Hydrogenosomes Organelles that Can Use Protons as Electron Acceptors

In marked contrast to the mitochondrial remnant organelles mentioned above, hydrogenosomes retained an ATP generating function (Müller 1993, 1998). They compartmentalize the terminal reactions of the cellular energy metabolism. Characteristically, hydrogenosomes import pyruvate (or malate) which is oxidatively decarboxylated to acetyl-CoA by the action of a pyru-vate:ferredoxin oxidoreductase (PFO), and not, as in aerobic mitochondria, by the action of a pyruvate dehydrogenase (PDH). An acetate: succinate CoA-transferase and a succinate thiokinase are believed to mediate the formation of acetate and ATP (Müller 1993, 1998), similar to the situation in the mitochondria of the kinetoplastidae and helminths (van Hellemond et al. 1998; Tielens et al. 2002). The reduction equivalents that are formed during the decarboxylation of pyruvate are not used to fuel an electron transport chain as in mitochondria; rather they are removed from the hydrogenosome by a hydrogenase, which reduces protons resulting in the formation of molecular hydrogen (Fig. 2; Müller 1993; Embley and Martin 1998; Martin and Müller 1998; Vignais et al. 2001). Notably, anaerobic chytridiomycete fungi followed a different evolutionary strategy: they avoid the generation of reduction equivalents by using pyruvate: formate lyase (PFL) instead of pyruvate ferredoxin oxidoreductase (PFO) for the non-oxidative splitting of pyruvate into acetate and formate rendering hydrogen production a marginal metabolic route (Fig. 3; Akhmanova et al. 1999; Boxma et al. 2004). The hydrogenosomes of certain ciliates such as Nyctotherus ovalis, on the other hand, retained a much more mitochondrial-type of hydrogenosomal metabolism (see below). Clearly, hydrogenosomes are not the same (Coombs and Hackstein 1995; Embley et al. 1997; Hackstein et al. 1999, 2001). Therefore, it is necessary to review publications dealing with the various hydrogenosomes in more detail, and, in particular, to discuss the metabolic variants with respect to their potential to support the growth of endo- or ectosymbionts, respectively.

Hydrogenosomes of Trichomonas vaginalis

The hydrogenosomes of the trichomonads (Parabasalia) have been studied intensively for more than 30 years (Lindmark and Müller 1973; Müller 1993).

Fdred Fdred

Trichomonas

Fdred Fdred pyruvate pyruvate ^=^>acetyl-CoA acetate >acetate succinate succinyl-CoA

j)PFO

ATP ADP

ATP ADP y (3STK

Fig. 2. Metabolic scheme of a generalised, anaerobic "textbook" protist (like Trichomonas vaginalis) with a hydrogenosome ("type II anaerobe"; Müller 1993, 1998). Pyruvate is formed in the cytoplasm (C) by glycolysis, imported into the hydrogenosome (H) and metabolised to acetate and CO2 under formation of H2. ATP is formed by substrate level phosphorylation by the enzymes acetate succinyl-CoA transferase (ASCT, 2) and succinate thiokinase (STK, 3). ATP is exported by an ADP-ATP carrier (AAC, J). The electrons resulting from the oxidative decarboxylation of pyruvate are transferred to a ferredoxin by pyruvate:ferredoxin oxidoreductase (PFO, 1) and to protons by a Fe-hydrogenase (HYD, 4). N: nucleus. From Hackstein et al. (2001), modified

Upon initial inspection, with the exception of a double membrane, these organelles were considered, both morphologically and biochemically, distinct from mitochondria (Fig. 4; Benchimol et al. 1996; Benchimol and Engelke

2003). Subsequent biochemical and molecular studies have changed this view, albeit sometimes in a way of wishful thinking. Trichomonad hydrogenosomes possess mitochondrial-like chaperonins, HSP 10, HSP 60 and HSP 70 (Clark and Roger 1995; Bui et al. 1996; Germot et al. 1996), proteins of the mitochondrial carrier family (HMP 31; see Dyall et al. 2000, 2004; Tjaden et al.

2004). The N-terminal extensions of many hydrogenosomal proteins favoured the assumption that a mitochondrial-type import machinery would facilitate

Fig. 3. Energy metabolism of the anaerobic chytridiomycete fungus Piromyces sp. E2. The panel shows a scheme of the metabolic pathways involved in the production of the major end products, which are representative for a bacterial-type mixed acids fermentation. The numbers in panel A indicate the following enzymes: (1) hexokinase, glu-cose-6-phosphate isomerase, phosphofructokinase 1, aldolase and triose phosphate isomerase; (2) glyceraldehyde 3-phosphosphate dehydrogenase; (3) phosphoglycerate kinase, phosphoglycerate mutase and enolase; (4) phosphoenolpyruvate car-boxykinase; (5) malate dehydrogenase; (6) fumarase; (7) fumarate reductase; (8) pyruvate kinase; (9) lactate dehydrogenase; (10) cytosolic pyruvate:formate lyase; (11) alcohol dehydrogenase E; (12) pyruvate import into hydrogenosomes; (13) malic enzyme; (14) hydrogenase; (15) hydrogenosomal pyruvate: formate lyase; (16) ace-tate:succinate CoA-transferase; (17) succinyl-CoA synthethase; (18) ADP/ATP carrier. Abbreviations; AcCoA acetyl-CoA; EtOH ethanol; FUM fumarate; G3P glyceralde-hyde-3-phosphate; MAL malate; OXAC oxaloacetate; PEP phosphoenolpyruvate; PYR pyruvate; SUCC succinate. An analysis of the metabolic fluxes revealed that the formation of hydrogen via the malate route can become marginal under certain culture conditions (Boxma et al. 2004). From Boxma et al. (2004), modified.

the import of nuclear-encoded hydrogenosomal proteins into the organelle (Dyall and Johnson 2000; Dyall et al. 2000, 2004; Embley et al. 2003). Also, the presence of acetate:succinate CoA-transferase activity (Müller 1993, 1998; van Hellemond et al. 1998), an enzyme activity that is shared by these organelles

Deep Mind Dreams

Fig. 4. A Trichomonas vaginalis, light microscopical picture of eosin-stained cells; natural size approximately 10 x 45 |m (courtesy of H. Aspöck, Vienna). B Electron micrograph of Tritrichomonas foetus: seven hydrogenosomes (H) can be identified in the cytoplasm (N nucleus; G Golgi apparatus; A axostyl). C A higher magnification reveals that a double membrane surrounds the hydrogenosomes. (M marginal plate). B and C were kindly provided by M. Benchimol, Rio de Janeiro Bar. in B and C 1 |m. From Hackstein et al. (2001), modified.

Fig. 4. A Trichomonas vaginalis, light microscopical picture of eosin-stained cells; natural size approximately 10 x 45 |m (courtesy of H. Aspöck, Vienna). B Electron micrograph of Tritrichomonas foetus: seven hydrogenosomes (H) can be identified in the cytoplasm (N nucleus; G Golgi apparatus; A axostyl). C A higher magnification reveals that a double membrane surrounds the hydrogenosomes. (M marginal plate). B and C were kindly provided by M. Benchimol, Rio de Janeiro Bar. in B and C 1 |m. From Hackstein et al. (2001), modified.

and certain mitochondria, seemed to suggest a "mitochondrial" ancestry for the hydrogenosomes of Trichomonas (Müller 1993, 1998; Dyall and Johnson 2000; Rotte et al. 2000). However, trichomonad hydrogenosomes are clearly different from mitochondria since they lack a genome, ribosomes, cytochromes, an electron transport chain, cardiolipin and cristae (Müller 1993, 1998; Benchimol et al. 1996; Clemens and Johnson 2000; Voncken et al. 2002a; Benchimol and Engelke 2003). Moreover, their import machinery seems to exhibit rather peculiar characteristics that are not shared with mitochondria (Dyall et al. 2003, 2004). Like mitochondria, they import pyruvate, which results from glycolysis, but trichomonad hydrogenosomes do not use a pyruvate dehydrogenase (PDH) for the catabolism of pyruvate. Rather, these hydrogenosomes metabolize pyruvate through pyruvate .ferredoxin oxidoreductase (PFO) and hydrogenase to acetate, carbon dioxide and hydrogen (Fig. 2; Müller 1993, 1998). Acetate formation from acetyl-CoA is believed to be coupled to the substrate level phosphorylation of succinate via the enzyme acetate.succinate CoA transferase (ASCT; Müller 1993); this route should yield 1 ATP per mol of pyruvate consumed. ASCT is one of the few enzymes, which are known to be shared between hydrogenosomes and certain mitochondria. However, it is still unknown whether or not the genes encoding proteins with ASCT activity are the same in trichomonads and ki-netoplastids (cf. Riviere et al. 2004). Additional ATP formation seems to be feasible by the generation of a PMF as in mitochondria (Humphreys et al.

1998). The generation of a PMF has not yet been studied in detail, but the generation of a proton gradient by trichomonad hydrogenosomes is likely (Turner and Lushbaugh 1991). Notably, potential F1F0 ATP synthases of Trichomonas have not been identified so far, and the observation that trichomonad hydrogenosomes can serve as cellular Ca2+-stores (Biagini et al. 1997) is rather circumstantial with respect of a "mitochondrial" ancestry. However, it might be concluded that trichomonad hydrogenosomes are capable of generating a PMF and/or a H+ gradient, although a mitochondrial-like electron-transport chain is absent (Humphreys et al. 1998). Thus, the relationship between the hydrogeno-somes of Trichomonas and aerobically-functioning mitochondria is much less evident than suggested in many publications (see below).

Hydrogenosomes of Anaerobic Ciliates: at Least One Appears to be a Mitochondrion that Produces Hydrogen

Ciliates belong to the "crown group" of eukaryotes (Sogin 1991), and in at least 8 of the 22 orders of ciliates as classified by Corliss (1979), anaerobic species evolved (Fenchel and Finlay 1995). There is a certain agreement that anaerobic ciliates evolved secondarily from aerobic ancestors since some higher ciliate taxa comprise both aerobic and anaerobic species (Embley et al. 1995, 1997, 2003; Fenchel and Finlay 1995; Hackstein et al. 2001, 2002). Bona fide hydrogenosomes are present in 7 of the 22 orders, but both the evidence that these hydrogenosomes evolved independently and repeatedly from mitochondria is rather circumstantial. Notably, Akhmanova et al. (1998a) and van Hoek et al. (2000a) have presented straightforward evidence for the presence of a mitochondrial genome in the hydrogenosomes of Nyctotherus ovalis, an anaerobic, heterotrichous ciliate that inhabits the intestinal tract of cockroaches (Gijzen et al. 1991; Akhmanova et al. 1998a; van Hoek et al. 1998, 1999, 2000b). This genome, identified by immunocytochemistry, hosts rRNA genes that are abundantly expressed, and phylogenetic analysis reveals a clustering among the mitochondrial rRNA genes of aerobic ciliates (van Hoek et al. 2000a; Hackstein et al. 2001). Since the phylogenies of the nuclear 18S rRNA genes of the ciliates are congruent with the SSU rRNA genes of their mitochondria and hydrogenosomes (Akhmanova et al. 1998a; van Hoek et al. 1998, 2000a,b; Hackstein et al. 2001), it is likely that the hydrogenosomes of N. ovalis evolved from the mitochondria of aerobic ciliates. Moreover, the hydrogenosomes of N. ovalis possess cristae, and thus morphologically resemble mitochondria (Fig. 5; Akhmanova et al. 1998a; Hackstein et al. 2001). In addition, there is evidence for quite a number of genes encoding mitochondrial proteins (located on the macronuclear or organelle genome, respectively (Boxma et al. 2005). Therefore, it seems reasonable to assume that the hydrogenosomes of heterotrichous ciliates evolved from mitochondria that adapted to anaerobic environments. There is evidence that the hydrogenosomes of rumen ciliates and plagiopylid ciliates, too, evolved from mitochondria (Embley et al. 1995). However, these hydrogenosomes are quite different from those of N. ovalis and it is likely that they evolved independently (see below).

Akhmanova et al. (1998a) have shown that the ciliate's hydrogenosomes possess an [Fe]-hydrogenase that is encoded by a macronuclear gene-sized chromosome. This hydrogenase represents a novel type of [Fe]-hydrogenase that allows H2-formation to be coupled directly to the reoxidation of NADH. The [Fe]-hydrogenase has been linked covalently with a protein, which possesses NAD and FMN binding sites, and a ferredoxin-like module that allows transferring electrons to the catalytic site of the hydrogenase (Akhmanova et al. 1998a; Vignais et al. 2001; Voncken et al. 2002b). The origin of the hydro-genase(s) is of central importance not only for the hydrogen hypothesis, but also for an understanding of the evolution of the various types of hydrogeno-somes. All eukaryotic [Fe] hydrogenases, including the hydrogenase-like proteins ("NARFs", cf. Balk et al. 2004), which neither produce nor consume hydrogen, seem to form a monophyletic cluster, with the possible exclusion of the N. ovalis hydrogenase (Akhmanova et al. 1998a; Horner et al. 2000, 2002; Nixon et al. 2003). However, due to the high conservation of the hydrogenases

Fig. 5. The hydrogenosome (H) of Nyctotherus ovalis at higher magnification looks like a mitochondrium (glutaraldehyde/OsO4 fixation). The inner and outer membrane, crista-like invaginations of the inner membrane (arrowheads), and putative 70S ri-bosomes can be identified (arrows). m methanogenic endosymbiont. Bar 1 |m. (From Akhmanova et al. 1998a, modified)

Fig. 5. The hydrogenosome (H) of Nyctotherus ovalis at higher magnification looks like a mitochondrium (glutaraldehyde/OsO4 fixation). The inner and outer membrane, crista-like invaginations of the inner membrane (arrowheads), and putative 70S ri-bosomes can be identified (arrows). m methanogenic endosymbiont. Bar 1 |m. (From Akhmanova et al. 1998a, modified)

and the sensitivity for species sampling, statistical support for either of these phylogenies is poor (Voncken et al. 2002b; Embley et al. 2003). Even the identification of the first a proteobacterial hydrogenase could not solve the phylogenetic puzzle (Davidson et al. 2002; Embley et al. 2003). The recent isolation of the genes encoding hydrogenases of rumen ciliates (Boxma 2004) supports a common eukaryotic origin of all eukaryotic hydrogenases and NARFs - with the exception of the hydrogenase of N. ovalis, which appears to be a mosaic of proteins of 8 proteobacterial and P proteobacterial origins (Akhmanova et al. 1998a; Horner et al. 2000, 2002; Voncken et al. 2002b). The [Fe]-hydrogenases of rumen ciliates and anaerobic chytrids clearly belong to the eukaryotic cluster. They are similar to the "long-type" [Fe] hydrogenases from Trichomonas. Interestingly, phylogenetic analysis of the hydrogenases of anaerobic chytrids clusters them with the extremely short hydrogenases of green algae, which function in the plastidic (and not in the mitochondrial) electron transport (Florin et al. 2001; Horner et al. 2002; Voncken et al. 2002b; Nixon et al. 2003). However, the origin of these eu-karyotic hydrogenases from a hypothetical hydrogenase-containing universal endosymbiont remains unclear; in particular, there is no statistical support for the assumption that all eukaryotic hydrogenases (including the NARFs) evolved from an a-proteobacterial ancestor (Horner et al. 2002; Voncken et al. 2002b; Embley et al. 2003; Stejskal et al. 2003).

Hydrogenosomes of Anaerobic Chytrids: an Alternative Way to Adapt to Anaerobic Environments

Anaerobic chytrids are important symbionts in the gastro-intestinal tract of many herbivorous mammals. Their life cycle consists of an alternating flagellated zoospore stage and a vegetative phase when a multi-nucleated mycelium is formed. The hyphae of the rhizomycelial system attach to the digesta and secrete a broad spectrum of fibrolytic enzymes that is very efficient in digesting plant polymers (Teunissen et al. 1991; Orpin 1994; Yarlett 1994). These organisms are highly adapted to intestinal environments; their optimal growth temperature coincides with the body temperature of their mammalian hosts, and during almost their whole life cycle, they live and multiply under anoxic conditions (Orpin 1994). The anaerobic chytrids evolved from mitochondria-bearing ancestors, since DNA sequence analysis reveals a clustering of both aerobic and anaerobic chytrids (Bowman et al. 1992; cf. Paquin et al. 1995; Paquin and Lang 1996). Also an analysis of biochemical and morphological traits consistently establishes a close relationship between chytrids and other fungi (Ragan and Chapman 1978), and Akhmanova et al. (1998b) demonstrated that several enzymes of mitochondrial origin, which lack putative targeting signals, were retargeted to the cytoplasm (in active form) being no longer present in the hydrogenosomes. Consequently, there is little doubt that the chytrids living in the gastro-intestinal tract of herbivorous mammals have secondarily adopted an anaerobic life style (Hackstein et al. 1999).

Anaerobic chytrids such as, for example, Neocallimastix and Piromyces possess hydrogenosomes, which, however, are structurally and functionally clearly different from the hydrogenosomes of the ciliate N. ovalis, the amoe-boflagellate Psalteriomonas lanterna and the parabasalid T. vaginalis (Fig. 6; Coombs and Hackstein 1995; Hackstein et al. 1999, 2001). Like the hydrogenosomes of the amoeboflagellate P. lanterna (Hackstein, unpubl.), and of the parabasalid T. vaginalis (Clemens and Johnson 2000) the hydrogenosomes of Neocallimastix and Piromyces lack a genome (van der Giezen et al. 1997; Hackstein, unpubl.). But unlike T. vaginalis hydrogenosomes, the chytrid hydrogenosomes rely on malate and not pyruvate for hydrogen formation. The imported malate is oxidatively decarboxylated by a hydrogenosomal malic enzyme, and it had been postulated that the resulting pyruvate is oxidized further by pyruvate:ferredoxin oxidoreductase (PFO) to acetyl-CoA. The reduced equivalents should be transferred via ferredoxin to hydrogenase thus maintaining the redox balance (Marvin-Sikkema et al. 1992, 1993, 1994). However, Akhmanova et al. (1999) and Boxma et al. (2004) showed that the hydrogenosomes of anaerobic chytrids perform a bacterial-type mixed acid fermentation during which pyruvate is split into acetyl-CoA and formate by a pyru-vate:formate lyase (PFL), and not oxidatively decarboxylated by a PFO. Consequently, the formation of reducing equivalents is avoided, and the hydro-genosome excretes formate and acetate as end products of its energy metabolism (Fig. 3). Moreover, the vast majority of the carbon flow through the hydrogenosome is mediated by pyruvate, which is imported from the cyto-sol and metabolized in the hydrogenosome without hydrogen formation (Boxma et al. 2004). Obviously, the hydrogenosomes of anaerobic chytrids followed a different strategy when adapting to anaerobic environments: avoiding the formation of reduced equivalents renders hydrogen production a rudimentary metabolic activity in these organelles.

Functional and phylogenetic analysis of the ADP/ATP carriers from anaerobic chytrid hydrogenosomes clearly supports a fungal mitochondrial origin for these organelles (van der Giezen et al. 2002; Voncken et al. 2002a; Tjaden et al. 2004). Given that chytrid hydrogenosomes lack a genome, the ADP/ATP carriers and HSP 60 are the "second-best" markers for tracing the evolutionary history of these organelles (Andersson and Kurland 1999). Phylogenetic analysis of both genes unequivocally reveal a fungal mitochondrial ancestry (Hackstein et al. 1999; Voncken 2001; van der Giezen et al. 2002, 2003; Voncken et al. 2002a), in agreement with the earlier finding that typical mitochondrial enzymes had been retargeted to the cytoplasm in the course of the evolution of the chytrid hydrogenosomes (Akhmanova et al. 1998b).

Fig. 6. A Electron micrograph of a hydrogenosome of the anaerobic chytrid Neocallimastix sp. L2, isolated from the faeces of a llama. This type of hydrogenosomes has a morphology very different from that of Trichomonas sp. or Nyctotherus ovalis. Arrows in A indicate the membranes (m) and vesicles. From Hackstein et al. 2001, modified. B Section through a hydrogenosome of Neocallimastix subjected to (hypotonic) osmotic treatment. The "peas in a pod" organisation of the hydrogenosome becomes visible. Bar 0.5 |m. C An artist's view of the organelle shown in B. From Voncken et al. (2002a), modified

Fig. 6. A Electron micrograph of a hydrogenosome of the anaerobic chytrid Neocallimastix sp. L2, isolated from the faeces of a llama. This type of hydrogenosomes has a morphology very different from that of Trichomonas sp. or Nyctotherus ovalis. Arrows in A indicate the membranes (m) and vesicles. From Hackstein et al. 2001, modified. B Section through a hydrogenosome of Neocallimastix subjected to (hypotonic) osmotic treatment. The "peas in a pod" organisation of the hydrogenosome becomes visible. Bar 0.5 |m. C An artist's view of the organelle shown in B. From Voncken et al. (2002a), modified

Chytrid hydrogenosomes are therefore clearly distinct from the hydrogenosomes of Trichomonas that formed the basis for the Martin-Müller 'hydrogen hypothesis' for the evolution of the eukaryotic cell. Both the origin of the organelle i.e., the universal endosymbiont in the case of Trichomonas and a differentiated fungal mitochondrion in the case of the anaerobic chytrids, and the evolutionary strategies to adapt to anoxic environments are different. Since also the hydrogenosomes of Trichomonas lack a genome (Clemens and

Johnson 2000), an analysis of the hydrogenosomal ADP/ATP carriers should provide clues for or against the hydrogen hypothesis. Surprisingly, trichomonad hydrogenosomes did not host mitochondrial type ADP/ATP carriers (Tjaden et al. 2004). Rather they use a different member of the mitochondrial carrier family, the hydrogenosomal membrane protein (HMP 31) for ADP/ATP exchange that is functionally and phylogenetically distinct from the mitochon-drial-type ADP/ATP carriers (Dyall et al. 2000; Tjaden et al. 2004). The gene encoding HMP 31 branches earlier than the mitochondrial-type ADP/ATP carriers, in agreement of the predictions derived from the hydrogen-hypothesis, and the deviating hydrogenosomal import machinery discussed by Dyall et al. (2003, 2004). Notably, also HSP 60 phylogenies cluster Trichomonas with HSP(cpn) 60 genes of "amitochondriate" (but mitochondrial-remnant-bearing) taxa (diplomonads, entamoebids) rather than with "true" mitochondrial chap-eronines (Voncken 2001; Voncken et al. 2002a; van der Giezen et al. 2003).

+1 0

Post a comment