Escherichia coli cells use long, thin structures called flagella to propel themselves. Source: NSF.gov.
Scientists, ever since the time of Aristotle, have been classifying creatures and looking at the relatedness among biological species or their “phylogeny”. Life on earth can be classified into three major types or domains:
Bacteria (like E. coli) and Archaea are microscopic creatures with their DNA open to the rest of their cell. Eukaryotes, on the other hand, are organisms with their DNA enclosed in a separate membrane within the cell. This includes us, the trees in our back yards, the mushrooms on our pizzas, and most of the living world that we see with our naked eyes.
Figuring out the relationships among eukaryotes can have very practical applications. For example, the discovery that the Microsporidia, once thought to represent an ancient lineage, are actually closely related to fungi,12,14 opened up unforeseen potential avenues of fighting this important agricultural and economic pest. Eukaryotic phylogeny may also yield answers to a number of questions about how:
- some eukaryotes acquired the ability to survive just by being out in the sun (photosynthesis)
- the energy-producing machinery in our cells (mitochondria) came to be there
- and, ultimately, how we got to be the way we are
History of eukaryotic phylogeny
Morphology-based phylogeny For a long time phylogeny was based only on morphology: the study of structural characters such as shape, size, or features that an organism might possess. This was how the visible biological world was initially tackled and, with the advent of the microscope, how much of the microbiological world was studied as well. This type of work has been and continues to be immensely important and is still how many organisms are classified.
The breakthrough of DNA sequencing technology gave rise to a new field of evolutionary studies: molecular phylogenetics. The idea underlying this field is that, in general, the more closely related two organisms, the more similar their gene sequences will be. This is because they wouldn’t have had time to accumulate many independent changes in their gene sequences. By statistically comparing the similarities and differences in the DNA sequence between the same gene from various organisms, we can deduce the pattern of how those organisms are related. This pattern is normally drawn out in a tree diagram much like a family tree. The closer two lineages are on a tree, the closer their evolutionary relationship. A standard convention is that the deeper the split in the tree, the longer ago that evolutionary split (or “speciation”) took place.
Computer analysis of the first gene extensively used to study molecular phylogeny — the so called ssu rDNA gene — gave us a seemingly definitive picture (see Figure 1) of the eukaryotic landscape.22 A number of lineages emerged sequentially from the base of the tree, apparently corresponding to lineages that evolved away from the rest of eukaryotes early in our history. These were followed by an unresolved or “crown radiation” of groups that included animals, plants, fungi and a number of unicelled eukaryotic lineages, collectively referred to as “protists.”
Our understanding of eukaryotic relationships circa 1991. This is the ssu rDNA tree redrawn from Sogin 1991  and modified from a figure in Dacks and Doolittle 2001 .
Unfortunately, as more data was accumulated and better methods of analysis were developed, the strength of this structure began to erode. The phylogeny derived from protein genes didn’t seem to agree with this picture8, and even more importantly a special type of pervasive problem called Long Branch Attraction was discovered to be in play in the computer analyses.10,18 The problem in these analyses is that genes that had evolved very quickly were incorrectly concluded to be related.
In the case of eukaryotic phylogeny, this
- hid some true relationships
- made others look real when they weren’t
- led us to the questionable conclusion that particular organisms were very ancient species
- AND, when compensated for, many of the clear answers in our initial picture of eukaryotic relationships disappeared
It must be pointed out, however, that not all ssu rDNA studies are subject to Long Branch Attraction. The sequence for the ssu rDNA gene is still the best place to start studying a new organism, and is very useful for figuring out specific relationships, as will be discussed later. However, it doesn’t look like THE answer to figuring out the overall relationship among eukaryotes.
In addition to the ssu rDNA gene, scientists have tried using a number of other single genes to reconstruct the eukaryotic tree. Each attempt has been met with similar problems of Long Branch Attraction and uncertainty.7,11,13,21 Long Branch Attraction is not a phenomenon restricted to ssu rDNA but will affect any set of molecular data where the different sequences have evolved at very different rates.
In the face of the uncertainty surrounding eukaryotic relationships it has even been proposed that this task might, for real biological reasons, be impossible.18,19 A theory called “the eukaryotic big bang” suggested that the reason researchers couldn’t find clear relationships between the major groups was that those groups had evolved from one another very rapidly. This separation of the groups into species had happened so quickly that there hadn’t been time to accumulate enough changes in genes that we study for scientists to be able to figure out the pattern of relatedness.
However, in the year 2000 and early 2001, a number of recent advances have emphasized that specific large-scale relationships are, in fact, discernible. Building comprehensive structure in the eukaryotic tree in a piece-by-piece fashion may be the right way to approach the study of eukaryotic phylogeny.
1. Are red and green algae related: who caught the first chloroplast?
In the 1970s a shocking discovery was made that several integral pieces of our cellular machinery, such as the energy producing mitochondrion and the light harvesting chloroplast, were actually captured and permanently resident bacteria. This idea, actually proposed as early as 190516 was summarized and updated in the “serial endosymbiotic theory.”15 Although wildly controversial at the time, the molecular phylogeny of mitochondrial and chloroplast genes strongly backed this theory and it has now been widely accepted. However, a number of major questions still remain.
Perhaps the biggest question is how many times did ancient eukaryotes capture bacteria and turn them into chloroplasts. This isn’t a small question but involves the relationships of three major groups of eukaryotes:
- the red algae (the outside wrappings of sushi rolls)
- the land plants /green algae (trees, grasses and green slime on fish tanks)
- a group of protists called the glaucocystophytes
All three groups look like they have captured bacteria and turned them into chloroplasts. If the groups are related then the capture was probably done once, by their common ancestor. If they aren’t related, however, they may each have independently captured their own chloroplasts. Analysis of chloroplast genes suggest that the three groups are closely related but nuclear genes have, so far, been inconclusive or moderately supportive of the groups being unrelated.
Early in 2000 however, a joint publication from French and Spanish groups weighed in heavily on one side.17 Using both single gene and a combined gene analysis, these teams showed that red and green algae are closely related. The glaucocystophytes were only weakly grouped with red and green algae in this analysis but other evidence points strongly to glaucocystophytes being related to red algae.23 Together this suggests that the three groups are indeed descended from a common ancestor (Note: for a dissenting opinion see ). The uniting of these algal groups (see Figure 2) is a major step forward in resolving the eukaryotic tree and points quite strongly towards a single original capturing of the bacterium that would eventually become the chloroplast.
Some suggested relationships among eukaryotes circa 2001. This image was modified from a figure in Dacks and Doolittle 2001 .
2. Algae and beyond: when does the captor become the captured?
Aquatic photosynthetic creatures are common and numerous.
- Many are classified in a large group called the stramenopiles, also called the heterokonts or chromists.4 These include the giant kelp, tiny diatoms and the golden algae in our lakes and oceans.
- Another common sea creature, the dinoflagellate causes Red Tide. These are related, in a group of protists called the alveolates, to things as diverse as the high school biology stand-by Paramecium (ciliates) and the parasite that causes malaria (Plasmodium, an apicomplexan).
Unfortunately it has never been clear where either of these consortiums fit in the larger scheme of relationships among eukaryotes.
In a blockbuster of a paper9 in 2001, a Canadian group from the University of British Columbia showed that the heterokonts and alveolates are evolutionarily linked (see Figure 2). This team discovered that these groups share a particular gene replacement in their photosynthetic apparatus. Not only does the computer analysis of this gene unite the groups, but also the fact that they all share this extremely rare feature makes it very likely that they share a recent common ancestor. This discovery not only unites a major percentage of known eukaryotic diversity but also answers some fundamental questions about the evolution of photosynthesis. The chloroplast is such a useful machine that the organisms that captured them the first time have themselves been captured and co-opted, creating in essence a cell within a cell within a cell. Uniting the alveolates and the stramenopiles into a group now called the Chromalveolates4, suggests that their common ancestor was a eukaryote that captured a red alga and made it act as a chloroplast.
3. From lakebeds to termite guts: lessons on losing mitochondria?
Sometimes organisms are discovered that have no apparent relationship to the rest of eukaryotic life. Trimastix, which lives in the sediments of lakes and oceans, is a good example. This organism, first described in 1886, but was practically forgotten and not studied until the late 1990s. It has no mitochondria but it has been suggested that it possesses a vestigial one. Trimastix shares a number of characteristics with a loose assemblage of parasites and marine organisms called the excavates. Some of the excavates also have no mitochondria, while others have mitochondria that most closely resemble what we think the original mitochondrion looked like. At one time it was thought that those excavates without mitochondria might even be ancient lineages that evolved before the ancestor of the rest of eukaryotes captured the original mitochondria.3 However, scientists now believe most of those excavates have probably had mitochondria at one time and then lost them.20
Long-forgotten organisms like Trimastix are exciting. Just as interesting, however, are organisms whose phylogenetic relationship is still a mystery after 100 years of study. Such was the case with the oxymonads (see Figure 3), a group of protists that live as symbionts in the guts of termites and allow termites to digest wood.2
Drawing of Oxymonas, an oxymonad protist. This image was modified from Brugerolle and Konig 1997 .
This year however, in a collaboration between myself and a number of international colleagues, the answer seems to have been found.6 We obtained the first ssu rDNA gene sequences from Trimastix and oxymonads and, after checking for possible Long Branch Attraction effects, showed by computer analysis of these genes that oxymonads and Trimastix are closely and specifically related.
- Since oxymonads have no mitochondria but are related to Trimastix (which probably have modified mitochondria), the oxymonads likely also had mitochondria at one time and then lost them.
- The oxymonads and Trimastix together may be a good model for examining the process of losing mitochondria.
- Two major groups of organisms (see Figure 2), whose phylogenetic relationships were previously open questions, are now anchored down.
4. Concatenated datasets: phylogenetic voices above the noise
Despite the seeming lack of resolution in the overall structure of the eukaryotic tree when only one type of evidence is used, a new approach may be able to break through this problem. While one gene might not have enough differences to solve the relationships among the major eukaryotic groups, using the accumulated differences from a number of genes may be able to give us answers. This approach was taken by researchers at Dalhousie University who, late in 2000, strung together the four most commonly used protein markers into a single combined (or concatenated) dataset with striking results.1
Their concatenated dataset analysis provides strong evidence for a number of previously weakly supported or unseen relationships.
- A large grouping of amoebae was demonstrated
- The amoebae were linked to a cluster of groups which includes animals and fungi
- The euglenozoa (including the causative agent for African sleeping sickness) and the heterolobosea (agent causing amoebic meningoencephalitis) were convincingly united
- Some of the relationships above were also supported, including the red plus green algae and the chromalveolates
The real breakthrough was not any single relationship recovered, but the fact that they were all demonstrated in the same analysis and that they were so consistent with the single pieces of convincing data above.
Where to next?
For a while it looked as if figuring out the relationships among the major eukaryotic groups was hopeless. Eukaryotes probably did evolve rapidly into the major groups. However, in the last year and a bit, we have seen that piecing together the eukaryotic tree should be possible (see Figure 2). We have to abandon our hopes for a quick fix. It’s unlikely that a single gene is going to give us all the answers but, by using various approaches, determining the relationships among eukaryotes should be possible. This will be a difficult task but what else would we expect from a challenge as daunting as reconstructing events billions of years old and navigating the web of life that we see all around us?
Editor’s note: Some of the topics in this article, as well as the reconstruction of the ancestral eukaryotic cell and the application of genomics to such questions is covered slightly more in depth in . Acknowledgements: The author would like to thank W. Ford Doolittle for supervision. He also thanks W. Ford Doolittle, Gurston Dacks, Patrick Keeling, Barbara Dacks, Karen Slevinsky, Maureen O’ Malley, David Dacks and an anonymous reviewer for critical comments on this manuscript. Financial support has been provided by the Walter C. Sumner Foundation and a Canadian Institute for Health Research (C.I.H.R.) doctoral research award, as well as C.I.H.R. grant MT4467 to W. Ford Doolittle.
© 2002, American Institute of Biological Sciences. Educators have permission to reprint articles for classroom use; other users, please contact firstname.lastname@example.org for reprint permission. See reprint policy.