1 Introduction

Hydrogenosomes and mitosomes are mitochondria in the evolutionary sense that they descend from one and the same eubacterial endosymbiont. But they differ from the mitochondria that most people know from textbooks in that neither hydrogenosomes nor mitosomes studied to date have been found to possess mechanisms of oxidative phosphorylation . Hydrogenosomes and mitosomes occur among eukaryotes that have oxygen-independent ATP synthesis . Accordingly, they occur among various protists that inhabit anaerobic environments and among various protists that have a parasitic lifestyle. The discovery and study of both organelle types have had substantial impact on our understanding of eukaryotic evolution. The role of mitochondria in eukaryote evolution has come more clearly into focus in recent years. Mitochondria were causal to the unique bioenergetic configuration of eukaryotic cells (Lane and Martin 2010), they contributed the lipid vesicles that formed the seed of the eukaryotic endomembrane system (Gould et al. 2016), they contributed three times as many genes to the eukaryotic lineage than the archaeal host did (Thiergart et al. 2012; Ku et al. 2015), and they were the physiological enabler for the origin of phagocytosis in eukaryotes (Martin et al. 2017). Hydrogenosomes and mitosomes carry the evolutionary heritage of mitochondria in eukaryotic anaerobes .

Hydrogenosomes were discovered in 1973 (Lindmark and Müller 1973) and turned out to be anaerobic forms of mitochondria from eukaryotes that inhabit anoxic environments (van der Giezen 2009). Mitosomes were discovered independently by Tovar et al. (1999) and Mai et al. (1999) in the human intestinal parasite Entamoeba histolytica. Mai et al. (1999) called the organelle a “crypton ,” but the name “mitosome” suggested by Tovar et al. (1999) has stuck. Both organelles are mitochondria in the evolutionary sense. The list of those who also agree on that issue would surely include those who discovered hydrogenosomes (Lindmark and Müller 1973), those who discovered mitosomes (Mai et al. 1999; Tovar et al. 1999), many authors of the first edition of this book, and authors of other publications on the topic (Müller 2007; Lane 2007; Allen et al. 2007; Sapp 2007; Tielens and van Hellemond 2007; Tachezy and Dolezal 2007; Hackstein et al. 2007; Cavalier-Smith 2007; Emelyanov 2007; Barberà et al. 2007; Tovar 2007). The evidence to support that view has been the subject of numerous topical reviews and minireviews (Embley et al. 1997; Biagini et al. 1997; van der Giezen et al. 2005; van der Giezen and Tovar 2005; Hackstein et al. 2006; Embley and Martin 2006; Mentel and Martin 2008; Tachezy 2008; van der Giezen 2009; Shiflett and Johnson 2010; Müller et al. 2012; Zimorski et al. 2014; Makiuchi and Nozaki 2014). The evidence from the genome sequences of organisms that possess hydrogenosomes, Trichomonas (Carlton et al. 2007), or mitosomes—Encephalitozoon (Katinka et al. 2001), Entamoeba (Loftus et al. 2005), and Giardia (Morrison et al. 2007)—is also fully consistent with the view that both organelles share a common ancestry with mitochondria. Hence there is no need to recite here once more the evidence in support of that view that is based in gene sequence comparisons (van der Giezen et al. 2005; Embley and Martin 2006), shared mechanisms and components of protein import (Dolezal et al. 2006), and shared enzymes (Tielens et al. 2002; Müller 2003; Müller et al. 2012).

There have been outspoken critics of the view of common ancestry for mitochondria and hydrogenosomes. Critics tend to call evidence indicating the common ancestry for these organelles into question, to dismiss it as equivocal, or to marginalize its possible evolutionary significance (Margulis et al. 2006; de Duve 2007). Other critics of common origin favor unspecified lateral gene transfer (LGT) mechanisms to explain hydrogenosome enzyme origin (Andersson and Kurland 1999; Andersson et al. 2003, 2007; Barberà et al. 2007; Hug et al. 2010; Hampl et al. 2011; Stairs et al. 2011, 2014, 2015; Leger et al. 2013, 2016). The LGT theory has it that the earliest eukaryotes were strict aerobes (see however Fig. 1) that acquired the ability to survive low O2 by means of interdomain and intradomain LGT of genes for proteins germane to anaerobe physiology. Yet a different camp disputes the endosymbiotic origin of mitochondria and hydrogenosomes (Harish and Kurland 2017). Most current evidence for LGT comes from the study of trees and branch lengths, and at present nobody knows which, what kind of, or how many phylogenetic trees we can really trust (see Dyall et al. 2004 vs. Hrdy et al. 2004 or Leger et al. 2016 vs. Martin 2017a).

Fig. 1
figure 1

Summary of oxygen accumulation of earth history . Data from Fischer et al. (2016), Lyons et al. (2014), Javaux and Lepot (2018), Lenton et al. (2016), and Stolper and Keller (2018). PAL percent of present atmospheric level

Full size image

There are thus two kinds of issues to consider: (1) the evolutionary relationship of hydrogenosomes and mitosomes to mitochondria as membrane-bounded compartments, on which everyone basically agrees and hence will not be the main focus here, and (2) the phylogenetic status and evolutionary significance of eukaryotic anaerobes that possess hydrogenosomes, mitosomes, and anaerobically functioning mitochondria in the geological context of life’s history on Earth. The latter issue is of some interest. We will not deal with the LGT theory for eukaryote anaerobe origin here in any detail because that has been dealt with recently in considerable detail elsewhere. The problems with LGT theories for eukaryote anaerobe origin are numerous and severe. LGT theories for eukaryote anaerobe origin are Lamarckian in nature (Martin 2018), but that is the least of their problems (Martin 2017a). The main problem with LGT theories is that the genes for anaerobic energy metabolism germane to mitochondria and hydrogenosomes (and the cytosol in eukaryotes that have mitosomes) trace to the eukaryote common ancestor, as Martin and Müller (1998) predicted, and that eukaryotes arose and diversified in low oxygen and anaerobic environments (Müller et al. 2012), as shown in Fig. 1. Eukaryotes required genes for anaerobic energy metabolism from the onset of the lineage, not as late lineage-specific lateral acquisitions.

2 Motivation and a Thirteenfold Insight About O2

This chapter serves to further develop the thesis presented in earlier papers (Martin et al. 2003; Martin 2007, 2017a, b, c, 2018; Mentel and Martin 2008; Müller et al. 2012), that (1) the role of oxygen in eukaryote evolution needs to be viewed in light of what geologists are telling us about oxygen in Earth history and that (2) mitochondriate anaerobes are altogether normal eukaryotes with no special kind of evolutionary rank or status. It will be argued that the common ancestry of ATP-synthesizing forms of mitochondria among eukaryotic aerobes and anaerobes and the presence of mitochondria in eukaryotes that synthesize their ATP in the cytosol are most readily interpreted in light of evolutionary specializations of the mitochondrion from a facultatively anaerobic ancestral eukaryotic state. Conversely, the strictly oxygen-dependent forms of the organelle that we know from organisms that live above the soil line on land are also most easily interpreted as evolutionary specializations of the organelle from a more generalized ancestral state—facultatively anaerobic—that existed during a protracted, anaerobic period of Earth ocean history.

Geochemists have been reporting that the oceans had very low oxygen levels and were subject to widespread anoxia until about 580 million years ago (Lyons et al. 2014; Lenton et al. 2016; Reinhard et al. 2016; Fischer et al. 2016; Brocks et al. 2017; van de Velde et al. 2018) with evidence indicating that the rise to roughly our current atmospheric levels of O2 was generated by early land plants only some 400 million years ago (Lenton et al. 2016), reaching present levels only about 450 million years ago (Lenton et al. 2016; Stolper and Keller 2018). The rise of land plants appears to have been essential for the permanent transition to fully oxygenated oceans (Wallace et al. 2017). Thus, the consensus view on oxygen in Earth history that was current in the 1990s, namely, that O2 rose to modern levels 2.4 billion years ago (Kasting 1993), has changed: O2 rose late, about 500 million years ago, aerobic metabolisms evolved late in Earth history, and anaerobes and aerobes have always been around. Eukaryotes have been around for 1.6 billion years; the first billion years of their lineage diversification took place in anoxic or low oxygen environments. This is sketched in Fig. 1.

A main goal of this chapter will be to convey to the readers the new view of oxygen accumulation during Earth history that has emerged over the last years. The newer view of oxygen in Earth history (oxic habitats late) would render anaerobic eukaryotes with mitochondria a natural result of evolution to be expected in many disparate lineages. Under that view, anaerobic eukaryotes with mitochondria are not, in an evolutionary sense, strange, obscure, or otherwise out of the ordinary; hence no special or unusual mechanisms are required to explain their origin. They are normal in every respect and so are their mitochondria.

What is the 13-fold insight about oxygen? It is this. We know that oxidative phosphorylation using O2 as the terminal acceptor generates about 30 ATP per glucose (Rich and Maréchal 2010), which corresponds to a sevenfold higher ATP yield than the anaerobic fermentations typical of hydrogenosomes (4 ATP per glucose) or a sixfold higher ATP yield compared to anaerobic mitochondria of the type found in liver fluke (5 ATP per glucose) (Müller et al. 2012). That sounds like a big energetic benefit conferred by O2. A similar calculation has been made for the ATP yield from amino acid oxidation (Martin et al. 2017) which is relevant because predatory eukaryotes eat mainly cells, and cells are mainly protein. But that is only half the calculation. Less well known is the circumstance that cells living in the presence of O2 require 13 times more energy to synthesize their cell mass and building blocks for replication than cells that live in the complete absence of O2 (Lever et al. 2015). Eukaryotes that use O2 as the terminal acceptor obtain, if we are generous, 10 times more energy per glucose, yes, but they have to spend 13 times more energy per glucose to use it, because they live in an oxic environment . Where is the energetic benefit ? It seems that if we look at the matter openly, there is none. The reason is that O2 is a strong oxidant. It can react with lipids, proteins, nucleic acids, and biosynthetic intermediates in such a way as to oxidize them. Living in an O2 atmosphere is costly because cells require energy input just to synthesize and maintain the reduced organic compounds that are the building blocks of life in an environment that constantly tends to oxidize them.

3 Older Views of How and Why the Mitochondrion Become Established

Current concepts relating to the origin of mitochondria are most readily understood before the backdrop of how those concepts arose. Since about 1967, mainstream reasoning on the rationale behind mitochondrial origins has focused on oxygen, ATP , and improved energy yield from glucose breakdown through oxidative phosphorylation . There are exceptions to that rule (Blackstone 1995), but the exceptions prove the rule, as set forth recently in a contribution (Martin 2017c) to a special issue of J Theoret Biol devoted to the 50th anniversary of Lynn Margulis’s 1967 paper (Sagan 1967) that revived endosymbiotic theory . In order for there to be a biochemical rationale behind mitochondrial origins, there has to be a null hypothesis about the nature of the host, so that one can explicate the possible nature of host-symbiont interactions that lead to a stable symbiosis and hence, in turn, to specify the selective advantages for either partner during the transition from endosymbiont to organelle. The host has always been the weak partner in endosymbiotic theory as it concerns the origin of mitochondria.

Following the rise and fall of endosymbiotic theory in the first half of the twentieth century (Mereschkowsky 1905; Sapp 1994), Margulis (named Sagan in 1967) suggested that the host that acquired the mitochondrial endosymbiont (hereafter just called “the host ”) was an anaerobic, heterotrophic, fermenting, cell wall-lacking (amoeboid, in the broad sense) prokaryote perhaps similar to modern Mycoplasma (Sagan 1967; Margulis 1970). This host corresponded, in terms of cell topology, to the nucleocytoplasmic component of eukaryotes. Margulis opined that the selective advantage that the mitochondrion conferred upon its assumedly fermenting host was improved ATP yield from glucose breakdown by virtue of oxygen respiration and oxidative phosphorylation . Sagan (1967, p. 229) wrote: “The anaerobic breakdown of glucose to pyruvate along the Embden-Meyerhof pathway occurred in the soluble cytoplasm under the direction of the host genome. […] The greater amounts of energy available after the incorporation of the mitochondrion resulted in large cells with amoeboid and cyclotic movement.”

This idea was highly compatible with another view that emerged at the time, namely, that the origin of eukaryotes (and their mitochondria) corresponded temporally and causally to the global rise in atmospheric oxygen levels ~2 billion years ago. For example, Sagan (1967, p. 225) wrote: “The subsequent evolution of aerobic metabolism in prokaryotes to form aerobic bacteria (protoflagella and protomitochondria) presumably occurred during the transition to the oxidizing atmosphere.” A main point of this chapter, to which will return shortly, concerns the last word of that quote, because microbial evolution did not take place in the atmosphere.

That dual assumption of Margulis’s original versions of mitochondrial origins, which one can call the “oxygen/ATP ” argument (Martin 2007), was accepted by many subsequent authors, most notably in the present context by those who contest the common ancestry of mitochondria, hydrogenosomes, and mitosomes. However, most people have always rejected the part of Margulis’s theory suggesting that flagella (and in later formulations, also the nucleus) stem from spirochaete symbionts. Similarly, most people have always rejected the part of de Duve’s (1969, 2007) modification of Margulis’s (Sagan 1967; Margulis et al. 2006) theory in which peroxisomes, instead of spirochaetes, are the additional endosymbiont. Possibly as a consequence of that and other early controversies about endosymbiosis in cell evolution (Cavalier-Smith 1975), a new view emerged about 1980 (van Valen and Maiorana 1980; Doolittle 1980) that placed the oxygen/ATP advantage in the context of a modified form of Margulis’s theory that (1) lacked the spirochaete endosymbiont and that (2) assumed the newly discovered archaebacteria to be relatives of the host lineage, flanked by the assumption that the host was a phagocytotic, anaerobic, fermenting eukaryote (possessing a nucleus and other salient eukaryotic features).

That 1980 view assumed that the prokaryote-to-eukaryote transition occurred via gradualist mechanisms such as point mutation and hence did not involve symbiosis at all (van Valen and Maiorana 1980; Doolittle 1980) and culminated with a cell that possessed a nucleus but lacked mitochondria. This is what Doolittle (1998) has called the “standard model ”. In that view, mitochondria are interpreted as a small tack-on to, and mechanistically unrelated to, the process that made eukaryotic cells nucleated and complex (Cavalier-Smith 2002). In the standard model, mitochondria (and chloroplasts) are descended from endosymbionts, but the nuts-and-bolts of the prokaryote-to-eukaryote transition (the origin of eukaryote-specific traits) was seen as having occurred independently from, and prior to, the origin of mitochondria. The paper by van Valen and Maiorana (1980) expresses that view in clear physiological terms: The host was assumed to be an amoeboid, anaerobic, fermenting cell related to archaebacteria ; the advantage of the mitochondrial endosymbiont was to supply ATP.

The most important sentence of this section probably got lost in the foregoing; hence it is repeated here: The 1960s idea that oxygen and ATP drove the origin of mitochondria was highly compatible with another 1960s view, namely, that the origin of eukaryotes (and their mitochondria) corresponded temporally and causally to the global rise in atmospheric oxygen levels ~2 billion years ago. In that 1960s view, there was no room for anaerobic eukaryotes and especially no room for their organelles. The evolutionary significance of hydrogenosomes and mitosomes is, today, still marginalized by proponents of theories that have one assumption in common: The origin of eukaryotes and their mitochondria was mechanistically related to global oxygen change. But that idea might be wrong. So, it is worthwhile to see if geochemical evidence tells a different story about the global history of oxygen in 2018 than in 1970. That is the subject of the next section. We will see that eukaryotic anaerobes fit much better in modern views of global oxygen change.

4 Anoxic and Sulfidic Oceans up to ~450 Million Years Ago

The older, traditional model of global oxygen history arguably starts off with a paper (Cloud 1968) that appeared at the time when Margulis was reviving endosymbiotic theory. Further developed and supported by much evidence into the 1990s, the model went more or less like this: The early Earth was devoid of O2; O2 in the atmosphere stems from photosynthesis; cyanobacteria produced several millions of billions of tons of O2, much of which served to oxidize Fe(II) in the early oceans; and once that titration process was complete, the planet was oxidized and O2 started accumulating in the atmosphere about 2–2.3 billion years ago as evidenced by the disappearance of particular uranium minerals and the appearance of redbeds (iron oxidized on continents) at that time (Holland and Beukes 1990; Kasting 1993). In that model, the transition from anoxic oceans to oxic oceans occurred in a very narrow window of time (<100 million years ago) and coincided with the appearance of atmospheric O2 (Kasting 1993). Under that view, the oceans were chemically much like today’s (oxic through and through) for about the last 2 billion years. Because the oceans were where life began (Weiss et al. 2016) and eukaryotic evolution was taking place, and because mitochondria were thought to be all about oxygen, anaerobes only fit on the evolutionary map if they lacked mitochondria, and the anaerobic mitochondria of the animals made no evolutionary sense whatsoever, other than as some late and evolutionarily insignificant adaptation to anaerobic habitats. That is the evolutionary position of eukaryotic anaerobes under the traditional view of ocean oxygen history. Thus, eukaryotes were seen as a very ancient lineage; the most primitive lineages branched off early and lack mitochondria; and with the advent of O2, mitochondria and the lineage leading to higher eukaryotes arose (Knoll 1992). This is also the essence of the recent contributions by Margulis et al. (2006) and de Duve (2007), who cast doubt upon the common ancestry of mitochondria and hydrogenosomes, who disregard mitosomes, and who derive eukaryotes via their own versions of endosymbiotic theory that entail hypothetical symbioses (flagella vs. peroxisomes, respectively) preceding that which gave rise to the mitochondrion. The main biological interpretation of those models and other models that have no room for hydrogenosomes, mitosomes, or anaerobes among evolutionarily advanced eukaryotic lineages is as follows: Anaerobic eukaryotes without mitochondria arose early (and are somehow poorly understood or mysterious), aerobic eukaryotes with mitochondria arose late, and the border between early and late coincides with O2 accumulation . First chinks in the old geological model came from unexpectedly late occurrences and reinterpretation of particular deposits called banded iron formations (BIFs) (Holland 1999), but the model still stood.

Over the last years, new models of ocean geochemistry arose. The new models of global oxygen history start off like the old one: The ancient Earth was devoid of O2, and O2 stems from photosynthesis at the beginning of the early Phanerozoic Eon. The increase of oxygen in the Earth’s atmosphere was caused by photosynthesis by cyanobacteria (ca. 3 Ga) and by eukaryotic algae (ca. 1.5 Ga) (Rensing 2018). The oxygen content leading to 100% of present atmospheric level may have originated from the advent of terrestrial photosynthesis and biomass production accompanied by carbon burial (Lenton et al. 2016; Daines et al. 2017). Early land plants, which evolved from charophytic algae (Nishiyama et al. 2018) and came onto land approximately 500 million years ago (Morris et al. 2018; Reski 2018), generated large amounts of oxygen (Rensing 2018) and O2 reached present levels only about 450 million years ago (Lenton et al. 2016; Stolper and Keller 2018). The accumulation of oxygen in the oceans was a much slower process. Oxygenation of the upper ocean to relatively sustainable near-modern conditions happened at about 200 million years ago (Lu et al. 2018). Many lines of evidence indicate that the oxygenation of the oceans came to a completion less than about 600 million years ago (Sahoo et al. 2012; Lyons et al. 2014).

First traces of microbial life are found in rocks that are 3.95–3.8 billion years (Ga) of age (Tashiro et al. 2017, Arndt and Nisbet 2012, Ueno et al. 2006). At that time the early Earth was a highly reducing environment; thus life on earth originated and evolved in anoxic environments (Fischer et al. 2016). Thus, “facultative anaerobes early” fits seamlessly with the newer findings that oceans were oxygen-poor and largely anoxic until about 580 million years ago (Poulton et al. 2004; Lyons et al. 2014; Fischer et al. 2016; Planavsky et al. 2016; Lenton et al. 2016; Reinhard et al. 2016, Laakso and Schrag 2017). In the view of facultative anaerobes early, no convoluted LGT narratives are required to deliver the same genes for anaerobic survival to many different eukaryotic lineages, rather the genes required to underpin anaerobic growth were present in the eukaryote common ancestor, whereby most symbiont genes were transferred to the host’s chromosomes (Martin and Müller 1998; Timmis et al. 2004; Ku et al. 2015), a select few for components of the electron transport chain were retained in organelles (Allen 2015), and the symbiotic origin of mitochondria, not oxygen, was the decisive factor in eukaryote origin (Lane and Martin 2010).

Early eukaryotes arose about 1.65 billion years ago and had diversified by 0.8 billion years ago (Porter et al. 2018). That occurred in anoxic oceans ; hence the widespread occurrence of anaerobic energy metabolism among eukaryotes no longer is a curious puzzle, nor does it require some kind of special explanation. The broad phylogenetic distribution of a uniform and recurrently conserved energy metabolic repertoire among anaerobic eukaryotes no longer has to be explained away as some adaptation to anaerobic niches from an assumed aerobic-specialized ancestral state. In many different lineages, the anaerobic lifestyle in eukaryotes can be readily understood as a direct holdover among diversified descendants of a kingdom’s anaerobic youth. The new models of “oxygen late ” and models for eukaryote origins that accommodate the ubiquity of mitochondriate eukaryotes in anaerobic habitats fit perfectly together.

5 Anoxic Oceans Give Anaerobic Eukaryotes Room to Breathe

Under the new model of Earth’s oxygen, hydrogenosomes, mitosomes, and anaerobic eukaryotes in general fit quite naturally into the overall scheme of Earth and life history, without any kinds of special explanation required to account for anaerobic lifestyles. Many protists today still inhabit environments where virtually no free O2 is available, such as anaerobic sediments (Fenchel and Finlay 1995; Bernhard et al. 2000; Stoeck et al. 2009; Beinart et al. 2018; Porter et al. 2018), and new species are still discovered (Hanousková et al. 2018). In addition, numerous species, in particular among the ciliates, harbor methanogenic archaebacteria that live as endosymbionts within their hosts (Embley and Finlay 1994; Edgcomb et al. 2011), whereby methanogens are among the strictest anaerobes known (Welte and Deppenmeier 2014). Thus, the terms microaerophilic and anaerobic can be equally applicable for many protists. Few eukaryotes studied in detail so far are truly strict anaerobes, in that most of them regularly encounter a bit of O2 in their natural habitats. Accordingly, they have biochemical means for dealing with O2 and can readily tolerate it in small amounts.

Globally widespread oxic (aerobic) marine environments are a very late arrival on the evolutionary scene, almost as late as life on land. In that sense, the ecological specialization to strictly oxic environments is a comparatively late evolutionary development. That statement stands in direct opposition to much traditional argumentation regarding the origin and evolution of mitochondria (de Duve 2007; Margulis et al. 2006; Cavalier-Smith 2004), so let’s contrast it to some readily available observations to see how it fares.

Let’s move back in evolution just one single step more, into the invertebrates. Early eukaryotes appeared ca. 1.65 Ga ago and diversified ca. 800 million years ago (Porter et al. 2018). Vertebrate life on land only goes back about 350 million years (Long and Gordon 2004), when the Earth’s atmosphere was enriched with oxygen. Thus, all large animals living on land are specialized to life with O2 in the atmosphere, specialized to oxic environments , because their mitochondria are specialized to oxic environments. They arose after the oceans were oxic and moved onto land at a time when they had evolved lungs and were able to crawl. Most marine invertebrates (worms, molluscs, and crustaceans) can also use oxygen as terminal electron acceptor in their mitochondria but can also endure sustained anoxia without any problems whatsoever, during which their mitochondria function anaerobically, excreting acetate, propionate, succinate, and CO2 as end products of mitochondrial ATP synthesis, without the participation of oxygen. Their energy metabolism has been summarized in various extensive reviews (de Zwaan 1991; Schöttler and Bennet 1991; Zebe 1991; Grieshaber et al. 1994; Müller et al. 2012). The biochemical steps involved in the synthesis of the main end products of mitochondrial ATP synthesis —succinate, propionate, acetate, and alanine (Schöttler and Bennet 1991; de Zwaan 1991; Grieshaber et al. 1994)—are the same in various free-living marine invertebrates as those found in various parasitic worms that afflict humans and livestock (Tielens 1994; Tielens et al. 2002; van Hellemond et al. 2003; Bringaud et al. 2010; Harder 2016). The steps involve fumarate reductase (FRD) , rhodoquinone instead of ubiquinone at the FRD step (Tielens and van Hellemond 1998; Tielens et al. 2002; Müller et al. 2012), and an acetate/succinate CoA transferase (ASCT) as occurs in hydrogenosomes (Müller 2003). The methylmalonyl-CoA route to propionyl-CoA in anaerobic mitochondrial energy metabolism of invertebrates is nothing special, it is simply a reversal of the pathway used by the mammalian digestive tract to incorporate propionate into glycogen (Schöttler and Bennet 1991), and it is also used by Euglena mitochondria (Schneider and Betz 1985), which contain rhodoquinone as required for the FRD step (Hoffmeister et al. 2004), during anaerobic respiration leading to synthesis of odd chain fatty acids and alcohols during Euglena’s wax ester fermentation (Tucci et al. 2007). Succinate production in anaerobic metazoan mitochondria is not a process of substrate-level phosphorylation; it involves rhodoquinone-dependent proton pumping at the FRD step and chemiosmotic ATP synthesis (Tielens et al. 2002; van Hellemond et al. 2003), and it could thus be described as fumarate respiration, the term commonly used when discussing energy metabolism in prokaryotes (Fenchel and Finlay 1995), although the term malate dismutation is also used (Tielens and van Hellemond 2007). The typical method for cultivating the marine invertebrates to characterize their anaerobic mitochondrial metabolism is to incubate them in N2-saturated water over longer periods of time (Grieshaber et al. 1994). For very small metazoans (less than ~2 mm in size; meiofauna), biochemical data about anaerobic metabolism are scarce at best, but it is known that many of them live in the anoxic layer of marine sediment (Nicholas 1991).

The circumstance that mitochondria of marine invertebrates (metazoans, higher eukaryotes in any sense of the word) can readily do their job of synthesizing ATP for the animal without the help of oxygen is not very well known or well appreciated in the evolutionary community (see Martin 2017b), even though the biochemistry and physiology of invertebrate anaerobic mitochondria have been studied in considerable detail. Grieshaber et al. (1994, p. 64) write: “Meanwhile, details of the anaerobic energy metabolism have been elucidated in all major phyla of invertebrates and even in some insect species (Zebe 1991) which are usually considered as completely oxygen dependent.” The widespread occurrence of anaerobic mitochondria among marine invertebrates might seem like an odd adaptation of some sort if we think that mitochondria arose because of oxygen, but it makes sense if we think that mitochondria arose at a time where only little amounts of oxygen were available and that the low availability of oxygen persisted throughout most of eukaryotic history in the oceans.

If we dig a bit deeper into eukaryotic phylogeny and look at the fungi, then we already arrive at the anaerobic chytridiomycetes, a basal group of true fungi that inhabit anaerobic habitats and have fully fledged hydrogenosomes (Yarlett et al. 1986; Yarlett 1994; Biagini et al. 1994; Boxma et al. 2004). The chytrid hydrogenosomes are anaerobic mitochondria that produce ATP without an electron transport chain (without chemisomosis) and that produce H2. The main end products of ATP synthesis in hydrogenosomes of Piromyces sp. E2 are formate and acetate, with a comparatively small amount of H2; additional end products of energy metabolism are succinate, lactate, formate, and ethanol produced in the cytosol (Boxma et al. 2004). The production of formate is the main difference to trichomonad hydrogenosomes (Mountfort et al. 1994); it entails the activity of pyruvate/formate lyase (PFL) , which is a comparatively rare metabolic enzyme among most eukaryotes studied so far, but the same PFL enzyme occurs in other chytridiomycetes and in the green alga Chlamydomonas reinhardtii (Gelius-Dietrich and Henze 2004; Atteia et al. 2006), suggesting the presence of PFL already in the eukaryote common ancestor. The ethanol produced by Piromyces is generated from acetyl-CoA by the action of bifunctional aldehyde/alcohol dehydrogenase (ADHE), which is found in a variety of anaerobic protists (Boxma et al. 2004), in particular many possessing mitosomes (Müller 2003), and which is also found in Chlamydomonas (Atteia et al. 2006, 2013; Mus et al. 2007) and its colorless chlorophyte relative, Polytomella sp. (Atteia et al. 2003). The presence of ADHE in the eukaryote common ancestor also seems likely (Atteia et al. 2003). The hydrogenase of chytrid hydrogenosomes is an iron-only hydrogenase ([Fe]-Hyd) , the same type as occurs in hydrogenosomes of other anaerobic eukaryotes and also in Chlamydomonas (Horner et al. 2002; Embley et al. 2003). The ATP-synthesizing enzyme of the chytrid hydrogenosome, succinyl-CoA synthase (also called succinate thiokinase ), is an enzyme of the citric acid cycle and is the same one (Brondijk et al. 1996; Dacks et al. 2006) as is found in other hydrogenosomes and in mitochondria that have an electron transport chain (Lahti et al. 1994).

Thus, the small repertoire of a few enzymes involved in anaerobic energy metabolism in chytridiomycete hydrogenosomes and their cytosol reveals nothing unique to the group; it entails the same enzymes as are found in other unicellular eukaryotes that deal with anaerobiosis permanently or on a regular basis and as also are found in eukaryotes that produce oxygen. In light of a predominantly anaerobic ocean past, that should not be surprising in the slightest. The reader might wonder how Chlamydomonas could encounter anaerobiosis, because it is often regarded as an oxygenic producer, but it is a generalist that commonly inhabits soil (Sack et al. 1994; Merchant et al. 2006), that can regularly encounter anaerobic conditions, and that avidly grows heterotrophically on acetate (Heifetz et al. 2000), a common currency of metabolic end product and carbon source among microbes (Wolfe 2005).

The example of the chytrid hydrogenosomes uncovers many links of shared common ancestry for enzymes regarded as “characteristic” for eukaryotic anaerobes , but the same enzymes also occur among oxygen-adapted forms like Chlamydomonas . That makes sense in the context of oxygen late in Earth history. So, if we look around in other eukaryotic groups, do we see anything really special among the anaerobes?

In the trichomonads, the group where hydrogenosomes were discovered, the map of metabolism is slightly different from the chytrids, because the trichomonad hydrogenosomes studied so far possess pyruvate/ferredoxin oxidoreductase (PFO) , an enzyme far more widespread among eukaryotic anaerobes (Müller 2003) than PFL . The end products of energy metabolism in trichomonad hydrogenosomes are equimolar amounts of H2, CO2, and acetate per mol of pyruvate that enters the organelle (Steinbüchel and Müller 1986). The H2 is generated by the same type of typical [Fe]-HYD as found in chytrids and Chlamydomonas (Horner et al. 2002). The same kind of PFO is found in some ciliate hydrogenosomes (Ellis et al. 1994) but also in the mitochondria of the photosynthetic protist Euglena (Rotte et al. 2001) and in Chlamydomonas (Atteia et al. 2006), where it is located in the chloroplast (van Lis et al. 2013).

In anaerobic eukaryotes that perform all of their ATP synthesis in the cytosol , like Entamoeba or Giardia, both of which have mitosomes (Tovar et al. 1999, 2003; Mai et al. 1999), several of the enzymes otherwise found in mitochondria or hydrogenosomes, for example, PFO and [Fe]-HYD (Müller 2003), have been relocalized to the cytosol. Entamoeba or Giardia tend to produce acetate and ethanol, or a mixture of both, as their main metabolic end products (Müller 2003), although Giardia can also produce hydrogen under highly anoxic conditions (Lloyd et al. 2002). The enzymes involved are PFO (Townson et al. 1996), ADHE (Sanchez 1998), and acetyl-CoA synthase (ADP-forming) (Sanchez and Müller 1996; Sanchez et al. 2000), that, like PFL , seems to be a rare enzyme in eukaryote energy metabolism, but the corresponding gene also occurs in the malaria parasite Plasmodium falciparum (Sanchez et al. 2000), which uses the mitochondrial electron transport chain in some stages of the life cycle (Painter et al. 2007). So again, there is no case to be made that the enzyme is genuinely specific to anaerobes. In some highly reduced parasitic (but not anaerobic) eukaryotes that possess mitosomes (Williams et al. 2002), it is not clear that the parasite actually produces its own ATP, so particularly among energy parasites, the presence of highly reduced mitochondria need not correlate with the anaerobic lifestyle.

Although hydrogenosomes and mitosomes are both mitochondrion-derived organelles (Tachezy 2008; van der Giezen 2009; Shiflett and Johnson 2010; Zimorski et al. 2014), there is one important difference between them—mitosomes do not produce ATP. The only function discovered so far that links mitosomes to mitochondria and hydrogenosomes is iron-sulfur cluster assembly (Tovar et al. 2003; Goldberg et al. 2008), which seems to be seen as evolutionarily significant. Iron sulfur cluster assembly is a notoriously O2-sensitive process, so it is not surprising that in oxygen-addicted vertebrates like humans, the process occurs in the mitochondrial matrix, which is generally oxygen-poor, similar to the situation in Bradyrhizobium, where N2 fixation in the cytosol takes place in the presence of active O2 respiration at the surrounding plasma membrane (Preisig et al. 1996).

To sum up this section, if we look around among the eukaryotic anaerobes that possess hydrogenosomes and mitosomes in search of enzymes or other biochemical attributes that set them distinctly apart from eukaryotes with mitochondria specialized to intermittently oxic or to permanently oxic environments , we come up with an empty set, hence the older ideas of that there are some eukaryotic lineages that diverged after a first endosymbiosis entailing spirochaetes (flagella) or some unspecified bacterium (peroxisomes) but before the origin of mitochondria should probably be abandoned; in 50 years no molecular evidence has come forth in support of either idea (Jekely and Arendt 2006; Gabaldon et al. 2006; Martin 2017c).

6 Conclusion

The common ancestry of mitochondria, hydrogenosomes, and mitosomes does not seem to be an issue anymore. But views concerning the origin, age, and polarity of the ability of eukaryotes to thrive in anaerobic habitats remain an issue of some debate. There are some who would possibly prefer the idea that the enzymes of anaerobic energy metabolism in eukaryotes are acquisitions via LGT in response to the adaptation to anaerobic habitats (Andersson and Kurland 1999; Andersson et al. 2003, 2007; Barberà et al. 2007; Hug et al. 2010; Hampl et al. 2011; Stairs et al. 2011, 2014, 2015; Leger et al. 2013, 2016), that the ancestral state of the eukaryotic lifestyle is aerobic (Cavalier-Smith 2004), or that eukaryotic anaerobes do not have mitochondria (Margulis et al. 2006; de Duve 2007). But if we actually look at the anaerobes, they are altogether normal eukaryotes, possessing no special sorts of attributes that would set them apart from the aerobes. If we had all learned the biochemistry of mitochondria in college using the example of marine invertebrates , then we would see mitochondria as organelles that produce ATP under aerobic and anaerobic conditions. But we learned mitochondrial biochemistry with the example of rat liver mitochondria. Rats do not survive without oxygen; marine invertebrates do.

A dozen or so genes for anaerobic fermentations also entered the eukaryotic lineage at mitochondrial origin, such that aerobic respiration, anaerobic respiration, and hydrogen-producing fermentations, in addition to heterotrophy in general, entered the eukaryotic lineage at mitochondrial origin, as a single inheritance from the facultatively anaerobic metabolism of the mitochondrial endosymbiont, followed by ecological specialization and differential loss in independent mitochondrion bearing eukaryotic lineages (Ku et al. 2015; Martin 2017a). While aerobic eukaryotes lost the ability to survive under anoxic conditions, anaerobic eukaryotes kept the anaerobic metabolism for specialization to anaerobic niches.

The anaerobic energy metabolism in eukaryotes carries the rather unmistakable imprint of a single acquisition via endosymbiosis in that it was acquired once, it represents a very narrow sample of preexisting prokaryotic metabolic diversity, and, at least with regard to an important quinone (rhodoquinone), it is traced physiologically to a particular group (facultatively anaerobic α-proteobacteria). This situation is similar to the cyanobacterial origin of plastids , which also entailed the single acquisition of a small sample of preexisting photosynthetic diversity via a plastid endosymbiont (Gould et al. 2008), followed by specialization and differential loss (e.g., of phycobilisomes in some lineages). Beyond the mitochondrion, there is no evidence for the participation of any other endosymbionts in the origin of the primarily heterotrophic eukaryotic groups, involving microbodies (Gabaldon et al. 2006), flagella (Jekely and Arendt 2006), or otherwise (Martin et al. 2001, 2015).

More robust characters, like the presence or absence of genes and enzymes, the presence of organelles, or the intermingling of aerobic and anaerobic lineages, as, for example, among the ciliates (Embley et al. 1995) or higher eukaryotic groups (Embley and Martin 2006; Müller et al. 2012), suggest that there is no phylogenetic distinction at all between eukaryotic aerobes and eukaryotic anaerobes. Similarly, there is no phylogenetic distinction between eukaryotes that possess or lack flagella (missing among many amoebae, all red algae, and most fungi) and no phylogenetic distinction between eukaryotes that possess or lack peroxisomes (typically lacking in anaerobes). That is, however, not to say that there are no phylogenetic distinctions among eukaryotes at all, for there clearly is a phylogenetic distinction between eukaryotes that possess or lack primary plastids, and the same is true for many other characters as well (see Adl et al. 2005 for a compilation). In light of the late rise of oxygen on Earth, eukaryotic anaerobes should hardly be surprising. The brunt of eukaryote evolution appears now to have occurred in anaerobic marine environments (Martin 2017d).

That is the reason that many eukaryotes have preserved enzymes of anaerobic energy metabolism even though they are aerobes. Many O2-producing algae have preserved enzymes of anaerobic energy metabolism (Atteia et al. 2013), for example. A more surprising case is perhaps the amoeba Naegleria gruberi. Its genome sequence revealed the presence of many enzymes involved in energy metabolism of hydrogenosomes, suggesting that it might use them for anaerobic growth (Fritz-Laylin et al. 2010). However, recent work showed that N. gruberi is an obligate aerobe that gains its ATP from O2-dependent lipid oxidation and that shuns both glucose and amino acids and will not grow in the absence of oxygen (Bexkens et al. 2018). Naegleria’s unique O2-dependent specialization toward lipid oxidizing energy metabolism (Bexkens et al. 2018) underscores the message of Müller et al. (2012) that studies of physiology and growth are essential, because investigations of genes alone can be positively misleading, if we are unlucky. Indeed, the lipid- and O2-dependent metabolism of N. gruberi appears to be fully conserved in N. fowleri, the deadly “brain-eating amoeba ,” explaining why it avidly consumes nerve tissue en route to a lipid-rich and the most O2-rich human organ—the human brain, which consumes 20% of our O2 intake but accounts for only 2% of our body weight (Erecińska and Silver 1989). Nonetheless, the retention of genes for anaerobic metabolism in the strict aerobe Naegleria is readily understood as a relic from the anaerobic past of eukaryotes and as an evolutionarily recent ecological specialization to late arising oxic environments.