Address correspondence and reprint requests to Dr R. Dringen, Physiologisch-chemisches Institut der Universität, Hoppe-Seyler-Str. 4, D-72076 Tübingen, Germany. E-mail: firstname.lastname@example.org
NADP+-dependent isocitrate dehydrogenases (ICDHs) are enzymes that reduce NADP+ to NADPH using isocitrate as electron donor. Cytosolic and mitochondrial isoforms of ICDH have been described. Little is known on the expression of ICDHs in brain cells. We have cloned the rat mitochondrial ICDH (mICDH) in order to obtain the sequence information necessary to study the expression of ICDHs in brain cells by RT-PCR. The cDNA sequence of rat mICDH was highly homologous to that of mICDH cDNAs from other species. By RT-PCR the presence of mRNAs for both the cytosolic and the mitochondrial ICDHs was demonstrated for cultured rat neurons, astrocytes, oligodendrocytes and microglia. The expression of both ICDH isoenzymes was confirmed by western blot analysis using ICDH-isoenzyme specific antibodies as well as by determination of ICDH activities in cytosolic and mitochondrial fractions of the neural cell cultures. In astroglial and microglial cultures, the total ICDH activity was almost equally distributed between cytosolic and mitochondrial fractions. In contrast, in cultures of neurons and oligodendrocytes about 75% of total ICDH activity was present in the cytosolic fractions. Putative functions of ICDHs in cytosol and mitochondria of brain cells are discussed.
sodium dodecyl sulfate–polyacrylamide gel electrophoresis
NADPH is an important source of reduction equivalents which is important for fatty acid synthesis, hydroxylation reactions and the reduction of glutathione disulfide (GSSG) (Numa 1974; Gunsalus et al. 1975; Dringen 2000). During these reactions NADPH is oxidized to NADP+. Subsequently, NADP+ has to be reduced again to NADPH. Enzymes which participate in cellular NADP+ reduction are glucose-6-phosphate dehydrogenase (G6PDH), 6-phosphogluconate dehydrogenase (6PGDH), malic enzymes (MEs) and mitochondrial transhydrogenase, as well as NADP+-dependent isocitrate dehydrogenases (ICDHs). Of these enzymes, G6PDH and 6PGDH are exclusively localized in the cytosol (Baquer et al. 1988) and transhydrogenases are localized in mitochondria (Hoek and Rydstrom 1988; Lee and Ernster 1989). In contrast, for ME both cytosolic and mitochondrial isoforms have been described (Frenkel 1972; Kurz et al. 1993; Vogel et al. 1998).
Since ICDHs can provide NADPH for the regeneration of the antioxidant GSH by glutathione reductase (Rall and Lehninger 1952; Vogel et al. 1999), a participation in the cellular defense against oxidative stress has been considered as one of the main functions of ICDHs. This view has recently been supported by reports showing that overexpression of either cICDH or mICDH improves the resistance of cells against oxidative stress, whereas cells with a reduced activity of one of the two ICDH isoforms are more vulnerable towards oxidative stress (Jo et al. 2001; Lee et al. 2002).
For the brain, presence of mICDH and cICDH has been reported (Loverde and Lehrer 1973; Bajo et al. 2002). On the cellular level, only cultured astroglial cells have been investigated so far for activities of ICDHs (Juurlink 1993). Here we demonstrate that both cICDH and mICDH are expressed in cultured neurons, astrocytes, microglial cells and oligodendrocytes. The high activities of ICDHs in neural cells suggest that these enzymes could contribute substantially to the NADPH supply for anabolic reactions as well as for cellular glutathione redox cycling.
Materials and methods
Bovine serum albumin (BSA), digitonin, insulin, progesterone, putrescine and 5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB) were obtained from Sigma (Deisenhofen, Germany). Acetyl-coenzyme A, anti-rabbit-IgG-alkaline phosphatase, deoxyribonucleoside triphosphate mixtures, fetal calf serum (FCS), the High-Pure PCR Product Purification Kit and oligo(dT)18 adapter primers were purchased from Roche Diagnostics (Mannheim, Germany). 5-Bromo-3-chloro-indolyl phosphate, NAD+, NADP+ and 4-nitroblue tetrazolium chloride were from Applichem (Darmstadt, Germany). Dulbecco's modified Eagle's medium (DMEM) was from Life Technologies (Eggenstein, Germany). d-(+)-threo-isocitrate and sodium selenite were obtained from Fluka (Deisenhofen, Germany). Horse serum, Superscript II reverse transcriptase, TOPO-Kit™ and transferrin were from InvitroGen (Karlsruhe, Germany). Highspeed Plasmid Midi Kit, QIAprep Spin Miniprep-Kit, QIAquick Extraction Kit, Rneasy® kit and Taq-Polymerase-Kit were obtained from Qiagen (Hilden, Germany). Penicillin G, streptomycin sulfate, Tween 20 and Triton X-100 were from Serva (Heidelberg, Germany). Skimmed milk powder was from Humana Milchunion (Herford, Germany). Primers were purchased from MWG-Biotech (Ebersberg, Germany). All other chemicals were obtained at analytical grade from E. Merck (Darmstadt, Germany). Sterile plastic material, 50-mm culture dishes, 175 cm2 flasks for cell culture and non-sterile 96-well microtiter plates were from Nunc (Wiesbaden, Germany) and Greiner (Frickenhausen, Germany).
Antibodies against cICDH and mICDH
Antibodies against cICDH and mICDH were raised by immunization of rabbits with purified rat liver cICDH and a peptide consisting of 14 amino acids of the C-terminus of mouse mICDH, respectively (Yoshihara et al. 2001; Haraguchi et al. 2003).
Astroglia-rich primary cultures derived from the brains of neonatal Wistar rats were prepared and maintained according to Hamprecht and Löffler (1985). Three million viable cells were seeded in 5 mL medium [DMEM containing 10% FCS as well as 20 U/mL penicillin G plus 20 µg/mL streptomycin sulfate (PS)] per 50 mm dish or 30 million viable cells in 50 mL medium per 175 cm2 flask. The cultures were used at an age of 14–21 days. These cultures contain about 90% glial fibrillary acidic protein (GFAP)-positive astroglial cells, up to 5% of both oligodendroglial and microglial cells as well as some ependymal cells (Hamprecht and Löffler 1985; Gutterer et al. 1999) but no neurons (Löffler et al. 1986).
Neuron-rich primary cultures were prepared from the brains of embryonal (E16) Wistar rats as previously described (Löffler et al. 1986; Dringen et al. 1999a). Three million viable cells were seeded into poly d-lysine-coated 50 mm dishes in 5 mL medium (90% DMEM/10% horse serum/PS) and were maintained as described previously. Experiments were conducted at an age of 6 days. These cultures contain approximately 5% astroglial cells (Dringen et al. 1999b) but no oligodendroglial or ependymal cells (Löffler et al. 1986).
Oligodendroglia-rich secondary cultures derived from astroglia-rich primary cultures in 175 cm2 flasks were prepared as recently described (Hirrlinger et al. 2002). One million cells were seeded per poly d-lysine-coated 50 mm dish in 5 mL medium (DMEM, 5 µg/mL transferrin, 5 µg/mL insulin and 25 ng/mL sodium selenite). The cultures were used at an age of 6 days. The cultures contain about 90% oligodendroglial cells and the majority of the remaining cells are astroglial cells (Hirrlinger et al. 2002).
A modification (Hirrlinger et al. 2000) of the method described by Giulian and Baker (1986) was used to prepare microglia-rich secondary cultures from astroglia-rich primary cultures. One million viable cells were seeded per 50 mm dish in 5 mL medium (90% DMEM/10% FCS/PS, half of which was preconditioned on astroglia-rich primary cultures) and used at an age of 3 days. These cultures contain about 90% microglial cells and small quantities of astroglial and oligodendroglial cells (Hirrlinger et al. 2000).
Cloning of rat mICDH cDNA
Total RNA was harvested from astroglia-rich primary cultures from rat brain and reverse transcribed (see below). From the known sequence of mouse mICDH (accession number, AF212319) two primers were designed (5′-CGTGCTCGGACCTCGCGT-3′ and 5′-ATGTCCCTAGAAAGGCCACC-3′) and used to amplify the coding region of rat mICDH cDNA. The PCR product was cloned twice using the TOPO-Kit™ according to the manufacturer's instructions. Four clones were sequenced from both sides to yield reliable sequence information for rat mICDH.
RT-PCR analysis for cICDH and mICDH
Using the RNeasy-kit, total RNA from neural cell cultures and tissues was isolated according to the instructions supplied by the manufacturer. Total RNA (1 µg) was reverse transcribed in 20 µL transcription buffer (75 mm KCl, 50 mm Tris/HCl pH 8.3, 10 mm dithiothreitol, 3 mm MgCl2, 0.5 mm deoxyribonucleoside triphosphates, 1 µg oligo(dT)18-adapter primer) with Superscript II reverse transcriptase (200 U) for 1 h at 37°C. The resulting cDNAs were purified using the High-Pure PCR Product Purification Kit according to the manufacturer's instructions. The sequence of rat mICDH (Fig. 1) as well as the published sequences for rat cICDH (accession number L35317) and rat β-actin (accession number V01217) were used to design primer pairs for the specific amplification of cDNA fragments. The following primers were used: mICDH: 5′-AAGGGAGCCAAATCCTTGAG-3′ and 5′-AACACCGACGAGTCCATTTC-3′; cICDH: 5′-AAAAATCCATGGCGGTTCTGTG-3′ and 5′-GGTCCCCATAGGCGTGTCG-3′; β-actin: 5′-GGGTCAGAAGGACTCCTACG-3′ and 5′-GGTCTCAAACATGATCTGGG-3′. For PCR, in a total volume of 50 µL of PCR buffer (supplied as 10 × reaction buffer by the manufacturer) containing 1.5 mm MgCl2, 1.25 U of Taq DNA polymerase and 0.5 µm forward and reverse primers, 5 µL of purified cDNAs were used as a template. Cycling conditions were as follows: 2 min at 94°C, followed by 30 cycles with 1 min at 94°C, 1 min at 50°C, 1 min at 72°C and a final 10 min at 72°C. Isolated RNA and reverse transcription integrity was tested by PCRs using β-actin specific primers. All PCR products were analyzed on a 1.8% agarose gel and had the size expected for the respective cDNA sequences.
Polyacrylamide gel electrophoresis and western blotting
Samples of different tissues, neural cultures and fractions of astroglia-rich primary culture were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) according to Laemmli (1970) using a 12% acrylamide separation gel and subsequent western blotting. After separation, proteins were transferred to a nitrocellulose membrane by electroblotting. Unspecific binding sites on the nitrocellulose membrane were blocked with 5% (w/v) skimmed milk powder and 0.05% (w/v) Tween 20 in phosphate-buffered saline (PBS, 10 mm potassium phosphate buffer pH 7.4 containing 150 mm NaCl). After incubation over night at 4°C on a roller with anti-cICDH or anti-mICDH (each 1 : 500 diluted in PBS/milk powder/Tween 20), four washing periods in PBS (5 min each), and incubation for 1 h with anti-rabbit-IgG-alkaline phosphatase (1 : 5000 diluted in PBS) the membrane was developed by incubation in 0.2 m Tris/HCl pH 9.5, 10 mm MgCl2, 0.39 mm 4-nitroblue tetrazolium chloride and 0.37 mm 5-bromo-4-chloro-3-indolylphosphate at room temperature. The reaction was stopped by washing the membranes with water.
Separation of cytosolic and mitochondrial fractions
To obtain fractions of neural cell cultures that are strongly enriched in cytosolic and mitochondrial enzymes the cells of the cultures were lysed using digitonin and the mitochondrial matrix enclosed by the inner mitochondrial membrane was separated from the cytosol by centrifugation. Cells grown on 50 mm dishes were washed once with ice-cold separation buffer (10 mm Tris/HCl pH 7.4, 320 mm sucrose and 1 mm EDTA), incubated for 10 min on ice with 100 µg/mL digitonin in separation buffer and scraped off the dish with a rubber policeman. Aliquot parts of the lysates were frozen (− 20°C) for determination of protein content or were mixed with Triton X-100 [final concentration of 1% (w/v)] to lyse the inner mitochondrial membrane (incubation for 30 min on ice) before the total enzyme activities were determined. The majority of the digitonin-lysate was centrifuged for 10 min (16 000 × g, 4°C). The supernatant was collected and was incubated after addition of Triton X-100 to a final concentration of 1% (w/v) on ice for 30 min (cytosolic fraction). The pellet of the centrifugation was resuspended in separation buffer and lysed by sonication [nine pulses of 1 s each, Sonifier Cell Disruptor B-30, Heinemann (Schwäbisch Gmünd, Germany)]. After addition of Triton X-100 to a final concentration of 1% (w/v) the resuspended pellet was incubated for 30 min on ice (mitochondrial fraction). Samples were kept on ice till enzyme activities were determined. Samples for western blotting were stored at −80°C.
Determination of enzyme activities and protein content
The activities of lactate dehydrogenase (LDH), citrate synthase (CS), and ICDHs in total cell lysates as well as in cytosolic and mitochondrial fractions of the lysates were measured at room temperature in a microtiter plate reader (MRX TC Revelation, Dynex Technologies Denkendorf, Germany) in a total volume of 360 µL per well. The activity of LDH was measured as described previously in detail (Dringen et al. 1998) using a reaction mixture containing 80 mm Tris/HCl, 200 mm NaCl, 1.6 mm pyruvate and 0.2 mm NADH at pH 7.2. Activities were calculated from the decrease in the absorbance at 340 nm using the molar absorbency coefficient for NADH. CS activity was determined in a reaction mixture containing 100 mm Tris/HCl pH 8.0, 0.1 mm acetyl-coenzyme A, 0.2 mm DTNB and 0.5 mm oxaloacetate. The coenzyme A generated in the CS reaction reduces DTNB and the rate of thionitrobenzoate formation was determined spectrophotometrically at 405 nm. Isocitrate dehydrogenase activity was determined by a modification of the assay described by Goldberg and Ellis (1983) for microtiter plates. The conversion of NADP+ to NADPH was monitored at 340 nm in a reaction mixture containing 65 mm triethanolamine hydrochloride/NaOH buffer pH 7.3, 3.35 mm D(+)-threo-isocitrate, 0.42 mm NADP+ and 1.67 mm MnCl2. The protein content of samples was determined by the method described by Lowry et al. (1951) using BSA as a standard.
Presentation of data
The experiments shown in the figures were carried out on at least three independent cultures with comparable results if not mentioned otherwise. The results shown are presented as mean values ± SD of all measured plates. Statistical analysis was performed using anova followed by Tukey's post hoc test.
Cloning of rat mICDH cDNA
cDNA was generated by reverse transcription of RNA harvested from rat astroglia-rich primary cultures and used as source for the cloning of the coding sequence of rat mICDH. Figure 1 shows the cDNA sequence obtained as well as the predicted amino acid sequence for rat mICDH. Comparison of the rat mICDH sequence with known mICDH sequences of other species revealed high identities in the nucleotide (85–96%) as well as in the predicted amino acid sequence (87–98%) (Table 1). In contrast, between rat mICDH and cICDH the identities were only 68% (Table 1). The predicted amino acid sequence of rat mICDH contains an N-terminal mitochondrial targeting sequence of 39 amino acids (Fig. 1). The molecular mass of rat mICDH was calculated to be 51.0 kDa (including this targeting sequence). In the absence of the mitochondrial targeting sequence the molecular mass is reduced to 46.6 kDa.
Table 1. Comparison of the cloned rat mICDH sequence on cDNA and amino acid level with sequences of known ICDHs
Identity to rat mICDH (%)
Accession number (EMBL)
The identities in cDNA and amino acid sequences were determined for the sequences lacking their respective mitochondrial targeting sequence. *The published pig sequence for mICDH appears to lack most of the mitochondrial targeting sequence. In the absence of the targeting sequence the length of the pig mICDH is identical to those of rat, mouse and cattle.
Presence of mRNAs and proteins of cICDH and mICDH in cultured neural cells
To demonstrate expression of mRNAs of cICDH and mICDH in cultures of four neural cell types RT-PCR analyses were performed using specific primers for the amplification of cDNAs of both isoforms of ICDH. Separation of the amplification products by agarose gel electrophoresis revealed amplification product for both ICDH isoforms (404 bp for cICDH and 314 bp for mICDH) of the expected sizes for all four types of neural cell cultures as well as for brain (Fig. 2). Control PCRs performed in the absence of cDNA did not generate any detectable band (Fig. 2).
Antibodies generated against purified cICDH and a peptide of mICDH were used to demonstrate the presence of the two ICDH isoenzymes by western blot analysis. Positive signals at about 47 kDa were obtained by western blots for cICDH and mICDH in liver and heart, respectively (data not shown), indicating that the antibodies used recognized specifically either mICDH or cICDH and did not cross-react with the other isoform. For homogenates of cultured astrocytes, oligodendrocytes, microglial cells and neurons protein bands of about 47 kDa were detected using antibodies against cICDH (Fig. 3). Antibodies against mICDH also detected a protein of about 47 kDa but recognized in addition proteins of higher molecular mass in the homogenates of neural cell cultures (Fig. 3).
Total activities of ICDH, LDH and CS in lysates of cultured brain cells
The total specific activities of ICDHs, LDH and CS determined in homogenates of the four types of neural cell cultures differed strongly (Table 2). Compared with the specific LDH activity [1.11 µmol/(min × mg protein)] and to the specific CS activity [116 nmol/(min × mg)] of cultured astrocytes, cultures of neurons, oligodendrocytes and microglial cells contained 30, 150 and 112% of LDH activity and 100, 31 and 50% of CS activity, respectively (Table 2). In addition, cultured astrocytes contained the highest specific total ICDH activity of the four culture types investigated as well as the highest protein content per dish (Table 2).
Table 2. Specific activities of LDH, CS, and ICDH in total lysates of neural cell cultures
Specific activity [µmol/(min/mg)] in total lysates
Protein content (mg/dish)
Culture dishes (n)
Lysates were prepared as described in Methods. The data represent mean values ±SD of n dishes derived from two (neuron-rich) or three (other culture types) independently prepared cultures. The significance of differences of the specific activities of neural cultures compared with astroglia-rich cultures were analyzed using anova followed by Tukey's post hoc test. *p < 0.001.
Separation of cell lysates in cytosolic and mitochondrial fractions
To quantify the activities of cICDH and mICDH in neural cell cultures, the cells were lysed by digitonin treatment and the lysate was separated in a mitochondrial fraction and a cytosolic fraction (see methods). The quality of this separation was tested by determining the activities of LDH and CS as marker enzymes for cytosol and mitochondria, respectively. The distribution of the marker enzymes LDH and CS revealed that more than 90% of the total cellular LDH activity was recovered in the cytosolic fraction for all cell culture types analyzed (Fig. 4a). Less than 10% of the total LDH activity was measured in the mitochondrial fraction. In contrast, the mitochondrial fraction contained at least 70% of the total CS activity and in the cytosolic fraction of cultured astrocytes, neurons, microglial cells and oligodendrocytes only 5%, 11%, 6% and 12%, respectively, of the total CS activity were determined (Fig. 4b). Therefore, marker enzyme analysis demonstrated that the two fractions separated were strongly enriched in cytosolic and mitochondrial enzymes.
cICDH and mICDH in cultured neural cells
The separation of cICDH and mICDH by the fractionation method applied was investigated by western blot analysis. Antibodies against cICDH detected in total lysates of astroglia-rich cultures as well as in the cytosolic fraction of the lysates a band of approximately 47 kDa. In contrast, no signal was obtained for the mitochondrial fraction. In reverse, the antibodies against mICDH detected mICDH in the lysate of the cultures as well as in the mitochondrial fraction, but not in the cytosolic fraction (Fig. 5).
ICDH activity was measured in both the cytosolic and the mitochondrial fraction of the four neural cell culture types. All four types of brain cells contained activities of both cICDH and mICDH (Fig. 4c). The total activity of ICDH determined for cultured astrocytes and microglial cells was distributed almost equally between cytosol and mitochondria. Cultured neurons and oligodendrocytes contained less than 30% of the total ICDH activity in the mitochondria and about 75% in the cytosol (Fig. 4c).
A cDNA sequence from astroglia-rich cultures was amplified using primers for mouse mICDH. The amino acid sequence postulated from the cloned cDNA sequence showed a much lower identity (68%) to the sequence of rat cICDH (Jennings et al. 1994) than to that of mICDHs of several species (87–98%), indicating that the cDNA cloned was the rat mICDH. The N-terminus of mICDH contains a 39 amino acid-long mitochondrial targeting sequence rich in amino acids with basic or hydroxyl side chains that are characteristic for such sequences (Nielsen et al. 1997; Emanuelsson et al. 2000). Similar mitochondrial targeting sequences have been described for mICDH of mouse (Jo et al. 2001) and cattle (Huh et al. 1993). Sequencing of purified mICDH revealed that the mitochondrial targeting sequence is not present in the enzymes purified from heart and kidney (Haselbeck et al. 1992; Huh et al. 1993). For neural cells it is not known so far whether the mitochondrial targeting sequence of mICDH is removed from the enzyme when it enters the mitochondria. All neural cell cultures analyzed contained mature mICDH as indicated by the protein band of an apparent molecular mass of 47 kDa. However, we cannot exclude that one of the bands of higher molecular mass in the mICDH western blots represents uncleaved mICDH that still contains the mitochondrial targeting sequence.
To quantify the activities of cICDH and mICDH in cultured neural cells, cytosolic and mitochondrial fractions of neural cell cultures were obtained by digitonin lysis of cells and centrifugation. The procedure used for fractionation was rapid and suitable also for the small numbers of cells present in neuron-rich, microglia-rich and oligodendroglia-rich cultures with protein contents of 0.2–0.4 mg per culture dish. This represents a useful alternative to other fractionation methods (Loverde and Lehrer 1973; Almeida and Medina 1998; Yoshihara et al. 2001), at least for studies that do not require coupled mitochondria. Since determination of marker enzyme activities in the cytosolic and mitochondrial fractions demonstrated a high recovery of total activities as well as low cross-contamination with the respective other fraction, the method used was considered suitable to quantify activities of ICDH in cytosolic and mitochondrial fraction. Western blot analysis of total lysates as well as cytosolic and mitochondrial fractions of astroglial cultures using specific antibodies for cICDH and mICDH confirmed that cICDH was present in the cytosolic fraction and mICDH in the mitochondrial fraction.
The expression of both cICDH and mICDH in cultured astrocytes, neurons, oligodendrocytes and microglial cells was demonstrated by RT-PCR and by western blot analysis. Under the conditions used, the RT-PCRs give only the qualitative information that mRNAs of the two isoforms of ICDH are present in the cultures investigated. The signals for the amplification products cannot be used for quantitative comparison of the amounts of ICDH mRNA present in the different neural culture types. In addition, since the cultures used contained besides the predominant cell type of the respective culture also minor amounts of other types of brain cells (see Materials and methods), we cannot exclude that mRNAs derived from these contaminating cells could contribute to the strong positive signals obtained for ICDHs by the highly sensitive RT-PCRs.
The presence of cICDH and mICDH protein in neural cell cultures was demonstrated by western blot analysis. The band intensities for cICDH and mICDH did not correspond well with the specific activities determined in the neural cell cultures, indicating that under the conditions used the western blots cannot be used to gain reliable information on the activities of ICDHs in the cells. The presence of inactive ICDHs, which could still be recognized as antigenes in the western blots, may be a reason for the discrepancy between specific activities of ICDHs and band intensities in western blots. Also, a comparison of the band intensities of mICDH and cICDH for one type of neural cell culture does not reflect the ratio of specific activities determined. Reasons for this are most likely that the two antibodies were applied in different concentrations and that they differ in their affinity for their respective antigenes. Consequently, the western blot data demonstrate the presence of the two isoforms of ICDH in neural cell cultures but do not allow quantitative comparison of the presence of active ICDHs. However, such information is provided by the activities of ICDH determined in the cytosolic and the mitochondrial fractions of cultured neural cells. These results confirmed the presence of cICDH and mICDH in neural cells that had been described for brain (Loverde and Lehrer 1973; Bajo et al. 2002) and astroglial cultures (Juurlink 1993).
Of the cell cultures investigated, astrocyte cultures contained the highest total specific activity of ICDH. This activity [65 nmol/(min/mg)] was lower than that determined earlier for ICDHs in astroglial cultures [Juurlink 1993: 158 nmol/(min/mg)]. Reasons for this discrepancy are most likely different culture conditions and/or the differences in sample treatment. The total specific ICDH activities of the four types of neural cell cultures investigated here varied only by a factor of two. For astrocyte and microglial cultures the total ICDH activities were equally distributed to cICDH and mICDH. In contrast, in cultures of neurons and oligodendrocytes 75% of total ICDH activity was found in the cytosolic fraction. The data obtained in the present study for the distribution of total ICDH activity between cytosolic and mitochondrial fractions are in the range reported for brain and astroglial cultures which contain 66% and 36%, respectively, of the total ICDH activity in the cytosolic fraction (Loverde and Lehrer 1973; Juurlink 1993).
The two enzymes of the oxidative part of the pentose phosphate pathway, G6PDH and 6PGDH, have been considered as predominantly responsible for NADPH regeneration in the cytosol (Ben Yoseph et al. 1994). However, under physiological conditions G6PDH is almost completely inhibited by the cellular concentration of NADPH (Eggleston and Krebs 1974; Loreck et al. 1987). In addition, G6PDH has to be supplied with glucose-6-phosphate as substrate. Moreover, at least in astrocytes, most G6PDH appears to be bound as inactive enzyme to a structural intracellular element and requires activation to be released into the cytosol (Garcia-Nogales et al. 2003). Therefore, cICDH could play an important role in addition to the pentose phosphate enzymes in cytosolic NADPH regeneration, especially under conditions when G6PDH is not activated or the supply of glucose-6-phosphate is insufficient. For hexose-depleted retina the importance of ICDH in providing NADPH for the regeneration of GSH from GSSG has been demonstrated (Winkler et al. 1986). In addition, in glucose-fed cells cICDH is involved in NADPH recycling for the GSH regeneration. Cells with a reduced activity of cICDH contain an elevated GSSG/GSH ratio and are more vulnerable against oxidative stress (Lee et al. 2002).
cICDH activity requires the availability of the substrate isocitrate in the cytosol. This compound is generated from citrate in reactions catalyzed by aconitases (Beinert and Kennedy 1993; Gruer et al. 1997). Alternatively, isocitrate could be taken up from the extracellular fluid. However, at least for astroglial cultures no evidence for a utilization of extracellular isocitrate has been observed (Kussmaul et al. 1999). Citrate and isocitrate are intermediates of the citric acid cycle in mitochondria. These tricarboxylates can be transported into the cytosol by the tricarboxylate carrier of the inner mitochondrial membrane (Kaplan et al. 1993). Therefore, the substrate isocitrate can either be provided to cICDH by direct export from mitochondria or by cytosolic aconitase-mediated isomerization of citrate exported from mitochrondria.
High activities of mICDH were found in all neural cell cultures investigated and, as the mitochondrial volume is small, a high volume activity of mICDH has to be assumed. Also, as the human brain consumes 20% of the oxygen used by the human body (Clarke and Sokoloff 1999) and because mitochondria produce reactive oxygen species as by-products of respiration (Halliwell and Gutteridge 1999; Wallace 2001), NADPH regeneration for glutathione redox cycling is likely to be the most important function of mICDH in brain cells. This hypothesis is supported by the findings that mICDH provides NADPH for GSSG reduction in brain mitochondria (Vogel et al. 1999) and that overexpression of mICDH protects cells against oxidative stress mediated damage (Jo et al. 2001).
In mitochondria, isocitate is continuously generated as intermediate of the citric acid cycle. This isocitrate can either be used as substrate of the NAD+-dependent IDH that provides NADH for the respiratory chain or by mICDH. In mitochondria of astrocytes, the specific activity of IDH is significantly lower than that of ICDH (Juurlink 1993). In addition, the KM value for isocitrate of IDH is higher than that of mICDH (Dalziel 1980). Consequently, it is most probably not the presence of isocitrate but the availability of the electron acceptors NAD+ and NADP+ that determines whether IDH or mICDH decarboxylates isocitrate in mitochondria. Under unstressed conditions the mitochondrial concentration of NADP+ is likely to be low compared with that of NAD+, as NAD+ is continuously generated from NADH by complex I of the respiratory chain. That would favour the reaction catalyzed by IDH. However, an increase in the mitochondrial NADP+ concentration due to consumption of NADPH by glutathione redox cycling would favour the reaction catalyzed by mICDH.
As an alternative to a function in NADP+ reduction, mICDH has been suggested to participate in an electron cycle mediated by the reactions catalyzed by IDH, transhydrogenase and mICDH (Sazanov and Jackson 1994). Reduction equivalents will be transfered from the NADH generated by the IDH reaction via the transhydrogenase to NADP+. In this cycle the isocitrate that has been decarboxylated by IDH would be regenerated by mICDH. The synthesis by mICDH of isocitrate from α-ketoglutarate and CO2 would be supported by the two orders of magnitude higher affinity of mICDH for NADPH than for NADP+ (Reynolds et al. 1978). Indeed, a reverse flux through mICDH has been demonstrated for liver (Des Rosiers et al. 1994) and recently for the heart (Comte et al. 2002). An isocitrate cycling by IDH, transhydrogenase and mICDH would have thermogenetic functions due to the uncoupling of the proton gradient over the inner mitochondrial membrane by transhydrogenase. In addition, the cycle could have regulative functions for the citric acid cycle (Sazanov and Jackson 1994; Comte et al. 2002). An involvement of mICDH in such a cycling in brain and in neural cells remains to be elucidated.
In conclusion, mitochondrial and cytosolic ICDHs are expressed in cultured brain cells. Since substantial activities of both isoenzymes of ICDH have been found for cultured astrocytes, neurons, oligodendrocytes and microglial cells, ICDHs play most likely an important role in NADPH regeneration in cytosol and in mitochondria of these cells in brain. Further studies are required to confirm the expression of mICDH and cICDH for the different neural cell types in vivo and to investigate the demand of the different brain cells for ICDH-derived NADPH under physiological and pathological conditions.
RD would like to thank the Neurosciences Victoria/Monash Institute of Neurological Diseases for a senior research fellowship.