Rhodoquinone and complex II of the electron transport chain in anaerobically functioning eukaryotes.

Many anaerobically functioning eukaryotes have an anaerobic energy metabolism in which fumarate is reduced to succinate. This reduction of fumarate is the opposite reaction to succinate oxidation catalyzed by succinate-ubiquinone oxidoreductase, complex II of the aerobic respiratory chain. Prokaryotes are known to contain two distinct enzyme complexes and distinct quinones, menaquinone and ubiquinone (Q), for the reduction of fumarate and the oxidation of succinate, respectively. Parasitic helminths are also known to contain two different quinones, Q and rhodoquinone (RQ). This report demonstrates that RQ was present in all examined eukaryotes that reduce fumarate during anoxia, not only in parasitic helminths, but also in freshwater snails, mussels, lugworms, and oysters. It was shown that the measured RQ/Q ratio correlated with the importance of fumarate reduction in vivo. This is the first demonstration of the role of RQ in eukaryotes, other than parasitic helminths. Furthermore, throughout the development of the liver fluke Fasciola hepatica, a strong correlation was found between the quinone composition and the type of metabolism: the amount of Q was correlated with the use of the aerobic respiratory chain, and the amount of RQ with the use of fumarate reduction. It can be concluded that RQ is an essential component for fumarate reduction in eukaryotes, in contrast to prokaryotes, which use menaquinone in this process. Analyses of enzyme kinetics, as well as the known differences in primary structures of prokaryotic and eukaryotic complexes that reduce fumarate, support the idea that fumarate-reducing eukaryotes possess an enzyme complex for the reduction of fumarate, structurally related to the succinate dehydrogenase-type complex II, but with the functional characteristics of the prokaryotic fumarate reductases.

Living with hypoxia or even anoxia is an everyday experience for many organisms. Not only many prokaryotes, but many eukaryotic organisms as well can function (temporarily) without oxygen. Parasitic helminths, freshwater snails, and some lower marine organisms are known to be able to survive anaerobic conditions by adaptation of their energy metabolism. In addition to simple fermentation in which glucose is degraded to ethanol or lactate, most of these facultative anaerobic eukaryotes contain another fermentation variant, malate dismutation ( Fig. 1). Malate dismutation is found in both strictly and facultative anaerobically functioning prokaryotes as well as in some eukaryotes that are capable of functioning anaerobically, like parasitic helminths (1), freshwater snails (2), mussels (3), oysters (4), and lugworms and other marine invertebrates (5). Although several variations of malate dismutation with various end products occur, the use of the production of succinate as an electron sink is universal. The reduction of malate to succinate occurs in two reactions that reverse part of the Krebs cycle, and the reduction of fumarate is the essential NADHconsuming reaction to maintain redox balance. Therefore, the possibility arises that during anoxia, fumarate reduction occurs by reversal of the succinate dehydrogenase (SDH) 1 already present in the aerobic respiratory chain or else is catalyzed by a fumarate reductase (FRD) specifically synthesized for this function (reviewed in Refs. 6 -8). Bacteria contain two distinct enzyme complexes, succinate-ubiquinone oxidoreductase (complex II) and menaquinol-fumarate oxidoreductase (FRD), for oxidizing succinate and reducing fumarate, respectively (7), although each enzyme will catalyze both reactions in vitro. These electron-transferring enzyme complexes of fumarate-reducing eukaryotes have not been studied extensively, but it has been shown that Haemonchus contortus possesses two different genes for the B-subunit of complex II that are differentially expressed during the development of this parasite (9). This differential expression during development was later confirmed for another parasitic worm, Ascaris suum, in which the existence of two different stage-specific forms of complex II was also demonstrated (10). The enzyme complexes responsible for fumarate reduction and the quinones of succinate-producing eukaryotes other than parasitic helminths, however, have not yet been studied.
Parasitic helminths are known to contain two distinct quinones, ubiquinone (Q) and rhodoquinone (RQ), for transport of electrons between the enzyme complexes of the respiratory chain. Free-living aerobic stages of several helminths contain a higher ratio of Q to RQ than parasitic anaerobic stages (11)(12)(13)(14). However, evidence for a quantitative correlation between the amount of RQ present and the importance of fumarate reduc-tion in the energy metabolism of parasitic helminths is still lacking. Furthermore, it is still unknown whether RQ is also present in other eukaryotes that reduce fumarate.
This report demonstrates that the percentage RQ of the total amount of quinones is correlated with the importance of fumarate reduction in vivo in all examined fumarate-reducing eukaryotes. In addition, during the development of the liver fluke Fasciola hepatica, a clear correlation was observed between the quinone composition and the importance of fumarate reduction in vivo. Therefore, RQ is an essential component for fumarate reduction in eukaryotes, in contrast to prokaryotes, where menaquinone serves this function (15,16).
The wild-type inbred strain SE (susceptible Edinburgh) of H. contortus was cultured and propagated in our own facilities (9). F. hepatica metacercariae were obtained from experimentally infected snails (Lymnaea truncatula). The snails were placed in aerated cold tap water to induce shedding of cercariae, which then encysted on cellophane paper. F. hepatica adults were isolated from infected sheep livers obtained at a slaughterhouse. F. hepatica adults were also obtained from male Wistar rats, which were orally infected with 30 metacercariae each. Lymnaea stagnalis and Dictyocaulus viviparus were a generous gift of Dr. M. de Jong-Brink (Faculty of Biology, Free University, Amsterdam) and Dr. M. Eysker (Department of Parasitology and Tropical Veterinary Medicine, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands), respectively. Mussels (Mytilus edulis), oysters (Crassostrea angulata), and lugworms (Arenicola marina) were collected at the Dutch coast.
Homogenization and Fractionation-Before homogenization, all isolated adult parasites were washed in physiological saline containing 10 mM glucose, 25 mM NaHCO 3 , and 20 mM Hepes (pH 7.4) at 37°C to remove host material. All tissues were homogenized in 20 mM Hepes (pH 7.4), 150 mM KCl with a Teflon-glass homogenizer. H. contortus L3 larvae and F. hepatica metacercariae were crushed with a pestle in a liquid nitrogen-cooled mortar or crushed with quartz sand in a mortar, respectively, before they were homogenized by an Ultraturrax homogenizer.
Mitochondrial fractions were prepared by differential centrifugation. Homogenates from which the cell debris was removed (10 min at 4°C and 600 ϫ g) were centrifuged for 20 min at 4°C and 14,000 ϫ g. The mitochondrial pellet was resuspended in homogenization buffer. Mitochondrial fractions were used in all experiments, except when the available amount of tissue was limited. The results obtained with homogenates and mitochondrial fractions of adult F. hepatica were the same, except for the quinone concentration/mg of protein, which was higher in mitochondrial fractions.
Enzyme Assays-To ensure that changes in fumarate reduction activities are not caused by changes in quinone composition, but really reflect changes in the enzyme complex itself, we used the direct assay of fumarate reductase activity with reduced benzyl viologen as electron donor described by Ackrell et al. (17), which has been used to distinguish SDH and FRD in prokaryotes. Succinate dehydrogenase activity was determined as described by Hä gerhäll et al. (18). The measured activity was corrected for nonspecific reduction of Q 0 by performing control assays with malonate instead of succinate. All enzyme assays were performed immediately after fractionation of freshly collected material. Protein was determined by a Lowry method as reported by Bensadoun and Weinstein (19) using bovine serum albumin as a standard.
Quinone Determination-Quinones were extracted from lyophilized homogenates or lyophilized mitochondrial preparations according to Zhu et al. (20). Lyophilized samples were crushed into powder before extraction with pentane. Four subsequent extractions were performed each time using five times as much pentane as the wet weight of the sample (v/w). During the third extraction, the sample was sonicated three times for 15 s on ice. Extracts were pooled, evaporated to dryness, and dissolved in ethanol, after which they were kept in the dark until HPLC analysis. Quinones were separated according to the HPLC method of Takamiya et al. (12) using a reversed-phase RP18 HPLC column (Ligosphere 100, 5 m, end-capped, 4 ϫ 250 mm; Merck, Darmstadt, Germany). The quinones were eluted using a linear gradient from 15 to 21% diisopropyl ether in methanol (v/v) in 24 min. The eluted quinones were identified by comparing their retention times with Q 7 , Q 9 , and Q 10 standards and purified RQ 10 of F. hepatica. In addition, RQ 10 of F. hepatica was identified as RQ 10 by mass spectrometry analysis. RQ of the other examined species demonstrated four characteristics identical to RQ of F. hepatica: (i) the retention time on reversedphase HPLC; (ii) the retention value on TLC (21); (iii) the oxidized and reduced (by potassium borohydride) absorption spectra (230 -600 nm), which were recorded using a split-beam spectrophotometer (Aminco DW2A); and (iv) a purple color in concentrated ethanol solution instead of the yellow color of ubiquinones.
The quinones were spectrophotometrically quantified at 275 nm (RQ 10 E 1 cm 1% ϭ 126.8 (21) and Q 10 E 1 cm 1% ϭ 165 (22)). Q 7 was used as an internal standard and was added to each homogenate or mitochondrial fraction (followed by gentle sonication, three times for 10 s) before the samples were lyophilized. Quinone amounts were corrected for the recovery of the Q 7 internal standard (between 55 and 90% for all samples).

RESULTS AND DISCUSSION
Quinone Composition-The electron transport chains of many bacteria employ menaquinone when fumarate is the final electron acceptor (15,16,(23)(24)(25). In parasitic helminths, which also utilize fumarate as final electron acceptor during anoxia, Allen (11) demonstrated, however, the presence of RQ. Since RQ is present mainly in anaerobic, fumarate-reducing stages of parasitic helminths, it was suggested that rhodoquinol functions as electron donor in fumarate reduction, similar to menaquinol in fumarate reduction in other organisms (11). It is unknown whether RQ occurs only in parasitic helminths or is present in other fumarate-reducing eukaryotes as well, which would imply an important difference between prokaryotic and eukaryotic fumarate reduction (8). All examined mitochondria from eukaryotes contain Krebs cycle activity and a respiratory chain. All these mitochondria contained Q (Table I), which is an essential component of the aerobic respiratory chain. The amounts of Q present in bovine and rat heart mitochondria (Table I) were in accordance with other reports (26 -28). The observed amounts of total quinones in the examined parasitic helminths were comparable to the amounts of quinone in A. suum and Paragonimus westermani, two other parasitic helminths (12,13). The amounts of Q in the abductor muscles of M. edulis and C. angulata were lower than those in parasitic helminths.
All examined organisms contained quinones with 9 or 10 isoprenoid units (Table I), which is consistent with the findings in other eukaryotes. The molecular basis that causes the differences in quinone chain length, as well as possible functional differences, is not resolved yet.
The mitochondria from rat or bovine heart, which are strictly aerobic and possess complex II functioning in vivo in the direction of succinate oxidation, contained only Q and no RQ (Table  I). On the other hand, RQ was present not only in the investigated parasitic helminths, but also in all examined eukaryotes that reduce fumarate under anaerobic conditions in vivo, like the marine organisms M. edulis, C. angulata, and A. marina and the snail L. stagnalis (Table I). In lower unicellular eukaryotes that reduce fumarate during anoxia, like Euglena gracilis (29), RQ was found (21), whereas in those unicellular species that do not reduce fumarate during anoxia, like rumen protozoa from cattle (30), no RQ could be detected. 2 These results demonstrate a general and essential function of RQ in fumarate reduction in eukaryotes, in contrast to prokaryotes, which use menaquinone in fumarate reduction.
Interestingly, fumarate-reducing members of the family of Rhodospirillaceae, which are purple non-sulfur bacteria, contain RQ as well (31). The reason for this enigmatic difference in these Rhodospirillaceae compared with all other known fumarate-reducing prokaryotes is still unknown. On the other hand, Ochman and Wilson (32) described a high nucleotide sequence similarity between 16 S rRNA of Rhodospirillaceae and that of mitochondria, indicating a phylogenetic relation between these prokaryotes and eukaryotes.
In addition to the presence of RQ in all examined fumaratereducing eukaryotes, a correlation was found between the ratio of RQ to Q and the dependence on fumarate reduction in the anaerobic energy metabolism of these species. F. hepatica adults produce almost exclusively acetate and propionate as end products of their energy metabolism, in which fumarate reduction is an essential step (33). Hence, they possess an enzyme complex that functions in the direction of fumarate reduction, which correlates with the very high RQ content as a percentage of the total amount of quinones (Table I). Adult H. contortus produces mainly acetate, succinate, and propionate (9), which correlates well with the observed high content of RQ. In contrast to F. hepatica, adult H. contortus is also partly dependent on Krebs cycle activity (9), which is reflected in the lower ratio of RQ to Q as compared with F. hepatica (Table I).
On the other hand, free-living stages of both F. hepatica and H. contortus contained mainly Q, which correlates with the aerobic energy metabolism in these stages (1). The snail L. stagnalis contained a low but substantial amount of RQ compared with Q, which correlates with its mainly aerobic metabolism in vivo together with its significant anaerobic capacity (2). The examined facultative anaerobic lower marine organisms contained a substantial amount of both Q and RQ. This correlates with their energy metabolism, which changes every 6 h with the tides of the sea. Part of the time, fumarate is reduced when the organisms function anaerobically, but when they rely on Krebs cycle activity, succinate is oxidized by complex II in these organisms. Hence, both Q and RQ are required within a limited time span, and therefore, both substances should be present in substantial amounts as the half-life of quinones is generally on the order of days (21,34,35). The absolute amount of RQ in these lower marine organisms is low as compared with parasitic helminths. However, marine invertebrates are known to contain, besides fumarate reduction, other fermentative pathways, which produce specific end products like octopine, alanopine, and strombine (36).
To confirm the functional importance of RQ in fumarate .4% of total quinone). d Q 10 was present, but was Ͻ10% of the total quinone amount. e RQ was not detectable (Ͻ0.04 nmol/mg of protein and Ͻ0.5% of total quinone). f Q 9 was present, but was Ͻ3% of the total quinone amount. g RQ 9 and Q 9 were present, but were Ͻ1% of the total quinone amount. h RQ 9 and Q 9 were present, but were Ͻ0.1% of the total quinone amount. i RQ 10 and Q 10 were present, but were Ͻ2% of the total quinone amount. j F. hepatica adults were isolated from infected sheep livers obtained from a slaughterhouse. k RQ 10 and Q 10 were present, but were Ͻ10% of the total quinone amount.
reduction in eukaryotic organisms, the correlation between the relative amount of RQ compared with Q and the importance of succinate production was investigated in the developing F. hepatica liver fluke. Tielens et al. (33) demonstrated that the aerobically functioning juvenile liver fluke (F. hepatica) uses Krebs cycle (and hence SDH) activity. This aerobic energy metabolism is gradually replaced by an anaerobic energy metabolism as the juvenile develops into an adult. The energy metabolism of the adult liver fluke is almost exclusively dependent on malate dismutation and hence fumarate reduction. From these metabolic studies, it can be calculated how many electrons have to be transferred per unit of time during the development of the liver fluke, from complexes I and II to complex III by Q in the aerobic respiratory chain, as well as how many electrons have to be transferred by RQ from complex I to complex II to reduce fumarate. As shown in Fig. 2, in the juvenile liver fluke, electrons have to be transferred exclusively by Q. However, during its development, gradually more electrons have to be transferred by RQ. As also shown in Fig. 2, during the development of F. hepatica, a clear correlation was found between the amount of RQ and Q and the amount of electrons that have to be transferred by RQ and Q, respectively. The amount of RQ compared with Q increased during the development of the liver fluke from ϳ17 to 84%. Interestingly, electron transport via Q, as well as the Q content, drastically decreased during development, in contrast to the rather constant electron transfer via RQ and the RQ content. This result corresponds with the observation that the juvenile liver fluke is already fully equipped for anaerobic functioning (37) as the RQ content is not substantially increased during development of the liver fluke. As it was not technically feasible to analyze the quinone composition of stages of F. hepatica from liver (6 -25 days after infection), we were not able to investigate whether the observed increased rates of electron transport via Q as well as via RQ are accompanied by increased concentrations of quinones in that period (Fig. 2). Surprisingly, adult F. hepatica isolated from infected sheep livers obtained from a slaughterhouse had a higher RQ content than adult F. hepatica isolated from experimentally infected male Wistar rats (compare Table I and Fig. 2). The reason for this difference is still unknown, but it might be correlated with the increased egg production in mature liver flukes, which requires an enhanced energy metabolism.
Our results demonstrated that RQ is an indispensable component for efficient electron transport in the anaerobic electron transport chain of eukaryotic organisms. This implies an important difference compared with fumarate reduction in prokaryotes, which utilize menaquinone instead of RQ for fumarate reduction. Both menaquinone and RQ have similar standard redox potentials, Ϫ74 and Ϫ63 mV, respectively, in contrast to Q (EЈ 0 ϭ ϩ100 mV). These standard redox potentials correlate with the function of RQ and menaquinone in fumarate reduction and with the function of ubiquinone in succinate oxidation (8). Interestingly, however, Q and RQ are both benzoquinones, whereas menaquinone is a naphthoquinone. Therefore, the complexes of eukaryotes or prokaryotes that function in vivo as an FRD interact with structurally distinct quinones, suggesting a difference in enzyme structure as well. This possible difference in structure between fumaratereducing enzyme complexes of eukaryotes and those of prokaryotes was therefore further investigated by analysis of their kinetic properties and diode-like behavior (see below).
Kinetic Properties of Complex II-In prokaryotes, two distinct enzyme complexes are known to exist, one to oxidize succinate (SDH) and another one to reduce fumarate (FRD) (but it should be realized that in vitro, both enzyme complexes are able to catalyze the reaction in both directions). Earlier investigations on parasitic helminths (9, 10) suggested that distinct enzyme complexes, which are adapted to their in vivo function, exist in eukaryotes as well.
When measured, the activity ratio of succinate oxidation and fumarate reduction reflects the catalytic capacity of the total amount of enzymes present as one cannot distinguish between a reversible reaction in a single type of enzyme and the occurrence of two distinct enzyme complexes. The activity ratio is, however, still informative because a low ratio indicates the presence of a fumarate-reducing type of enzyme complex. Analysis of the activity ratios of succinate oxidation and fumarate reduction in mitochondrial fractions showed that complexes II from rat and bovine heart mitochondria were twice (1.9 and 2.2, respectively) as active in the direction of succinate oxidation as in that of fumarate reduction. This corresponds with their in vivo function as these strictly aerobically functioning mitochondria contain Krebs cycle activity and therefore have to oxidize succinate. Mitochondria of the other species listed in Table I, which all contain RQ, demonstrated a low SDH/FRD activity ratio (0.09 -0.35). On the other hand, significant differences in activity ratios were not detected between different stages of F. hepatica and H. contortus, although in vivo, the free-living stages of these parasites oxidize succinate, whereas the parasitic stages reduce fumarate. Apparently, the observed change in the activity ratio that was observed between free-living and parasitic stages of A. suum (10) does not occur in F. hepatica and H. contortus.
The low SDH/FRD activity ratios, which we observed in all fumarate-reducing eukaryotes (parasitic helminths as well as the lower marine organisms), are mainly caused by a markedly increased fumarate reduction activity compared with rat and bovine heart mitochondria. This correlates with the significant capacity for fumarate reduction of mitochondria from fumarate-reducing eukaryotes and, in this respect, shows that the fumarate-reducing enzyme complexes of these eukaryotes resemble the FRDs of prokaryotes.
Diode  Tielens et al. (33) demonstrated that the energy metabolism of the developing liver fluke changes gradually from an aerobic energy metabolism, in which Krebs cycle activity occurs, to an anaerobic energy metabolism, in which malate dismutation occurs and fumarate is reduced. From these experiments, it was calculated how many electrons are transported by ubiquinone (UQ) from complexes I and II to complex III in the aerobic respiratory chain as well as how many electrons are transported by RQ from complex I to complex II in the anaerobic electron transport chain. E and Ç, calculated use of Q and RQ during development of the liver fluke, respectively; q and å, the detected amounts of Q and RQ, respectively, in liver flukes isolated from experimentally infected male Wistar rats at certain time points after infection. The means of three independent experiments are shown with standard deviations. onstrated by cyclic voltammetry that bovine SDH, when measured in the direction of fumarate reduction, is severely retarded when the driving force (over-potential) is increased. The authors suggested that bovine SDH functions just like a tunnel diode: preferentially unidirectional (in the direction of succinate oxidation). In contrast to bovine and Escherichia coli SDHs (17), E. coli FRD does not demonstrate this diode-like behavior (39), which correlates with its in vivo function, the reduction of fumarate.
Ackrell et al. (17) demonstrated that these contrasting behaviors could also be observed via the reduced benzyl viologen fumarate reductase assay. Diode-like behavior in this assay results in negative-order kinetics instead of the usual positiveorder kinetics. E. coli, rat, cattle, and human SDHs demonstrated negative-order kinetics and therefore diode-like behavior (17). On the other hand, E. coli and Saccharomyces cerevisiae FRDs demonstrated positive-order kinetics. Hence, enzymes showing a positive order may be identified as FRDs, while those displaying a negative order are SDHs inasmuch as they appear to catalyze the reduction of fumarate only under conditions of moderate driving force (17).
Typical examples of positive-and negative-order kinetics in the reduced benzyl viologen fumarate reductase assay of parasitic helminths are shown in Fig. 3. Diode-like behavior of complex II was demonstrated for mammalian-type mitochondria isolated from rat and bovine heart; for homogenates of the facultative anaerobic marine organisms M. edulis, C. angulata, and A. marina; and for homogenates of the investigated freeliving aerobic stages of parasitic helminths (F. hepatica metacercariae and H. contortus L3). These results suggest that all mitochondria utilizing Krebs cycle activity in vivo contain an SDH-type complex II possessing diode-like behavior. The opposite, however, that enzyme complexes known to reduce fumarate in vivo show positive-order kinetics, was not observed as adult A. suum (17) and F. hepatica (Fig. 3), although both completely dependent on fumarate reductase activity (32,40), showed diode-like behavior (negative order). Adult H. contortus (Fig. 3) and D. viviparus, on the other hand, which are not solely dependent on fumarate reductase activity (9,41), showed positive-order kinetics and not the diode-like behavior. Hence, the absence or presence of the diode-like behavior in crude membrane preparations from these eukaryotic organisms gives no indication as to the importance of fumarate reduction in vivo.
Comparison of Prokaryotic and Eukaryotic Enzyme Complexes That Reduce Fumarate-Recent reports suggested that eukaryotes that reduce fumarate also contain, in addition to an enzyme complex for succinate oxidation, a distinct enzyme complex for fumarate reduction (9,10). It can now be concluded that these enzyme complexes for fumarate reduction in eukaryotes differ significantly from the prokaryotic type of FRD since (i) all known primary structures of eukaryotic enzyme complexes functioning in vivo in the direction of fumarate reduction demonstrate a higher amino acid sequence similarity to mammalian and E. coli SDHs than to prokaryotic FRD (1,9,42,43); and (ii) eukaryotic complexes that reduce fumarate interact with a benzoquinone (rhodoquinone), whereas prokaryotic FRDs interact with a naphthoquinone (menaquinone). Therefore, in addition to the two known types of enzyme complexes in prokaryotes (SDH and FRD), which differ in primary structure, activity ratio of succinate oxidation and fumarate reduction, diode-like behavior, and interacting quinone, another type of complex exists. This type is present (next to the SDH-type complex) in fumarate-reducing eukaryotes and belongs to the FRD category with respect to the functional properties of the complex, like the SDH/FRD activity ratio and the relatively low standard redox potential of the interacting quinone, RQ (EЈ 0 ϭ Ϫ63 mV). On the other hand, this type of enzyme belongs to the SDH category of complex II with respect to its primary structure and the type of interacting quinone, a benzoquinone, as our results demonstrated that in eukaryotes the essential component in fumarate reduction is RQ.
Our results support the idea that fumarate-reducing eukaryotes possess an enzyme complex for the reduction of fumarate, which is structurally related to SDH-type complex II, but has the functional characteristics of prokaryotic FRDs. For definitive conclusions, however, on the differences between enzyme complexes for succinate oxidation and fumarate reduction in eukaryotes, further experiments, like sequence comparisons and analysis of the kinetic properties of the purified enzymes, will be necessary.