of a separation medium (0.25 M sucrose, 20 mM are capable of bringing about equilibration of these 3-(N-morpholino)propane sulfonate, pH 7.0, 3 mM substrates between the two spaces [7,8]. The present EDTA) at -2 "C containing digitonin ( 2 mM, final investigation was carried out to determine, with the concentration) and, after 15 s, centrifugation through aid of the indicator metabolite system of NADP2 ml of a silicone oil mixture (Wacker Chemie AR 200/ linked isocitrate dehydrogenase (for a recent discusGeneral Electric SF 96/100, S / l , v/v, and in some exsion, see [9]), whether a similar redox potential difperiments, 6/1, v/v) into 2 ml of perchloric acid, ference exists within the intact hepatocyte. For this 0.94 M ; sample 3, same as sample 2, but without purpose, isolated rat hepatocytes were subfractionated digitonin. with the digitonin and silicone oil fractionation techPart of the supernatant fractions of samples 2 and 3 nique of Zuurendonk and Tager [lo] and metabolite were taken for assay of lactate and glutamate dehydropatterns determined. genases as marker enzymes of the cytosolic and miIn perfused liver, urea formation from ammonia tochondrial matrix spaces, respectively. Total enzyme is accompanied by a substantial oxidation of NADPH activities in the incubation were measured after treat[I 1,121 and, indeed, ureogenesis from ammonia has ment of a sample with 0.5% lubrol WX, and were been identified as a major NADPH-requiring process taken as reference ( = 100%). For metabolite deterof the mitochondrial space in isolated hepatocytes minations, 4 ml of the supernatant fractions were 1131. It was, therefore, of interest to assess the extent quickly deproteinized with perchloric acid. The perto which this metabolic transition is reflected in the chloric acid extracts were neutralized with KOH-tritwo compartments. It was found that the steady-state ethanolamine to a pH of about 6. oxidation in the isocitrate dehydrogenase indicator The pellet fractions of samples 2 and 3 were desystems during ureogenesis from ammonia is largely signated mitochondrial, and intracellular, respectively, confined to the mitochondrial matrix space. Thus, whereas the corresponding supernatant fractions were the possibility exists that during enhanced rates of designated cytosolic, and extracellular, respectively. NADPH utilization the equilibration across the mitochondrial inner membrane may be limited, e.g. by the tricarboxylate carrier. Monitoring of the Dihydro Band of Nicotinamide Nucleot ides MATERIALS AND METHODS Hepatocyte Prepuration and Incubation Livers from male Wistar rats (1 50 - 200 g), fed on Altromin stock diet, were perfused and used for isolation of hepatocytes as described [14]. For incubation, hepatocytes were transferred into 25 ml conical flasks containing an incubation medium at 37 "C, equilibrated with a gas mixture of O,/CO, (95/5, v/v). Final concentrations were: D-glucose, 10 mM ; L-lactate, 2.1 mM ;pyruvate, 0.3 mM; ~,~-3-hydroxybutyrate, 0.6 m M ; acetoacetate, 0.3 m M ; L-ornithine, 2.4 mM ; and the salt mixture of the standard medium [14]: NaCl, 115 m M ; KCl, 5.9 m M ; MgCl,, 1.2 mM; NaH,PO,, 1.2 m M ; CaC12, 2.5 m M ; Na2S04, 1.2 mM ; and NaHCO,, 25 mM. Hepatocytes were added to a final concentration of 30-40 mg dry wt per ml.
Hepatocytes were incubated as above in l-cm glass cuvettes and stirred magnetically. The absorbance difference between 350 and 380nm was measured with a special photometer designed and constructed in the Electronics Department of Sonderforschungsbereich 51, as described [12,14]. M e tub0 lite Assuys Concentrations of citrate, isocitrate, 2-oxoglutarate, glutamate, NADP', ATP, ADP, AMP, urea and ammonia in the neutralized extracts were determined by enzymatic optical tests, either fluorometrically or spectrophotometrically, based on the procedures described in [16]. Measurements were carried out on Eppendorf fluorometers, on a Perkin-Elmer dual wavelength spectrophotometer (model 156), and on a Hitachi spectrophotometer (model 181).
H. Sies, T. P. M. Akerboom, and J . M. Tager
303
Chemicals The silicone oils, AR 200 and SF 96/100, were generously provided by Wacker Chemie (Burghausen) and General Electric (Bergen op Zoom), respectively. Chemicals and Biochemicals were obtained from Boehringer (Mannheim), Merck (Darmstadt), Serva (Heidelberg), and Sigma (Miinchen).
RESULTS A N D DISCUSSION Oxidation of NADPH in Isolated Hepatocytes Isolated hepatocytes rapidly respond to addition of ammonia (either added as an ammonium salt, or by generation with urea plus urease (Fig.1)) with a substantial oxidation of whole cell NADPH. This is indicated by the decrease of the dihydro band absorbance of the nicotinamide nucleotides (Fig. 1) and by recent enzymatic analyses of the reduced nicotinamide nucleotides, NADPH and NADH [17]. The further addition of tert-butyl hydroperoxide in the
Urea
Urease
tert-Butyl hydroperoxide
(10rnM) (14bg/rnl) (0.75rnM)
experiment indicates a further oxidation of NADPH, catalyzed by glutathione reductase reacting on GSSG formed by glutathione peroxidase. The latter transition involves both the mitochondria1 and cytosolic compartments. A substantial oxidation of NADPH also occurs upon addition of ammonia to isolated mitochondria [18-20). Since about z/3 of the total NADPH + NADP' in the liver is localized in the mitochondria and 1 / 3 in the cytosol [21], the question arises of whether the oxidation of NADPH observed in the intact hepatocyte on addition of ammonia is confined to the mitochondrial compartment. Digitonin Fractionation of the Hepatoc\~tcs In order to investigate this problem, the experiments described below, which were essentially similar to that of Fig. 1, were carried out. Hepatocytes were transferred to an incubation medium and further processed as described in the Materials and Methods section. A separation of the mitochondrial and cytosolic compartments was achieved, as indicated by the marker enzymes (Tablel) and by the ATP/ADP ratios (Table2). In the digitonin-treated cells, 83 '%, of lactate dehydrogenase and 14% of glutamate dehydrogenase activity were found in the supernatant (cytosolic) fraction (Table 1). The ATP/ADP ratios in the supernatant (cytosolic) and pellet (mitochondria]) fractions of 7.7 and 2.2, respectively, are
Table 2. A TPIADP ratios in t o t d extracts of isolated Iiepatocytes and in pellel i~nirochonclsiuf J ~iro'.c.~t~c'l.tlcir[~i7f ic:1-to.rolic. t /rwriom of digitonin-trerrted heputocytes in ubsence und presrnw ( 5 nijn J of ammoniu Data are given as means +S.E.M. (6 observations)
T
r -
~~
0
T
~~~
10
1 20
Fiaction
Addition
ATPIADP
Total
none 10 mM NH,CI
5.07 -I 0.14 4.06 k 0.21
Mitochondria1
none 10 mM NH,CI
2.29 2 0.33 1.49 k 0.20
Cytosolic
none 10 mM NH,CI
7.71 F 0.58 9.36 2 1.11
Incubation time (rnin)
Fig. 1. Oxidution of N A DPH in isoluted heputocytes upon uddition of unzmoniu (generated from ureu plus ureuse) and of tert-butyl hvdroperosidr. Dual-wavelength recording at 350 - 380 nm
Table 1. Murkrr enzymes in the supernatunts of digztonin-treated und untreated heputocytes Data are given as means t_S.E.M. (6 observations) Enzyme
Total activity in sample
Activity in supernatant of untreated cells (extracellular fraction)
Lactate dehydrogenase Glutamate dehydrogenase
U/g dry wt
'x total
1200 -I 165 855 62
4.0 i 1.5
8.3 k 1.4
digitonin-treated cells (cytosolic fraction)
83 & 16 14 2 3.6
Mitochondria1 and Cytosolic NADPH Systems
304
Table3. Citrate, isocitrate and 2-oxoglutarate in isolated hepatocytes in the absence of ammonia and 5 min afier the addition of 10 m M ammonium chloride The total, mitochondrial and cytosolic contents of metabolites were obtained as described in Materials and Methods, except for the isocitrate contents in the cytosolic fractions which were calculated by difference, between total and mitochondrial data. Calculated cytosolic values as obtained by difference between intracellular and mitochondrial values, are also given. Redox potentials were calculated with = -433 mV, assuming constant CO, concentration of 1.2 mM. Urea production rate was 9.6 pmol rnin-' (g dry wt)-'. Data are given as means fS.E.M. (4-8 observations) Fraction
Addition
Citrate
Isocitrate
2-0x0glutarate
Citrate/ isocitrate
Isocitratel 2-oxoglutarate
nrnol/g dry wt Total
none 10mM NH4CI
1780 i: 110 770 ? 70
Mitochondrial, from pellet
none 10mMNH4CI
870 & 110 380 2 70
Cytosolic, from supernatant
none 10 mM NH4C1
1010 f I0 670 +_ 130
Cytosolic, intracellular minus mitochondrial
none 10 mM NH,C1
47 5 210
Redox potential of free NADPH mV
95 f 8 38 f 6
1990 k 376 489 -i- 58
30 k 3
345 i 45 96 k 23
33.8 k 7.0 0.097 kO.015 > 148 c 0.042
-402 > -391
65 f 6 > 34
1470 k 179 289 i 34
15.8 f 3.4 0.047 f0.005 < 18.2 > 0.130
-392 < -406
28 k 4 > 13
497 I1 122 -t 22
0.062 ? 0.024 > 0.079
-396 < -399
i3
comparable to those obtained previously with cells from fed rats [22]. The marker enzyme activities in the supernatant fractions were similar in the presence of ammonia (not shown). The ATP/ADP ratio in the hepatocytes (total fraction) decreased slightly (Table 2). Interestingly, the addition of ammonia appears to be accompanied by a relatively pronounced decrease of the mitochondrial ATP/ADP ratio concomitant with a slight increase of the cytosolic ATP/ADP ratio. This phenomenon, together with additional information on the adenine nucleotide systems, will be reported in a separate paper (in preparation). Levels of Tricarboxylates and 2-Oxoglutavate in Steady State without Added Ammonia
Data on citrate, isocitrate and 2-oxoglutarate levels in the hepatocytes of fed rats are presented in Table3. The relative proportions of these metabolites in the total extracts are similar to those obtained from freeze-quenched liver in situ [4,6]. Regarding the mitochondrial values, it is seen that the aconitase mass action ratio is about 34 and the isocitrate/2-oxoglutarate ratio about 0.1, compared to values in the total extract of 20 and 0.06, respectively. From the isocitrate/2-oxoglutarate ratio, the redox potential of mitochondrial free NADPH is calculated to be -402mV, which is equivalent to a [NADPH]/ [NADP'] ratio of 83.5 at pH 7.0. The corresponding values for the cytosolic space are less readily obtained. The problem is approached in two ways, resulting in similar values. First, the values for cytosolic fraction in Table3 are directly used for calculation of the isocitrate/2-oxoglutarate ratio and the citrate/isocitrate ratio. The former is
19.7 f 1.3 22.9 k 2.7
17 < 16.2
0.060 f 0.012 0.085 f0.020
about 0.05, i.e. considerably lower than that obtained for the mitochondrial space, corresponding to a redox potential of free NADPH of -392 mV, or a [NADPH]/[NADP+] ratio of 39.5. This indicates that the extramitochondrial NADPH system is operating at a redox potential 10 mV more positive than inside the mitochondria, in agreement with the findings obtained with isolated mitochondria and external isocitrate dehydrogenase [5]. However, the direct cytosolic values in Table 3 represent not only the cytosolic metaboIite contents but also the extracellular contribution. This is eliminated in the second way of calculation as follows. The cytosolic values are calculated by the difference of the intracellular minus mitochondrial data, as shown in the lower part of Table 3. The isocitrate/2-oxoglutarate ratio obtained is 0.06, corresponding to a redox potential of free NADPH of - 396 mV, again more positive than that inside the mitochondria. It cannot be decided at present which method of calculation is a better reflection of the situation within the cell. While the uncorrected supernatant values would require the assumption of equilibration of isocitrate and 2-oxoglutarate across the plasma membrane, the values calculated from intracellular minus mitochondrial data depend on the difference between two separate samples. However, the results obtained with the two methods are in fairly close agreement, and indicate a more positive extramitochondrial redox potential of NADPH. Therefore, it may be reasonable to interpret the data as being in support of the operation of the isocitrate dehydrogenase reaction in equilibrating intramitochondrial and extramitochondrial NADPH/NA DP' . Total activity of the enzyme is 22.4 U/g liver wet wt 141, of which about 80% is localized extramitochondrially 111. The more
H. Sies, T. P. M. Akerboom, and J. M. Tager
305
Table4. Calculation of tricarboxylate and 2-oxoglutarate concentrations in mitochondrial and cytosolic water spaces Data from Table3. The water contents of 0.21 ml/g dry wt and 2.0 mljg dry wt for the two spaces, respectively, were used according to Williamson’s data [5]. In the lower section, the resulting mitochondriaI/cytosolic concentration gradients are given Fraction
Data obtained Ti-oin
Addition
Citrate
lsocitrate
2-Oxoglutarate
mM Mitochondrial
pellet
-
10 mM NH,CI Cytosolic
intracellular minus mitochondrial supernatant
_________ Mitochondi-ial/cytosolic concentralion gi-adieiit
-
10 mM NH,CI
0.142 < 0.014
1.64 0.46
0.24 0.11
0.014 > 0.0065
0.25 0.06
0.032 > 0.017
0.14 0.14
17.2 16.4
10.1 2.2
6.6 7.5
8.1 5.3
4.4 0.8
2.2 3.3
0.51 10 mM NH,Cl 0.34 ________
~~lilochondrial/intracellular
-
minus mitochondria1
10 mM NH,CI
pellet/supernatant
4.14 1.81
-
10 mM NH,CI
negative redox potential within the mitochondria presumably results from the pH difference (more alkaline inside), favoring a relative accumulation of the tricarboxylate with respect to the dicarboxylate C23L Levels of Tricarboxylates and 2-Oxoglutarate in Steady State during Ureogenesisfrom Ammonia
The metabolite pattern observed in the presence of added ammonia exhibits some striking differences to the situation without added ammonia (Table 3). First, it is seen that all three metabolites are present in lower amounts, while the content of glutamate is increased (not shown). The total levels of citrate and isocitrate decrease by about 60% and that of 2-0x0glutarate by about 80 %, in agreement with the results obtained with freeze-quenched liver [24,25]. The major point of interest is the following: In the mitochondrial fraction, the isocitrate level decreases from 30 nmol/g dry wt to values smaller than are detectable with the employed analytical methods. This causes the mitochondrial citrate/isocitrate ratio to rise from 34 to > 148, and the isocitrate/2-oxoglutarate ratio to decrease from about 0.1 to < 0.042. Thus, the redox potential of mitochondrial NADPH increases from -402 mV to less than -391 mV. In contrast, the isocitrate/2-oxoglutarate ratio in the extramitochondrial space (either from the supernatant or from intracellular minus mitochondrial data) is increased concomitantly. This is taken to indicate that during the enhanced rate of urea production and the associated flux through the glutamate dehydrogenase reaction the difference of redox potentials of NADPH in mitochondrial and cytosol not only is diminished but actually is reversed in sign. N A D P H in the mitochondrial space becomes more oxidized than that in the cytosol.
Calculation of Metabolite Concentrations in Mitochondria1 and Cytosolic Spaces
For a further evaluation of the cell-physiological aspects, it is necessary to have information on the free cellular concentrations of the metabolites in the respective spaces. In a first approximation, this is obtained by referring the measured metabolite amounts to the respective water contents (Table4), although clearly a further distinction must be made between free and bound metabolites. Nevertheless, the figures in Table4 provide useful information. The concentration of isocitrate in the mitochondrial space, calculated to be 0.142 mM without added ammonium chloride, decreases to less than 0.014 mM in its presence. Thus, the isocitrate concentration decreases to values much lower than the K, (isocitrate) of the NAD-dependent dehydrogenase, which is of the order of 0.1 mM [26]. On the other hand, the K , (isocitrate) of the NADP-dependent dehydrogenase is of the order of 0.01 mM or lower [26]. Thus, the NADP-linked mitochondrial enzyme is not only favored because of its higher activity compared to the NAD-linked enzyme, but also for kinetic reasons. A second piece of information resulting from the calculation of concentrations is obtained from the mitochondrial/cytosolic gradients. As expected from the observations on isolated mitochondria [23], the gradient is higher for the tricarboxylates than for the dicarboxylate. The results for citrate are of the same order of magnitude as those calculated by Williamson [5] for perfused liver in presence of oleate and by Greenbaum et al. [6] for freeze-quenched liver from fed rats. The distribution of 2-oxoglutarate is also in favor of the mitochondrial space, in marked contrast to the value of less than 0.1 for the mitochondrial/ cytosolic gradients calculated for fed rat liver by the authors of [6],but roughly in agreement with the recent
306
data for fed rats obtained by Soboll et al. [27] with the non-aqueous fractionation procedure [28].
Concluding Remarks First, it should be mentioned that in the present experiments the particular emphasis was on determining the low levels of isocitrate, thus requiring a relatively high cell concentration in the incubation ; the latter obviously is unfavorable to a perfect separation by the digitonin procedure, and, therefore, a compromise was sought between cell concentration on the one hand and the digitonin concentration (plus reaction time) on the other. As indicated by the ATP/ADP ratios, the separation of the particulate fraction from the soluble fraction was comparable to results in the literature and was considered acceptable. The somewhat larger proportion of glutamate dehydrogenase in the supernatant fraction indicates that some back-mixing of the two fractions is not excluded. However, the extent of such back-mixing must be rather limited, since it allowed the isocitrate levels in the mitochondrial fraction to decrease to < 3 nmol/g dry wt (Table 3), for example, when ammonia was present . The following conclusions can clearly be drawn from the present experiments. Firstly, the isocitrate dehydrogenase system can be used as an indicator metabolite system for NADPH in the mitochondrial and cytosolic spaces. Secondly, the isocitrate concentration in the mitochondrial matrix space decreases during ureogenesis from ammonia by at least one order of magnitude, to less than approx. 10 pM. Isocitrate removal occurs at a high rate causing a deviation of the aconitase reaction from near-equilibrium, and also exceeding isocitrate import rates on the tricarboxylate carrier. Thirdly, although the mitochondrial/cytosolic concentration gradients can only be calculated with respect to the order of magnitude, due to possible pitfalls mentioned above, it is clear that the isocitrate gradient is substantially decreased in presence of ammonia. On the other hand, there is relatively little effect on +! ie citrate and 2-oxoglutarate concentration gradients. The first conclusion is in agreement with the conclusions obtained by Krebs and his colleagues regarding the cytosolic space [4] but not the mitochondrial space [29], and the following further comments are given for clarification that mitochondrial NADPH-dependent isocitrate dehydrogenase may be suitable as a redox indicator system (see also the recent discussion in [9,30]). In the absence of added ammonia, the activity of the enzyme is sufficiently high, about 5 U/g liver wet wt, to maintain near-equilibrium, in view of the approximate rate of metabolite flow of 0.8 pmol/min per g liver wet wt
Mitochondria1 and Cytosolic NADPH Systems
through this step 191 and the high concentration of isocitrate compared to the K, (Table4). During the increased rate of flow in the presence of ammonia, activity of the enzyme is high enough to lead to a substantial decrease in the mitochondrial isocitrate/2oxoglutarate ratio in spite of the rapid utilization of 2-oxoglutarate by glutamate dehydrogenase, occurring at a rate of about 2 pmol/min per g liver wet wt in the steady state. The latter can be estimated because under such condition the rate of urea formation from ammonia is equal to the rate of amination of 2-0x0glutarate (cf: Scheme 1 in [17]). It is possible that isocitrate dehydrogenase activity may no longer be in excess under such extreme condition with isocitrate concentration in the region of the K , or lower, leading to an indicator error. At present, the methodological problem of distinguishing isocitrate levels lower than 3 nmol/g liver dry wt from zero precludes an evaluation of the possible indicator error; in other words, it is difficult to estimate whether the redox potential of mitochondrial free NADPH is significantly more positive than -391 mV (Table3) in the presence of ammonia. The present experiments provide further insight into the compartnientation of NADPH redox potentials and the organisation of NAD-dependent and NADP-dependent isocitrate dehydrogenases and energy-linked transhydrogenase, an area of research that lo date has not been completely elucidated [2.9,3 I ] . It is evident that the increased NADPH demand during ureogenesis is accompanied by a corresponding shift in the isocitrate dehydrogenase system : whereas without added ammonia part of isocitrate oxidation proceeds via the NAD-dependent, non-equilibrium enzyme [3], and another part through the NADPdependent enzyme coupled to the energy-linked transhydrogenase, the increase in the NADPH redox potential (as initiated here by the addition of ammonia) causes a shift in favor of the NADP-dependent enzyme, in agreement with other lines of evidence obtained by previous investigators [3,20,32,33]. A new question arising from the present study is whether under the condition of enhanced mitochondrial NADPH utilization there is a transfer of NADPH equivalents from the cytosol into the mitochondria, i. e. in the opposite direction to that usually envisaged, for example in lipogenesis; and whether the mitochondria] oxidation of NADPH leads to an increased formation of NADPH by the pentose phosphate pathway. However, the tricarboxylate carrier may not be capable of equilibrating, since the citrate gradient between mitochondria and cytosol is still considerable in the presence of ammonia, and the cytosolic citrate concentration is decreased toward the K , value of 0.12 mM for the carrier [34]. This limitation may explain the higher degree of reduction of cytosolic NADPH compared to mitochondria1 NADPH as
H. Sies, T. P. M. Akerboom, and J. M. Tager
307
indicated by the isocitrate dehydrogenase system, in the presence of ammonia.
15. Zuurendonk, P. F.,Akerboom, T. P. M. & Tager, J. M . (1976) in Use of' Isolated Liver Ce1l.s and K i d n q Tubules it7 Meruholic Studies (J. M. Tager, H . D. Soling & J. R. Williamson, Expert technical assistance was provided by Ingrid Linke and eds) pp. 17-27, North Holland, Amsterdam. Annegret Marklstorfer. Thanks are due to Prof. Theodor Bucher 16. Bergmeyer, H. U., ed. (1 974) Merhoden der En:yyrnuiischen for his helpful comments and suggestions and to Dr Karl-Hein7 Anul,yse, 3rd edn, Verlag Chemie, Weinheim. Summer for his help in part of the assays. This investigation wii\ 17. Sies. H.. Summer, K . H . , Haussingel-. D & Uiichel-. T l i . (1076) supported by Deursche Forschungs~emeinschufi,SonderJOr.rchun~~.\.in U,se of lsokzted Liver Cells and K i r h q Tiihuli~.~ in Mrlubereich 51, Medizinisrhe Molekularbiologie und Biochemie (Grant holic Studies (J. M. Tager, H. D. Soling & J. R. Willlainson. D/8) and by a grant from the Netherlands Foundation for Pure eds) pp. 311 -316, North Holland, Amsterdam. Scientific Research (ZWO) under auspices of the Netherlands 18. Klingenberg, M. & Slenczka, W. (1959) Bioclim?. Z. 331,486Foundation for Chemical Research (SON). 517. 19. Tager, J . M. (1 966) in Regulution of' Metabolic Procrssra in Mitochondriu (Tager, J. M., Papa, S.. Quagliariello, E. & Slater, E. C., eds) pp. 202-216, Elsevier, Amsterdam. 20. Nicholls, D. G. &Garland, P. B. (1969) Biocheni. J . 114,215REFERENCES 225. 1. Hogebooni, G . H. & Schneider, W. C. (1950) J . Biol. Chem. Chrm. 11, pp. 21. Klingenberg, M . (1961) Colloy. Grs. Ph~~siol. 186,417-427. 82-114. 2. Plaut, G. W. E. (1970) Curr. Top. Cell Regul. 2, 1 -27. 22. Siess, E. & Wieland, 0. (1 975) FEBS Lett. 52, 226 - 230. 3. Goebell, H. & Kiingenberg, M. (1964) Biochem. 2.340, 441 23. Palmieri, F., Quagliariello, E. & Klingenberg, M. (1970) Eur. 464. J . Bioc,hern. 17, 230-238. 4. Veech. R. L., Eggleston, L. V. & Krebs, H. A. (1969) Biochem. 24. Nordmann. R., Petit, M . A . & Nordmann, J. (1972) Biochin7ir. 54, 1473- 1478. J . 115, 609-619. 5. Williamson, J . R. (1969) in The Energy Level and Metabolic 25. Brosnan, J. T . & Williamson, D. H. (1974) Biochem. J . 138, Control in Mitochondria (S. Papa, J. M. Tager, E. Quaglia453 - 462. riello & E. C. Slater, eds) pp. 385-400, Adriatica Editrice, 26. Colman, R. F. (1975) Adv. Enzymr Rcgul. 13, 413-433. Bari. 27. Soboll, S., Scholz, R., Freisl, M.. Elbers, K. & Heldt, H. W. 6. Grecnbaum, A . L., Gumaa, K. A. & McLean, P. (1971) Arch. (1 976) in Use of Isolrrted Liver Cells and Kidney Tubules it7 Bio~he177.B / O , O ~ I(43, ~ , S .617 - 663. Metub~licStudies (J. M. Tager, H.D. Soling & J. R. Williamson, eds) pp. 29 -40, North Holland, Amsterdam. 7. Hoek, J. B. & Ernster, L. (1974) in Alcohol und Aldehyde Mctubolizink. S.vsiems (R. G. Thul-man, T. Yonetani, J . R . Wil28. Elbers, R., Heldt, H. W., Schmucker, R., Soboll, S. & Wiese, H. (1 974) Hoppe-Sc,.vler'.s Z.Plf.vsiol. Chem. 355. 378 - 393. liamson & B. Chance, eds) pp. 351 -364, Academic Press. 29. Krebs, H. A. & Veech, R. L. (1969) in Thr Energ! Lruel ( r i d New Yoi-k. i ~ cuitl hli~rtrholic~ Coiitrol iu Metubolic Control in Mitochondria ( S . Papa, J. M. Tager. 8. Papa. S (1969) in The I k y ~ Loivl Mitochondrirr ( S . Papa, J. M. Tager, E. Quagliariello BL E. C. E. Quagliariello & E. C. Slater, edsj pp. 329-382, Adriatica Slate)', eds) pp. 401 -409, Adriatica Editrice, Bari. Editrice, Bari. Its Regulution in 9. Biicher. Th. & Sies, H. (1976) in Use of Isolated Liver Cells 30. Williamson. J. R . (1 976) in GI~lc~oneogene.sis. und Kidntj, Tubules in Metabolic Studies (J. M. Tager, Mammcrliun Species (R. W. Hanson & M. A. Mehlman, H. D. Soling & J. R. Williamson, eds) pp. 41 -64, North eds) pp. 165-220, J. Wiley & Sons, New York, London, Sydney, Toronto. Holland, Amsterdam. 10. Zuurendonk, P. F. & Tager, J. M. (1974) Biochrm. 5iophy.s. 31. Rydstrom, J., Hoek, J. B. & Ernster, L. (1976) in The EllActu, 333, 393 - 399. rymes (P. D. Boyer, ed.) 3rd edn. vol. 13, pp. 51 -88, Aca11. Chamalaun, R. A. F. M . & Tager, J. M. (1970) Biochini. Biodemic Press, New York. pllys. act^, 222. 1 19 - 134. 32. Stein, A. M., Stein, J . H. & Kirkman, S. K . (1967) 5iochnni.stry.6, 1370-1379. 12. Sies, H., Hiiussingcr, D. & Grosskopf, M . (1974) Hoppc,-S'.v33. Smith, C. M., Beach, R . L. & Plaut. G . W. E. (1975) Ahstr. Ier'x Z. Phy.siol. Cheru. 355, 305 -318. Corninun. 10th M w r . Fed. Eur. Biochm. Sot.., 717. 13. Sies, H., Summer, K. H. & Bucher, Th. (1975) FEBS Lett. 34 Palmieri, F., Stipani, I., Quagliariello, E. & Klingenberg. M. 54,274-278. (1972) Eur. J . Biochem. 26, 587 - 594. 14. Sies, H. & Grosskopf, M. (1975) ELIT.J . Bioc,hrrn.57, 513 -520.
H. Sies, Institut fur Physiologische Chemie, Physikalische Biochemie und Zellbiologie dei- Ludwig-Maximilians-Universitiit Munchen, GoethestraBe 33, D-8000 Munchen 2, Federal Republic of Germany T. P. M. Akerboom and J. M. Tager. Laboratorium voor Biochemie. Universiteit van Amsterdam, B.C.P. Jansen Instituut, Plantage Muidergracht 12, Amsterdam C, The Netherlands
Note Addrdin Pioof(December 28, 1976). The theoretical redox potential difference of the NADP-linked isocitrate dehydrogenase system between mitochondria and cytosol can be calculated from the tricarboxylate and dicarboxylate gradients as was indicated above. Assuming equilibration across the mitochondria] inner membrane according to the Donnan relationship, and with equal COz concentration on either side of the membrane, a pH difference
(more alkaline inside) of 0.41 results from the data in Table 4 for citrate3- and 2-oxoglutarate2-, and 0.33 for isocitrate3-. This corresponds t o a calculated redox potential difference of 0.41 x 30.7 = 12.6 mV (or 10.1 mV, respectively) more negative inside the mitochondria than outside. The analytical data obtained with isolated mitochondria [7] and with the hepatocytes (Table3), thereCore, arc within the range of theoretically expected values.
