Note to users. If you're seeing this message, it means that your browser cannot find this page's style/presentation instructions -- or possibly that you are using a browser that does not support current Web standards. Find out more about why this message is appearing, and what you can do to make your experience of our site the best it can be.
Comment on "The Evolution of Modern Eukaryotic Phytoplankton"
Falkowski et al. (1) reviewed the evidence that three disparategroups of algae—dinoflagellates, diatoms, and coccolithophores,each with plastids derived from red algae by secondary endosymbiosis—havecome to dominate the oceans' flora over the past 250 millionyears and speculated about the forces responsible for this domination.Central to this speculation is the "portable plastid hypothesis"(1, 2), which posits that the likelihood with which plastidswill be transferred between eukaryotes by secondary endosymbiosisis directly related to the number of genes in their genomes.The more genes, the argument contends, the more portable theplastid. This hypothesis rests on three claims: (i) red algalplastids retain more genes than do green algal plastids; (ii)gene transfer from the (primary) endosymbiont nucleus to the(secondary) host nucleus is rare; and (iii) red algae have beenacquired by secondary endosymbiosis more often than have greenalgae. Although the limited number of red algal plastids examinedto date do have more genes (3), claims (ii) and (iii) are notconsistent with the available data, thus rendering the hypothesiseffectively unsupported.
Although red algal plastids may contain more genes than thoseof green algae, this difference pales against the nuclear contributionto plastid function. All plastid genomes encode only a smallfraction of the proteins needed for plastid function—atmost 10 to 20%, and only 1% in the case of dinoflagellates.The vast majority of plastid proteins are encoded by nucleargenes; most of these genes are derived from the plastid buthave been transferred to the nuclear genome over time, and theirprotein products are now targeted back to the plastid (4, 5).Indeed, most algae with secondary plastids have entirely losttheir endosymbiont nucleus and, even where retained, this nucleusis largely vestigial and encodes only a few plastid-targetedproteins (6). Therefore, in all known cases of secondary endosymbiosis,the host nucleus must have acquired hundreds of genes encodingplastid-targeted proteins from the endosymbiont nucleus (Fig. 1).Falkowski et al. (1) state that such gene transfer "seldomoccurs,"but all studies based on single genes, genome surveys,or complete genomes show this to be clearly wrong (7–11).In the face of such massive nucleus-to-nucleus gene transfer,it is very unlikely that the presence, in green algae, of asmall proportion of additional nuclear genes encoding plastid-targetedproteins would substantially hinder the portability of theirplastids.
Fig. 1. Algal evolution and the origin and spread of plastids by endosymbiosis. At the top is the single origin of plastids by primary endosymbiosis between a cyanobacterium and a eukaryotic host. This endosymbiont was reduced and integrated, and part of this process involved the transfer of hundreds of genes from the cyanobacterium/plastid to the eukaryotic host nucleus (red arrow). Glaucophytes, red algae, and green algae all descended from this fully integrated partnership. Next, plastids spread to other eukaryotic groups by secondary endosymbiosis (middle). Green algae were most likely involved in two independent events, giving rise to euglenids (turquoise) and chlorarachniophytes (orange). A single endosymbiosis involving a red alga probably gave rise to the chromalveolates (yellow); this group is supported by several molecular characters and gene trees (plotted on the figure). Plastids have apparently been lost in ciliates and Cryptosporidium (and perhaps other lineages), and photosynthesis has been lost in apicomplexa and many other individual lineages. Numbers indicate the approximate number of protein genes in the plastid genomes of the various lineages with secondary red plastids. Finally, dinoflagellates have substituted their ancestral plastid several times, most notably by tertiary endosymbioses involving other chromalveolates (a cryptomonad, a haptophyte, and a diatom) and by serial secondary endosymbiosis involving a green alga (24).
[View Larger Version of this Image (39K GIF file)]
What about the claim that red algae have been involved in secondaryendosymbiosis twice as often as have green algae? Current evidencesuggests that the opposite is true. The "chromalveolate hypothesis"(12)—which posits that all algae believed to possess secondaryred plastids [dinoflagellates, heterokonts (including diatoms),haptophytes (including coccolithophores), cryptomonads, andapicomplexa] acquired them by a single common endosymbiosis—isnow supported by considerable data. First, there are a numberof morphological and biochemical characters that unite someor all chromalveolates, including plastid membrane topology,storage carbohydrates, flagellar structure, and accessory pigments.Most important, all photosynthetic chromalveolates contain chlorophyllc, which is absent from red algae and best interpreted as ashared derived character (12). Second, two plastid-targetedproteins—glyceraldehyde-3-phosphate dehydrogenase (GAPDH)and fructose-1,6-bisphosphate aldolase (FBA)—have unusualevolutionary histories that are unique to chromalveolates, whichalso indicates a common origin of their plastids (13–15).Third, phylogenies of concatenated plastid genes support a cladecomprising cryptomonads, heterokonts, and haptophytes, whichsuggests that their plastids are derived from a single secondaryendosymbiotic event (16, 17). Fourth, phylogenies of individualand concatenated cytosolic proteins and rRNAs indicate a sister-grouprelationship of alveolates (dinoflagellates, apicomplexans,and ciliates) and heterokonts but do not yet resolve the positionof haptophytes and cryptomonads (18–21). Taken together,there is increasingly strong evidence for a single, common originfor these organisms and their plastids, but no strong evidencefor any alternative.
We agree with Falkowski et al. (1) that there have probablybeen three independent secondary endosymbioses of green algae.Therefore, red algae have most likely been involved in fewer,not more, secondary endosymbiotic events than have green algae(Fig. 1). Does this mean that green plastids are somehow moreportable? The answer is almost certainly no, because the totalnumber of secondary endosymbioses is so low and the differencesbetween these small numbers (four versus two, or one versusthree) are insubstantial.
The foundations of the portable plastid hypothesis do not holdup to scrutiny. This is broadly important because the emergingview that "red"secondary plastids probably originated only oncehas obvious implications for how we interpret not only the processof endosymbiosis but also the fossil record, the evolutionaryhistory of marine algae and their plastids, and their role inshaping today's oceans.
Patrick J. Keeling
Department of Botany University of British Columbia Vancouver, British Columbia V6T 1Z4, Canada E-mail: pkeeling{at}interchange.ubc.ca
John M. Archibald
Department of Biochemistry and Molecular Biology Dalhousie University Halifax, Nova Scotia, B3H 1X5, Canada E-mail: jmarchib{at}dal.ca
Naomi M. Fast
Department of Botany University of British Columbia Vancouver, British Columbia, V6T 1Z4, Canada E-mail: nfast{at}interchange.ubc.ca
Jeffrey D. Palmer
Department of Biology Indiana University Bloomington, IN 47405, USA E-mail: jpalmer{at}bio.indiana.edu
2. D. Grzebyk, O. Schofield, C. Vetriani, P. G. Falkowski, J. Phycol.39, 259 (2003). [ISI]
3. The claim that red algal plastids contain more protein genes (200) than do green algal plastids (at most 100) is consistent with existing data, but only three red algal plastid genomes have been sequenced. The ancestor of red algae must have had at least this many genes, but this does not necessarily mean that the genome of the ancestor (or ancestors) of secondary red algal plastids retained all of these genes. If, as Falkowski et al. (1, 2) postulate, there have been multiple red plastid secondary endosymbioses, one must identify the donor red lineage for each endosymbiosis and show that it contained 200 plastid protein genes to infer that red plastids are in general twice as portable, with respect to plastid gene content, as are green plastids. Indeed, this is a premature claim, given that the few sequenced plastid genomes from cryptomonads, heterokonts, haptophytes, and dinoflagellates contain only about 150, 140, 110 (22, 23), and 15 protein genes, respectively. Therefore, under the hypothesis of multiple red secondary endosymbioses, it is entirely possible that the red algal progenitors of some or even all of these plastids had many fewer than 200 plastid protein genes.
4. G. I. McFadden, J. Eukaryot. Microbiol.46, 339 (1999). [ISI] [Medline]
22. V. Sanchez Puerta, C. Delwiche, personal communication.
23. M. Turmel, F. Lang, C. Lemieux, personal communication.
24. This depiction of plastid evolution differs from figure 4 in (1) in several ways. This scheme takes into account evidence for a single origin of chromalveolate plastids and includes plastid-bearing groups not considered in (1). The most important of these groups is the apicomplexa, but we also show that tertiary endosymbiosis has occurred between dinoflagellates and diatoms (Kryptoperidinium) as well as cryptomonads and haptophytes. Falkowski et al. show the green alga-containing dinoflagellates arising from an ancestral "alveolate" in a secondary endosymbiotic event at the same level as that which gave rise to peridinin-containing dinoflagellate plastids. However, there are only two closely related genera of dinoflagellates with green algal plastids, and it is widely accepted that they originated relatively recently from a peridinin-containing dinoflagellate host, not from an ancestral alveolate. Falkowski et al. show all plastids to be morphologically identical and, most important, show all secondary and tertiary plastids as having two membranes. In reality, secondary plastids have four membranes, except those of euglenids and peridinin-containing dinoflagellates, which are bound by three membranes. Membrane structure of tertiary plastids varies. Falkowski et al. also omitted the relict nuclei (nucleomorphs) found in the secondary endosymbionts of chlorarachniophytes and cryptomonads (as well as the relict diatom nucleus in the tertiary plastid of Kryptoperidinium).
Received for publication 10 August 2004. Accepted for publication 2 December 2004.
The editors suggest the following Related Resources on Science sites:
In Science Magazine
TECHNICAL COMMENTS
Daniel Grzebyk, Miriam E. Katz, Andrew H. Knoll, Antonietta Quigg, John A. Raven, Oscar Schofield, F. J. R. Taylor, and Paul G. Falkowski (24 December 2004) Science306 (5705), 2191c.
[DOI: 10.1126/science.1105297] |Full Text »|PDF »
REVIEW
Paul G. Falkowski, Miriam E. Katz, Andrew H. Knoll, Antonietta Quigg, John A. Raven, Oscar Schofield, and F. J. R. Taylor (16 July 2004) Science305 (5682), 354.
[DOI: 10.1126/science.1095964] |Abstract »|Full Text »|PDF »|Supporting Online Material »
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES:
Evolutionary Relationships of Apusomonads Inferred from Taxon-Rich Analyses of 6 Nuclear Encoded Genes.
E. Kim, A. G. B. Simpson, and L. E. Graham (2006)
Mol. Biol. Evol.
23, 2455-2466
|Abstract »|Full Text »|PDF »
10 to 20%, and only 
