Photo: Nils Kroger, Regensburg University

Last summer's publication of the first diatom genome provided insight into the workings of a tiny organism with huge potential for environmental, industrial, and research applications.1 A growing appreciation of the sequence, however, has begun to divulge one of nature's wilder and most productive experiments.

Diatoms, a diverse division of one-celled ocean algae with gemlike silica casings, are thought to collectively absorb as much carbon dioxide through photosynthesis as all the world's rainforests. They appear to have descended from organisms near the plant-animal divergence in evolution, and their unique and intricate structures serve both as inspiration and model for nanofabrication technologies.

Yet beneath their sometimes striking enclosures, diatoms and many other algae reveal a sordid past of endosymbiotic events. They are, in essence, three organisms in one – or more precisely, one within another, within another. Researchers say this odd arrangement has major implications...


Secondary endosymbiosis was considered "just a wild hypothesis," says Sarah Gibbs, professor emeritus at McGill University, Montreal, who first proposed the idea more than 25 years ago.2 But, says Geoffrey McFadden, a professor at the University of Melbourne's School of Botany, the idea has inched closer to acceptance over time. "We're pretty much there now," he says. Further study of this special case may elucidate processes for all eukaryotes.

In the diatom's history, an ancestor of a red alga engulfed a cyanobacterium. Its descendants became the alga's chloroplasts and lost or ceded most of their genes to the nucleus. Later, an ancestral diatom engulfed both organisms and took over most of their genes.3 The red alga is now virtually invisible, having withered to no more than a pair of extra membranes around the chloroplast.

The story itself had languished for a time. Gibbs provided its basic outlines in a 1978 paper on Euglena, an organism distinct from diatoms that has a single extra membrane around its chloroplast.2 She deduced it "from just a few little scraps of information," says McFadden. Nature, which had rejected her paper, later published similar findings on secondary endosymbiosis,4 "as soon as it was confirmed at the molecular level by sequencing," Gibbs says. Last summer's genome clinched the case, adds Daniel Rokhsar, department head for computational genomics at Lawrence Berkeley National Laboratory and a coauthor on the paper. "Any skepticism is probably excluded by the genome," he says, which contains homologs of red algal and cyanobacterial genes.

Researchers say studies of secondary endosymbiosis can clarify the relationships among the algae, a somewhat artificial grouping that spreads uncooperatively into various phyla. Traditionally defined as photosynthetic, plantlike eukaryotes lacking true stems, roots and leaves, many algae are easily classifiable either as plants or protozoans, and others as fungi. But algae do have one thing in common: all arose from primary or secondary endosymbioses involving chloroplasts, McFadden says. Some algae contain chloroplasts pure and simple – which descended from cyanobacteria, the first photosynthesizers, once generally considered algae themselves. Others contain chloroplasts wrapped in remains of other former algae, mostly either red or green algae.

Taxonomists can sketch reasonable algal classification systems, or "trees," based on endosymbiotic origins, McFadden argues, although developing true evolutionary trees will require much additional evidence. "Endosymbiotic events will certainly have a tremendous influence on the look of the tree," says Edward Theriot, a professor of integrative biology at the University of Texas at Austin.

Certain types of endosymbiotic events may have happened multiple times, for example. A case in point is a debate over "red lineage" algae.5These, which include diatoms and other algae that got chloroplasts via red algae, dominate marine ecosystems more than any other algal group. Some propose the lineage descends from unrelated organisms that engulfed red algae at different times. This view is based on a greater apparent diversity of the "red lineage" compared to other groups, and a finding that red algal chloroplasts contain more genes critical to their own functioning than do other types, which might have helped them successfully invade multiple organisms.5 Other researchers counter – based on recent findings that these algae share genetic quirks beyond their distinctive chloroplasts – that the red lineage is monophyletic and stems from just one secondary endosymbiosis event.6 Somewhat similar considerations suggest a single event spawned chloroplasts as a whole, McFadden says.


Furthermore, cells formed by secondary endosymbiosis serve as unique models to explore gene swapping between formerly free living organisms. In endosymbiosis, the host-cell nucleus eventually takes over most, but not all, of the chloroplast genes. In secondary endosymbiosis, the host also usually appropriates all the genes from the nucleus of the chloroplast-containing eukaryote that was engulfed. Although in some species, vestigial nuclei hang on. These "nucleomorphs" interest researchers in their own right;3 but considerably more attention and debate centers, as Rokhsar says, on "why certain sets of genes are preferentially transferred" from the organelle to the main nucleus.


© 2004 Botanical Society of America

Many of the primary (top), secondary (middle), and tertiary and serial (bottom) endosymbiosis events in plastid-containing eukaryotes are represented here. Secondary endosymbiosis involving red algae created a large and diverse group of eukaryotes including the major lineage known as heterokonts, which includes diatoms. But whether the red algae lineages trace back to a single secondary endosymbiotic event or to several remains controversial. (From P.J. Keeling, Am J Botany, 91:1481–93, 2004.)

Such questions are important for human health, too. Efforts to defeat mitochondrial diseases by engineering nucleus-encoded, functional versions of defective mitochondrial genes might fail, if certain hypotheses are correct, writes Aubrey de Grey of the University of Cambridge, UK.7

Some hypothesize that genes that still haven't moved to the nucleus will never work there. The most popular hypothesis is that organelles retain certain genes vital to their energy-producing functions to provide critical fine-tuning possible only with on-site location.8 This predicts that such genes will never turn up in the nucleus, no matter what the species, says John Allen, professor of biochemistry at the University of London and author of the hypothesis. "I'm sort of heartened" by the diatom genome, Allen says. He claims it bears out his prediction in showing that the chloroplast retained the photosynthetic reaction center gene psbA.

"Chloroplasts and mitochondria have to be permitted a degree of autonomy without prior consultation with the nucleus," Allen explains. But de Grey raises arguments in favor of older hypotheses: that some mitochondrial genetic codes differ from the standard nuclear code, so that transferred genes would encode incorrect amino acids; or that most mitochondria-encoded proteins are hydrophobic, which could hinder import. A fourth possibility is that the transfer of organelle genes to the nucleus is still in progress.

Whether complete or not, why organellar genes began to migrate is less a source of controversy. An oft-cited reason is a need for central coordination: "The nucleus needs to take control of the situation," says Deborah Robertson, an assistant professor of biology at Clark University, Worcester, Mass. Other reasons given are the presence of free radicals in energy-producing organelles that could damage sensitive genes; and further possible damage to those genes from a permanently asexual existence in the organelle. Although it's unclear just how genes move to the nucleus, recent research suggests it happens readily.9 McFadden says one gene, cbbX, apparently entered cyanobacteria through a plasmid, journeyed to the red alga nucleus, and finally settled down in the diatom nucleus. Like cbbX, he adds, many genes have unique stories within the general framework of secondary endosymbiosis: some exist both in chloroplast and nuclear copies, which apparently compete for survival.

Finding out how these oddities fit into a bigger picture and quieting the debates surrounding secondary endosymbiosis will require more genomes, researchers claim. The diatom sequence "will be important, [but] the problem right now is there simply isn't a lot to compare it to," Theriot notes. Rokhsar says that will change: "We will soon, I think by the end of the year, have a couple more diatom genomes in hand."

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?