The Origin of Plastids

Plastids are core components of photosynthesis in plants and algae. Scientists are currently

debating the events leading to the appearance of plastids in eukaryotic cells.

Organelles, called plastids, are the main sites of photosynthesis in eukaryotic cells.

Chloroplasts, as well as any other pigment containing cytoplasmic organelles that enables

the harvesting and conversion of light and carbon dioxide into food and energy, are

plastids. Found mainly in eukaryotic cells, plastids can be grouped into two distinctive

types depending on their membrane structure: primary plastids and secondary plastids.

Primary plastids are found in most algae and plants, and secondary, more-complex plastids

are typically found in plankton, such as diatoms and dinoflagellates. Exploring the origin of

plastids is an exciting field of research because it enhances our understanding of the basis

of photosynthesis in green plants, our primary food source on planet Earth.

Primary Plastids and Endosymbiosis

Where did plastids originate? Their origin is explained by endosymbiosis, the act of a

unicellular heterotrophic protist engulfing a free-living photosynthetic cyanobacterium and

retaining it, instead of digesting it in the food vacuole (Margulis 1970; McFadden 2001;

Kutschera & Niklas 2005). The captured cell (the endosymbiont) was then reduced to a

functional organelle bound by two membranes, and was transmitted vertically to

subsequent generations. This unlikely set of events established the ancestral lineages of the

eukaryote supergroup "Plantae" (Cavalier-Smith 1998; Rodriguez-Expeleta et al. 2005;

Weber, Linka, & Bhattacharya 2006), which includes many photosynthetic algae and land

plants.

The idea of endosymbiosis was first proposed by Konstantin Mereschkowski, a prominent

Russian biologist, in 1905. He coined the term "symbiogenesis" when he observed the

symbiotic relationship between fungi and algae (Mereschkowski 1905). The term

"endosymbiosis" has a Greek origin (endo, meaning "within"; syn, meaning "with"; and

biosis, meaning "living"), and it refers to the phenomenon of an organism living within

another organism. In 1923, American biologist Ivan Wallin expanded on this theory when

he explained the origin of mitochondria in eukaryotes (Wallin 1923). However, not until

the 1960s did Lynn Margulis, as a young faculty member at Boston University, substantiate

the endosymbiotic hypothesis. Based on cytological, biochemical, and paleontological

evidence, she proposed that endosymbiosis was the means by which mitochondria and

plastids originated in eukaryotes (Sagan 1967; Margulis 1970). In those days, the research

community viewed her unconventional idea with much skepticism, but her work was

eventually published in 1967 (Sagan 1967) after being rejected by fifteen scientific

journals! Today, endosymbiosis is a widely accepted hypothesis to explain the origin of

intracellular organelles.

Besides these original and bold ideas, what else have we learned? Since 1990 we have seen

rapid advancement in techniques in molecular biology and bioinformatics. Using molecular

phylogenetic approaches, numerous comparative studies have demonstrated the

cyanobacterial origin of genes encoded in the Plantae plastid and provide evidence for gene

transfer from the endosymbiont genome to the "host" nucleus (Bhattacharya & Medlin

1995; Delwiche 1999; Moreira, Le Guyader, & Phillippe 2000; McFadden 2001; Palmer

2003; Bhattacharya, Yoon, & Hackett 2004; Rodriguez-Ezpeleta et al. 2005; Reyes-Prieto,

Weber, & Bhattacharya 2007). These studies complement several independent lines of

evidence based on protein transport and the biochemistry of plastids (McFadden 2001;

Matsuzaki 2004; Weber, Linka, & Bhattacharya 2006; Reyes-Prieto & Bhattacharya 2007).

The establishment of primary plastids in eukaryotes is estimated to have occurred 1.5

billion years ago (Hedges 2004; Yoon et al. 2004; Blair, Shah, & Hedges 2005), but dating

such an ancient event based on molecular data remains controversial due to the limited

support provided by the fossil records (Douzer et al. 2004).

Whereas endosymbiosis involving a cyanobacterium explains the establishment of primary

plastids in Plantae, the story is more convoluted in other photosynthetic eukaryotes, which

harbor secondary plastids with more complex structures. The plastids found in Paulinella

chromatophora (a filose amoeba) are an exception to the rule. These organisms are derived

from a far more recent cyanobacterial primary endosymbiosis that occurred about 60

million years ago (Bhattacharya, Helmchen, & Melkonian 1995; Marin, Nowack, &

Meklonian 2005; Yoon et al. 2006). This plastid traces its origin to a cyanobacterial donor

of the Prochlorococcus-Synechococcus type (Yoon et al. 2006). The closely related

Paulinella ovalis, although lacking a plastid, is an active predator of cyanobacteria that are

commonly localized within food vacuoles (Johnson, Hargraves, & Sieburth 2005).

Therefore, the cyanobacterium-derived plastid in the photosynthetic P. chromatophora

provides an independent example of the phagotrophic origin of a primary plastid.

Secondary Plastids and the Current Scientific Debate

Figure 1: The concept of endosymbiosis

The chromalveolate hypothesis can explain some endosymbiotic events in dinoflagellates.

In comparison to Plantae and P. chromatophora, the origin of plastids in other

photosynthetic eukaryotes is more complicated. These organisms possess secondary

plastids, which have one or two additional membranes surrounding the existing two

membranes of primary plastids. However, an endosymbiosis event involving a

cyanobacterium cannot explain the origin of three — and four — membrane-bound

plastids. Instead, these extra membranes likely formed due to secondary endosymbiosis,

when an existing Plantae cell containing a primary plastid was engulfed and reduced to a

plastid. There is, however, disagreement among scientists about the number of secondary

endosymbioses in eukaryotes. In general, there are opposing viewpoints regarding their

establishment, namely, the chromalveolate hypothesis (Cavalier-Smith 1999, 2003; Keeling

2009) and the independent acquisition hypothesis (also known as serial eukaryote-

eukaryote endosymbiosis, or the serial EEE) (Archibald 2009; Bodyl, Stiller, &

Mackiewicz 2009; Baurain et al. 2010) (Figure 1).

What is the chromalveolate hypothesis? This controversial idea proposes that a free-living

red algal cell was captured and retained by a nonphotosynthetic heterotrophic protist soon

after the evolutionary split of red and green algae (i.e., secondary endosymbiosis), giving

rise to the pigmented ancestor of the supergroup "Chromalveolata" (Cavalier-Smith 1999).

The red algal endosymbiont was retained in a variety of chromalveolate lineages such as

cryptophytes, haptophytes, stramenopiles, and dinoflagellates. This hypothesis is based on

numerous phylogenetic studies that support the red algal origin of the plastid in most

chromalveolates and of most plastid-localized proteins that are encoded in the nucleus of

these taxa (Fast et al. 2001; Harper & Keeling 2003; Bhattacharya, Yoon, & Hackett 2004;

Li et al. 2006; Nosenko et al. 2006). Although independent horizontal gene transfer events

could explain this observation, the more reasonable explanation is that a secondary

endosymbiosis event in red alga involved multiple gene transfers.

What other changes have occurred in organisms containing secondary plastids? With the

exception of the cryptophytes, which retain a remnant of the red algal cell (Greenwood

1974), other chromalveolates do not have the nucleus of the red algal endosymbiont, which

means that the genes necessary to ensure a fully functional plastid have been transferred to

the host nucleus (Douglas & Penny 2001; Douglas et al. 2001). In addition, for a subgroup

of dinoflagellates that contain the peridinin pigment, including species that cause "red tide"

in the ocean (e.g., Alexandrium), the plastid genomes are highly reduced in size due to

substantial gene transfer from the endosymbiont to the host nucleus (Hackett 2004).

To further complicate matters, besides tracing their origin not only to red algae as expected,

plastid-targeted proteins in many dinoflagellates traced also include proteins derived from

other unicellular eukaryotes such as excavates and other chromalveolates (Ishida & Green

2002; Hackett 2004; Yoon et al. 2005; Nosenko 2006), suggesting additional (tertiary)

endosymbioses in their evolutionary histories. The recent discovery of a substantial number

of red and green algal-derived genes in diatoms (Moustafa et al. 2009) supports the idea of

an additional secondary endosymbiotic event with green alga during the early evolution of

chromalveolates.

If the chromalveolate hypothesis is true, then why are not all chromalveolates (ciliates and

apicomplexans) photosynthetic? Under the assumption of the chromalveolate hypothesis,

the lack of plastids in ciliates is explained by subsequent loss of the captured algal cell or

by genes of the endosymbiont that were acquired during serial endosymbiosis (Reyes-

Prieto, Moustafa, & Bhattacharya 2008). Interestingly, the parasitic Apicomplexans (e.g.,

Toxoplasma), although nonphotosynthetic, possess unique organelles called the apicoplast

(also known as the nonphotosynthetic "plastids") that share similar features with secondary

plastids (Maréchal & Cesbron-Delauw 2001; Ralph et al. 2004). The apicoplasts may have

shared common origins with secondary plastids of the closely related dinoflagellates, but

subsequently lost their photosynthetic capability, likely due to the transition to obligate

parasitism (Funes et al. 2002; Waller & McFadden 2005).

Alternative Explanations for the Origin of Secondary

Plastids

Is there an alternative explanation for the origin of secondary plastids? Indeed, the

independent acquisition hypothesis suggests that the origin of secondary plastids in

different groups of chromalveolates (e.g., haptophytes, cryptophytes, and stramenopiles)

results from independent, serial endosymbioses involving unicellular eukaryotes, not

necessarily red algae (Archibald 2009; Bodyl, Stiller, & Mackiewicz 2009; Baurain et al.

2010). Experimental evidence for this hypothesis is not as strong as that for the

chromalveolate hypothesis, however. In fact, many recent analyses are tailored to disprove

the chromalveolate hypothesis using selective data sampling (Baurain et al. 2010).

What is the advantage of the independent acquisition hypothesis? We know that over

evolutionary time plastid-lacking chromalveolates are likely to lose the plastid and nuclear-

encoded plastid targeted proteins derived from the red algal endosymbiont. In evolutionary

biology, we assume that the least complex (most parsimonious) explanation for an

observation is more plausible. The independent acquisition hypothesis avoids a convoluted

explanation of organelle and gene losses to explain the sporadic plastid distribution

observed today in nonphotosynthetic chromalveolates. In the extreme case, some scientists

who support this hypothesis claim that the fundamental grouping of the chromalveolates is

itself inaccurate. Given the paucity of empirical data, we need to understand this complex

chain of plastid acquisition events much better.

What about the origin of plastids in other eukaryotes? There are other photosynthetic

eukaryotes that are not members of Plantae or Chromalveolata, such as the photosynthetic

excavates (e.g., Euglena) and the chlorarachniophyte amoebae (Rhizaria), but the plastid

origins of these organisms are less well-studied. Nonetheless a number of phylogenetic

analyses show that these organisms have acquired their plastids in more recent instances of

secondary endosymbiosis, during which a green alga was independently captured by their

common ancestors (Rogers et al. 2007).

Summary

There is no simple way to explain the gain and loss of plastids in all eukaryotes. The origin

of primary plastids via endosymbiosis involving a cyanobacterium is well-established, but

the origin of secondary plastids is still controversial. However, the chromalveolate

hypothesis (and the secondary endosymbiosis involving a red alga) is the best-supported

hypothesis to date based on numerous empirical studies. In addition, subsequent tertiary

endosymbioses involving other free-living eukaryotes explain plastid origins in other

eukaryote lineages. Additional studies and biochemical validation (where possible) are

needed to better test existing hypotheses about the evolutionary origins of plastids in

eukaryotes.

Mitochondria arose through a fateful endosymbiosis more than 1.45 billion years ago.

Many mitochondria make ATP without the help of oxygen.

What variety is there in mitochondria? Mitochondria occur in various forms across various

eukaryotic groups, yet considerations on the origin of mitochondria sometimes neglect this

understanding. Four main mitochondrial types can be distinguished on the basis of

functional criteria concerning how or whether ATP is produced. These functional types do

not correspond to natural groups, because they occur in an interleaved manner across the

tree of eukaryotic life. Instead they correspond to ecological specializations.

Mitochondria: A Ubiquitous and Diverse Family of

Organelles

The mitochondria typical of mammalian cells respire O2 during the process of pyruvate

breakdown and ATP synthesis, generating water and carbon dioxide as end products. The

Krebs cycle and the electron transport chain in the inner mitochondrial membrane enable

the cell to generate about 36 moles (mol) of ATP per mole of glucose, with the help of O2–

respiring mitochondria. Such typical mitochondria also occur in plants and various groups

of unicellular eukaryotes (protists) that, like mammals, are dependent on oxygen and

specialized to life in oxic environments.

In contrast, the mitochondria of many invertebrates (worms like Fasciola hepatica and

mollusks like Mytilus edulis being well–studied cases) do not use O2 as the terminal

acceptor during prolonged phases of the life cycle. These mitochondria allow the

anaerobically growing cell to glean about 5 mol of ATP per mole of glucose, as opposed to

about 36 with O2. The typical excreted end products are carbon dioxide, acetate, propionate,

and succinate, which are generated mostly through the rearrangement of Krebs cycle

reactions and the help of the mitochondrial electron transport chain. These organelles are

commonly called anaerobic mitochondria.

Figure 1: Enzymes and pathways found in various manifestations of mitochondria

Proteins sharing more sequence similarity to eubacterial than to archaebacterial

homologues are shaded blue; those with converse similarity pattern are shaded red; those

whose presence is based only on biochemical evidence are shaded grey; those lacking

clearly homologous counterparts in prokaryotes are shaded green. (A) Schematic summary

of salient biochemical functions in mitochondria, including some anaerobic forms.

(B)Schematic summary of salient biochemical functions in hydrogenosomes. (C) Schematic

summary of available findings for mitosomes and Eukaryotic evolution, changes and

challenges. Nature 440, 623–630 (2006)

Mitochondria of yet another kind yield even less ATP per molecule of glucose. These are

mitochondria of several distantly related unicellular eukaryotes (protists) that lack an

electron transport chain altogether. They synthesize ATP from pyruvate breakdown via

simple fermentations that typically involve the production of molecular hydrogen as a

major metabolic end product. These mitochondria are called hydrogenosomes and allow the

cell to gain about 4 mol of ATP per mole of glucose. Hydrogenosomes were discovered in

1973 in trichomonads, a group of unicellular eukaryotes. They were later found in

chytridiomycete fungi that inhabit the rumen of cattle, as well as some ciliates, and they

continue to be found in other groups. The enzymes of hydrogenosomes are not unique to

these anaerobes. They are found also in the mitochondria, the cytosol, or even the plastids

of other eukaryotes (Figure 1).

A fourth category of eukaryotes possesses small, inconspicuous mitochondria that are not

involved in ATP synthesis at all. These eukaryotes synthesize their ATP in the cytosol with

the help of enzymes that are otherwise typically found in hydrogenosomes. They obtain 2-4

mol of ATP per mole of glucose. Their typical end products are carbon dioxide, acetate,

and ethanol, and their mitochondria are called mitosomes. Mitosomes were discovered in

the human intestinal parasite Entamoeba histolytica in 1999, and were subsequently found

in many additional eukaryotes, including Giardia lamblia in 2003.

Knowledge about these different forms of mitochondria comes from decades of

biochemical and physiological investigations of eukaryotic anaerobes, many of which are

important pathogens or parasites of humans and livestock. Well into the 1990s it was

widely thought that several anaerobic eukaryotes, such as Giardia lamblia, lack

mitochondria altogether and had never possessed them in the evolutionary past. Newer

work, however, has shown that mitochondria are just as defining and ubiquitous among

eukaryotes as is the nucleus itself. That realization has had considerable impact on current

views about the origin of mitochondria.

The Endosymbiotic Origin of Mitochondria

There are currently two main, competing theories about the origin of mitochondria. They

differ with regard to their assumptions concerning the nature of the host, the physiological

capabilities of the mitochondrial endosymbiont, and the kinds of ecological interactions that

led to physical association of the two partners at the onset of symbiosis.

Figure 2: Models for eukaryote origins that are, in principle, testable with genome data

(A-D) Models that propose the origin of a nucleus-bearing but amitochondriate cell first,

followed by the acquisition of mitochondria in a eukaryotic host. (E-G) Models that

propose the origin of mitochondria in a prokaryotic host, followed by the acquisition of

eukaryotic-specific features. The relevant microbial players in each model are labelled.

Archaebacterial and eubacterial lipid membranes are indicated in red and blue, respectively.

The traditional view posits that the host that acquired the mitochondrion was an anaerobic

nucleus-bearing cell, a full-fledged eukaryote that was able to engulf the mitochondrion

actively via phagocytosis (Figure 2). This view is linked to the ideas that the mitochondrial

endosymbiont was an obligate aerobe, perhaps similar in physiology and lifestyle to

modern Rickettsia species; and that the initial benefit of the symbiosis might have been the

endosymbiont's ability to detoxify oxygen for the anaerobe host. Because this theory

presumes the host to have been a eukaryote already, it does not directly account for the

ubiquity of mitochondria. That is, it entails a corollary assumption (an add–on to the theory

that brings it into agreement with available observations) that all descendants of the initial

host lineage, except the one that acquired mitochondria, went extinct. The oxygen

detoxification aspect is problematic, because the forms of oxygen that are toxic to

anaerobes are reactive oxygen species (ROS) like the superoxide radical, O2-. In eukaryotes,

ROS are produced in mitochondria because of the interaction of O2 with the mitochondrial

electron transport chain. In that sense, mitochondria do not solve the ROS problem but

rather create it; hence, protection from O2 is an unlikely symbiotic benefit. This traditional

view also does not directly account for anaerobic mitochondria or hydrogenosomes, and

additional corollaries must be tacked on to explain why anaerobically functioning

mitochondria are found in so many different lineages and how they arose from oxygen-

dependent forebears.

An alternative theory posits that the host that acquired the mitochondrion was a prokaryote,

an archaebacterium outright. This view is linked to the idea that the ancestral

mitochondrion was a metabolically versatile, facultative anaerobe (able to live with or

without oxygen), perhaps similar in physiology and lifestyle to modern Rhodobacteriales.

The initial benefit of the symbiosis could have been the production of H2 by the

endosymbiont as a source of energy and electrons for the archaebacterial host, which is

posited to have been H2 dependent. This kind of physiological interaction (H2 transfer or

anaerobic syntrophy) is commonly observed in modern microbial communities. The

mechanism by which the endosymbiont came to reside within the host is unspecified in this

view, but in some known examples in nature prokaryotes live as endosymbionts within

other prokaryotes. In this view, various aerobic and anaerobic forms of mitochondria are

seen as independent, lineage-specific ecological specializations, all stemming from a

facultatively anaerobic ancestral state. Because it posits that eukaryotes evolved from the

mitochondrial endosymbiosis in a prokaryotic host, this theory directly accounts for the

ubiquity of mitochondria among all eukaryotic lineages.

Eukaryotes are genetic chimeras. They possess genes that they inherited vertically from

their archaebacterially related host. Genes for cytosolic ribosomes in eukaryotes, for

example, reflect that origin. But eukaryotes also possess genes that they inherited vertically

from the endosymbiont - for example, mitochondrially encoded genes for mitochondrial

ribosomes. But even the largest mitochondrial genomes possess only about sixty protein-

coding genes, while typical mitochondria harbor up to a thousand proteins or more that are

encoded in the nucleus. During the course of mitochondrial genesis, many genes were

transferred from the genome of the mitochondrial endosymbiont to the genome of the host.

This kind of endosymbiotic gene transfer is nothing unusual; endosymbiosis very often

entails gene transfers from the endosymbiont to the host. It happened during the origin of

plastids too, and it is still ongoing in our own genome: Mitochondrial DNA constantly

escapes from the organelle and becomes integrated as copies into nuclear DNA. The vast

majority of mitochondrial proteins are encoded by nuclear genes, and many of these are

endosymbiotic acquisitions from the mitochondrial ancestor.

When and How Often Did Mitochondria Arise?

Figure 3

The oldest undisputedly eukaryotic microfossils go back 1.45 billion years in the fossil

record. Given the coincidence of mitochondria with the eukaryotic state, this can also be

seen as a minimum age for mitochondria and a rough best-guess starting date for eukaryotic

evolution. According to newer geochemical views, this date of origin corresponds to a

protracted phase in Earth history when the oceans were mostly anoxic — from 1.8 billion

years ago until about 580 million years ago — because of the workings of marine, H2S-

producing bacteria. Eukaryotes thus arose and diversified in an environment where anoxia

was commonplace. Accordingly it is hardly surprising that many independent eukaryotic

lineages have preserved anaerobic energy-producing pathways in their mitochondria

(Figure 3).

Like eukaryotes themselves, mitochondria appear to have arisen only once in all of

evolution. The best evidence for the single origin of mitochondria comes from a conserved

set of clearly homologous and commonly inherited genes preserved in the mitochondrial

DNA across all known eukaryotic groups. In the case of hydrogenosomes (which usually

lack DNA) and mitosomes (which so far always lack DNA), the strongest evidence for their

common ancestry with mitochondria is twofold. First, aspects and components of the

mitochondrial protein import process are conserved in hydrogenosomes and mitosomes,

arguing strongly for common ancestry with mitochondria. Second, all known lineages of

eukaryotes that possess hydrogenosomes or mitosomes branch as sisters to mitochondrion-

bearing lineages.

Summary

Mitochondria arose once in evolution, and their origin entailed an endosymbiosis

accompanied by gene transfers from the endosymbiont to the host. Anaerobic mitochondria

pose a puzzle for traditional views on mitochondrial origins but fit nicely in newer theories

on mitochondrial evolution that were formulated specifically to take the common ancestry

of mitochondria and hydrogenosomes into account. The presence of mitochondria in the

eukaryote common ancestor continues to change the way we look at eukaryote origins, with

endosymbiosis playing a more central role in considerations on the matter now than it did

twenty years ago. The integral part that mitochondria play in many aspects of eukaryote

biology might well reflect their role in the origin of eukaryotes themselves.