Commentary

Continued evolutionary surprises among

Clifford W. Morden*†‡ and Alison R. Sherwood*

*Department of Botany and †Hawaiian Evolutionary Biology Program, University of Hawaii, 3190 Maile Way, Honolulu, HI 96822

t is well established that in (also in green ), fucoxanthin, chl c1 gene encoding the carbon assimilation Igreen and are derived from a and c2 (also in and hapto- protein Rubisco, rbcL, has frequently primary endosymbiotic event between a phytes) and chl c1 and phycobilins (also in been used to examine deep phylogenetic cyanobacterium and a eukaryotic organ- cryptophytes), and are believed to be the relationships (12–14). However, recent ism Ϸ1 billion years ago (Fig. 1; refs. 1 and products of further endosymbioses with studies have demonstrated that there are 2). Although these two groups account for species from those groups (5–8). In this several problems associated with using many of the world’s photosynthetic spe- issue of PNAS, Yoon et al. (9) provide rbcL for this purpose. First, there appears cies, most other major taxonomic groups startling new evi- to have been ram- of photosynthetic organisms (strameno- dence that implicates pant horizontal piles—including diatoms, phaeophytes, plas- gene transfer chrysophytes—and ) have tids containing fucox- Long-held theories suggested that among various al- derived from a photosynthetic eu- anthin and chl c1 and primitive dinoflagellates were gal groups (14, c (derived from a 15). Reliance on karyote implying a secondary endosym- 2 heterotrophic and the addition biosis (1, 2). Still other groups, such as the ancestor) such a gene to de- dinoflagellates, have more complicated as being ancestral to of a peridinin-containing termine potential associations believed to be derived from those with peridinin. occurred relatively recently. endosymbiotic re- tertiary endosymbioses involving the en- This new paradigm in lations would be gulfment of a secondary endosymbiont. the relationship of fraught with po- Each endosymbiotic event has character- these species forces tential misinter- istic structural changes associated with it, us to rethink many aspects of dinoflagel- pretations. Second, peridinin-containing the most notable of which is the addition late evolution. dinoflagellates have a nuclear encoded of two membranes surrounding the plastid In many respects, the findings by Yoon form II Rubisco (L2 structure rather than (the inner representing the cell membrane et al. (9) allow a more parsimonious view the L8S8 structure typical of most eu- of the engulfed organism and the outer of dinoflagellate evolution. Previously karotic plastids) that is similar in sequence representing the phagocytosis long-held theories suggested that primi- homology to anaerobic (16–18). membrane) (2). Dinoflagellates, although tive dinoflagellates were heterotrophic Third, multiple copies of Rubisco have believed to be tertiary endosymbionts, and the addition of a peridinin-containing been found in some organisms (14). Al- have only 3 membranes surrounding their plastid occurred relatively recently (10), though identical or nearly identical copies plastids (1, 2), suggesting that the acqui- which was borne out by the fact that of related genes are often sition of too many membranes may be approximately 50% of dinoflagellate spe- found in (19–21), both functionally unstable and can cause some cies are heterotrophic (1) and that many form I and form II Rubisco have been to be lost. photosynthetic dinoflagellates have re- found in some species. Dinoflagellates are fascinating organ- tained heterotrophic behavior (11). How- The genes chosen by Yoon et al. (9) were psaA (encoding the photosystem I isms that have intrigued researchers for ever, recent studies have demonstrated P700 apoprotein A1) and psbA (encoding many years. They are most well known for that photosynthetic dinoflagellates are the photosystem II thylacoid protein D1). toxic blooms associated with red tides and clearly ancestral, and that heterotrophy These genes are intimately tied to the symbiotic relationships with corals (zoox- has been independently derived numerous photosystem reaction centers and have a anthellae) (2). They contain an astound- times within the group (8). Fucoxanthin low rate of base substitution (nonsynono- ing array of unique features that has been and chl c1 and c2 are the predominant mous replacement rate K ϭ 0.022 and the impetus for continued evolutionary form of light harvesting pigments among A 0.009 for psaA and psbA, respectively; ref. studies. One is their close phylogenetic stramenopiles and haptophytes. With the 22) compared with other genes such as finding that haptophytes are the sister link with apicomplexans, organisms that rbcL (K ϭ 0.041, considered slowly evolv- group to dinoflagellates in phylogenetic A are best known for causing some of our ing; ref. 22), so that significant phyloge- analyses and that dinoflagellates with fu- most deadly infectious diseases (3, 4). netic signal should be expected even at coxanthin and chl c and c are apparently Another is the diverse array of light har- 1 2 these deep evolutionary positions. The vesting pigments within the group. Peri- ancestral to the peridinin-containing spe- resulting phylogenies were rigorously ex- dinin is a xanthophyll found exclusively in cies, there is no need to hypothesize an amined using various positions of the pro- dinoflagellates and, together with chl a independent origin of these pigments nor tein codons, amino acid sequences, and (ubiquitous among photosynthetic organ- a later endosymbiosis of a haptophyte in isms), makes up the light harvesting com- dinoflagellates. plex found in most species. Dinoflagel- Selection of the proper gene for exam- See companion article on page 11724. lates with other combinations of plastid ination of deep phylogenetic relationships ‡To whom reprint requests should be addressed. E-mail: pigments are also known, including chl b was critical for carrying out this study. The [email protected].

11558–11560 ͉ PNAS ͉ September 3, 2002 ͉ vol. 99 ͉ no. 18 www.pnas.org͞cgi͞doi͞10.1073͞pnas.192456999 Downloaded by guest on September 26, 2021 Fig. 1. Theoretical steps of primary, secondary, and tertiary plastid endosymbiosis. (1) A nonpho- tosynthetic engulfs a cyanobacterium. (2) The engulfed cyanobacterium becomes the cell’s plastid, and gene transfer occurs between the cyanobacterium cell and the host . The cyanobacterial membrane becomes the inner plas- Fig. 2. Hypothetical evolutionary relationships among algal groups and derivations of the various tid membrane and the eukaryotic vacuolar mem- dinoflagellate lineages based on possible plastid endosymbiosis events. (1) Initial symbiosis of a cyano- brane becomes the outer plastid membrane (pri- phyte/ progenitor (see text). (2) Loss of residual peptidoglycan . (3) Loss of phycobilins mary endosymbiosis). (3) A nonphotosynthetic and gain of chl b.(4) Secondary endosymbiosis of green alga with retention of algal nucleus as nucleo- eukaryote engulfs a eukaryote with a primary plas- morph between second and third layers of plastid membrane. (5) Secondary endosymbiosis of green alga COMMENTARY tid. (4) Four plastid membranes are present from by dinoflagellate (Dinoflagellates 1). (6) Secondary endosymbiosis of red alga by cryptophyte with the primary and secondary plastids (see text); the retention of algal nucleus () between second and third layers of plastid membrane. nucleus of the secondary plastid is reduced to a (7) Tertiary endosymbiosis of cryptophyte by dinoflagellate (Dinoflagellates 2). (8) Secondary endosym- nucleomorph (which is present in the chlorarach- biosis of red alga; loss of phycobilins with gain of chl c1 and c2 as light harvesting pigments; gain of niophytes and the cryptophyes); gene transfer oc- xanthophylls (predominantly fucoxanthin) as accessory pigments. (9) Tertiary endosymbiosis of diatom by curs between the nucleomorph of the secondary dinoflagellate (Dinoflagellates 3). (10) Possibly a tertiary endosymbiosis of by haptophyte plastid and the nucleus of the host cell (secondary (see text). (11) Possibly a tertiary or quaternary endosymbiosis of haptophyte by dinoflagellate (Dinoflagellates endosymbiosis). (5) A eukaryotic cell with a second- 4) (see text and figure 3 of ref. 9 for further details). (12) Loss of fucoxanthin and chl c1 and c2 after development ary plastid is engulfed by a nonphotosynthetic eu- of peridinin as light harvesting pigment (Dinoflagellates 5). (13) Loss of plastids; roughly half of peridinin- karyote (tertiary endosymbiosis). The remnant containing dinoflagellates (Dinoflagellates 6). (14) Loss of photosynthesis in apicomplexans. nucleomorph of the secondary plastid is lost. (6)Up to six plastid membranes are theoretically possible (see text). The nucleus of the engulfed cell has been sis (9) (Fig. 2). Haptophyte and This then begs the question, how many reduced to a nucleomorph. (7) Gene transfer occurs dinoflagellate plastids, in turn, form a endosymbiotic events are required to ac- between the nucleomorph of the tertiary plastid sister group to the stramenopiles, suggest- count for all of the diversity present and the host cell nucleus. (8) The nucleomorph of the tertiary plastid is lost. Abbreviations: 1° End, ing that they were derived from a single among photosynthetic algae? This is a primary endosymbiosis; 2° End, secondary endo- common ancestor. It cannot be ruled out difficult question that we may never fully symbiosis; 3° End, tertiary endosymbiosis; Cy, cya- that this ancestor was more red alga-like understand. However, evidence from nu- nobacterium; E, eukaryote; GT, gene transfer; N, (23), but given the features shared be- clear small subunit RNA sequences may nucleus; NM, nucleomorph; P, plastid. tween stramenopiles and haptophytes help shed light on the answer (24, 25). (four membrane layers surrounding the Taken at face value, red algae, strameno- plastid and similar pigment composition piles, haptophytes, and dinoflagellate͞ statistical testing, all pointing to a hapto- phyte being the sister group to both peri- among other characters) (2), it is likely apicomplexans form four distinct and these share a close relationship. It also unrelated . Given the plastid phy- dinin and fucoxanthin, chl c1 and c2 dinoflagellates. Supporting analyses were seems clear that dinoflagellate (and pos- logeny (Fig. 2), this would suggest that also performed with 16S rRNA data, al- sibly apicomplexan) plastids are derived plastids of red algae are primary endo- though data sets were not as complete. from haptophyte ancestors (Fig. 2). Yoon symbionts and those of stramenopiles are What are the relationships among stra- et al. (9) hypothesize a separate origin for secondary endosymbionts (both consis- menopiles, haptophytes, and dinoflagel- the plastids within apicomplexans, yet a tent with previous theories) (1, 2). How- lates? It seems clear that stramenopile recent study suggests a common origin for ever, if haptophyte plastids are derived plastids form a monophyletic group and the plastids of both groups (4). If so, it is from a primitive stramenopile plastid, as were derived from a red alga-like progen- likely that both are derived from a hapto- believed, its occurrence would necessarily itor by means of a secondary endosymbio- phyte progenitor. represent a tertiary endosymbiotic event.

Morden and Sherwood PNAS ͉ September 3, 2002 ͉ vol. 99 ͉ no. 18 ͉ 11559 Downloaded by guest on September 26, 2021 Similarly then, it follows that the a common ancestry is that most Kryptoperidinium foliaceum (6, 7) and dinoflagellate (and apicomplexan) plastid dinoflagellates have retained a heterotro- Durinskia baltica (8, 28), and of crypto- could conceivably represent a quaternary phic history; most species still gain phytes in Gymnodinium acidotum (5). A endosymbiotic event! Such a sequence of nourishment through phagocytosis of final conjectured example is the transient events seems contrary to morphological other organisms (11), perhaps indicating symbiont of an anaerobic bacterium hy- and physiological patterns in these groups, an incomplete transfer of nuclear-en- pothesized to have been engulfed by an especially among the haptophytes. coded plastid genes from their endosym- early dinoflagellate, apparently while the A closer examination of the nuclear biotic eukaryotic plastid (see below). Al- haptophyte plastid was still in place (9), phylogeny trees (24, 25) indicates weak though the evidence is less strong, the and contributed its form II rbcL gene to branch support for the lineages that sep- fossil record also suggests a different or- the nuclear genome (16–18). arate haptophytes and stramenopiles. It is igin. Dinoflagellate-like organisms are Acquisition of new plastids for photo- therefore possible that haptophytes are an present as early as in Silurian sediments Ϸ synthesis is only one of several necessary ancient derivation from stramenopiles af- ( 400 million years ago), whereas modern changes for an organism to become com- ter fucoxanthin and chl c1 and c2 evolved dinoflagellate fossils are only found in pletely autotrophic. Much of the cya- in the lineage. This relationship is sup- sediments from the late Triassic (200 mil- nobacterial genome responsible for pho- ported by Yoon et al. (9) in that the lion years ago), well after haptophytes had tosynthesis and other metabolic functions haptophyte lineage is sister to that of become established in the Carboniferous was transported from the plastid to the stramenopiles rather than from within it to Triassic periods (290 to 220 million nuclear genome in the early endosymbi- as might be expected if resulting from a years ago) (2). ont. Should this organism have been taken tertiary endosymbiosis. Another indication that dinoflagellate up as a secondary endosymbiont, it would Separation of haptophytes and plastids are derived from a separate en- similarly have been necessary to move dinoflagellates is similarly weakly sup- dosymbiosis is their propensity toward these genes from the nuclear genome of ported in nuclear phylogenetic trees (24, forming endosymbiotic relationships. Sev- 25). Might these disparate organisms also eral examples exist of dinoflagellates ac- the primary symbiont to the nucleus of the share a common ancestor? This possibility quiring extra symbionts that eventually secondary symbiont. Our review of the seems less likely. Structurally, the plastids displace the original plastid symbiont literature has determined that over 300 are different in the two groups (as pointed (note multiple positions of dinoflagellates nuclear-encoded plastid genes are present out previously) with dinoflagellates hav- in the plastid tree in Fig. 2). The recruit- in flowering systems, nearly three ing only three membranes surrounding the ment of a new plastid does not appear to times the number of genes encoded in plastid and the outer most membrane not be linked to strictly heterotrophic organ- plant DNA (29, 30). Although being continuous with that of the nuclear isms; heterotrophic species are scattered not functionally photosynthetic, it is likely envelope (1). Plastid chromosomes are throughout the dinoflagellate phylogeny that apicomplexans have maintained their also apparently different; photosystem (8) and some taxa that have undergone plastids for these other functions (31). genes in dinoflagellates are found sepa- plastid replacement have maintained their It is a tribute to dinoflagellates that they rate on minicircular DNA chromosomes former plastid as a three membrane- have readily taken up additional symbi- rather than a single large chromosome bound eyespot within the cell (1, 2). Ex- onts in so many different lineages. Such a typical of most plastid DNA (26, 27). This amples of the secondary endosymbiosis of proclivity must also have a readily acti- occurrence may also help to explain why by a dinoflagellate are found vated system for moving genes from the Yoon et al. (9) found dinoflagellate plastid in Lepidodinium viride and Gymnodinium symbiont to the nuclear genome of the DNA to evolve at a rate much faster than chlorophorum (8). The tertiary endosym- host. Characteristics such as these make other plastid DNAs. A third strike against biosis of diatoms has been well studied in dinoflagellates kings of symbioses.

1. Van Den Hoek, C., Mann, D. G. & Jahns, H. M. 11. Schnepf, E. & Elbraechter, M. (1999) Eur. J. 23. Durnford, D. G., Deane, J. A., Tan, S., McFadden, (1995) Algae: An Introduction of Phycology (Cam- Protistol. 28, 3–24. G. I. & Green, B. R. (1999) J. Mol. Evol. 48, 59–68. bridge Univ. Press, Cambridge, U.K.). 12. Morden, C. W. & Golden, S. S. (1991) J. Mol. Evol. 24. Van de Peer, Y. & de Wachter, R. (1997) J. Mol. 2. Graham, L. E. & Wilcox, L. W. (2000) Algae 32, 379–395. Evol. 45, 619–630. (Prentice–Hall, Englewood Cliffs, NJ). 13. Morden, C. W., Delwiche, C. F., Kuhsel, M. & 25. Van de Peer, Y., Baldauf, S. L., Doolittle, W. F. & 3. McFadden, G. I., Reith, M. E., Munholland, J. & Palmer, J. D. (1992) BioSystems 28, 75–90. Meyer, A. (2000) J. Mol. Evol. 51, 565–576. Lang-Unnasch, N. (1996) Nature (London) 381, 14. Delwiche, C. F. & Palmer, J. D. (1996) Mol. Biol. 26. Zhang, Z., Green, B. R. & Cavalier-Smith, T. 482. Evol. 13, 873–882. (1999) Nature (London) 400, 155–159. 4. Fast, N. M., Kinninger, J. C., Roos, D. S. & 15. Palmer, J. D. (1995) BioEssays 17, 1005–1008. 27. Barbrook, A. C. & Howe, C. J. (2000) Mol. Gen. Mol. Biol. Evol. 18, Keeling, P. J. (2001) 418–426. 16. Morse, D., Salois, P., Markovic, P. & Hastings, Genet. 263, 152–158. 5. Wilcox, L. W. & Wedemeyer, G. J. (1984) J. J. W. (1995) Science 268, 1622–1624. 28. Chesnick, J. M. & Cox, E. R. (1987) BioSystems 21, Phycol. 20, 236–242. 17. Palmer, J. D. (1996) Plant Cell 8, 343–345. 69–78. 6. Chesnick, J. M., Morden, C. W. & Schmeig, A. M. 18. Rowan, R., Whitney, S. M., Fowler, A. & Yellow- 29. Shinozaki, K., Ohme, M., Tanaka, M., Wakasugi, (1996) J. Phycol. 32, 850–857. lees, D. (1996) Plant Cell 8, 539–553. T., Hayshida, N., Metsubayasha, T., Zaita, N., 7. Chesnick, J. M., Kooistra W. H. C. F., Wellbrock, 19. Golden, S. S., Brusslan, J. & Haselkorn, R. (1986) U. & Medlin, L. K. (1997) J. Eukaryotic Microbiol. EMBO J. 5, 2789–2798. Chunwongse, J., Obokata, J., Yamaguchi- 44, 314–320. 20. Golden, S. S. & Stearns, G. W. (1988) Gene 67, Shinozaki, K., et al. (1986) EMBO J. 5, 2043– 8. Saldarriaga, J. F., Taylor, F. J. R., Keeling, P. J. & 85–96. 2049. Cavalier-Smith, T. (2001) J. Mol. Evol. 53, 204– 21. Morden, C. W. & Golden, S. S. (1989) Nature 30. Hiratsuka, J., Shimada, H., Whittier, R., Ishibashi, 213. (London) 337, 382–385. T., Sakamoto, M., Mori, M., Kondo, C. Honji, Y., 9. Yoon, H. S., Hackett, J. D. & Bhattacharya, D. 22. Wolfe, K. H. (1991) in Cell Culture and Somatic Sun, C.-R., Meng, B.-Y., et al. (1989) Mol. Gen. (2002) Proc. Natl. Acad. Sci. USA 99, 11724–11729. Cell Genetics of , eds. Bogorad, L. & Vasil, Genet. 217, 185–194. 10. Taylor, F. J. R. (1980) BioSystems 13, 65–108. I. K. (Academic, London), Vol. 7B, pp. 467–482. 31. Palmer, J. D. (1997) Science 275, 790–791.

11560 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.192456999 Morden and Sherwood Downloaded by guest on September 26, 2021