Figures
Abstract
In the summer of 2008, immediately prior to the Beijing Olympics, a massive green tide of the genus Ulva covered the Qingdao coast of the Yellow Sea in China. Based on molecular analyses using the nuclear encoded rDNA internal transcribed spacer (ITS) region, the Qingdao strains dominating the green tide were reported to be included in a single phylogenetic clade, currently regarded as a single species. On the other hand, our detailed phylogenetic analyses of the clade, using a higher resolution DNA marker, suggested that two genetically separate entities could be included within the clade. However, speciation within the Ulva clade has not yet been examined. We examined the occurrence of an intricate speciation within the clade, including the Qingdao strains, via combined studies of culture, hybridization and phylogenetic analysis. The two entities separated by our phylogenetic analyses of the clade were simply distinguished as U. linza and U. prolifera morphologically by the absence or presence of branches in cultured thalli. The inclusion of sexual strains and several asexual strains were found in each taxon. Hybridizations among the sexual strains also supported the separation by a partial gamete incompatibility. The sexually reproducing Qingdao strains crossed with U. prolifera without any reproductive boundary, but a complete reproductive isolation to U. linza occurred by gamete incompatibility. The results demonstrate that the U. prolifera group includes two types of sexual strains distinguishable by crossing affinity to U. linza. Species identification within the Ulva clade requires high resolution DNA markers and/or hybridization experiments and is not possible by reliance on the ITS markers alone.
Citation: Hiraoka M, Ichihara K, Zhu W, Ma J, Shimada S (2011) Culture and Hybridization Experiments on an Ulva Clade Including the Qingdao Strain Blooming in the Yellow Sea. PLoS ONE 6(5): e19371. https://doi.org/10.1371/journal.pone.0019371
Editor: Simon Joly, Montreal Botanical Garden, Canada
Received: September 28, 2010; Accepted: April 3, 2011; Published: May 5, 2011
Copyright: © 2011 Hiraoka et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: WZ is employed by a commercial company, Xianshan Xuwen Seaweed Development Co., Ltd. There was no other external funding. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have read the journal's policy and have the following conflicts. WZ is employed by a commercial company, Xianshan Xuwen Seaweed Development Co., Ltd. This does not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials.
Introduction
Excessive growth of green macroalgae termed “green tides” have over the last four decades increasingly occurred in coastal areas worldwide, causing both ecological and economic impacts to coastal environments and human activities [1], [2]. In late June 2008, before the opening of the Beijing Olympics, a massive green tide of filamentous Ulva covering approximately 600 km2 along the coast of Qingdao, host city for the sailing regatta, lasted over two weeks, and over one million tons of the Ulva thalli were removed by more than 10,000 people (Fig. 1) [3]–[6]. As the cost to manage the impact of the 2008 Qingdao bloom is estimated as more than U.S. $100 million, proper management with efficient methods to suppress the Ulva growth is needed to prevent future massive blooms [6].
Different growth responses among various Ulva species to light, temperature, salinity and nutrient concentrations under controlled conditions have been reported [7], [8]. Different Ulva species bloom at different times and at different locations, so it is predicted that there are between-species differences of ecological impacts [9]. Recently, green tides composed of different species have been found to have different ecological functions, and are causing specific spatiotemporal occurrences [10], [11]. In order to understand the characteristics of green tides, a correct identification is important and it is an essential step to draw up an ecological profile of the species [4].
The genus Ulva presents extreme taxonomic difficulties, particularly in the types that form free floating blooms, because of their plasticity with season and environmental conditions and their broad intraspecific variations, whereas morphological differences between species are small [12]. Therefore, species identification in recent taxonomic studies of such green macroalgae has been confirmed using molecular markers such as the nuclear encoded rDNA internal transcribed spacer (ITS) region or plastid encoded large subunit ribulose 1,5-bisphosphate carboxylase gene (rbcL) [12]–[15]. Molecular analyses showed that the Qingdao strains are included in the Ulva linza-procera-prolifera clade (LPP clade) [4], [16], which is a monophyletic group on the basis of sequence analysis of the ITS region, including provisionally three Ulva species (U. linza, U. prolifera and U. procera) [17]. Taxonomically, the entire clade is currently regarded as U. linza [14]. However, in some publications the Qingdao strains are regarded as U. prolifera (or Enteromorpha prolifera) due to the filamentous, intensively branched morphology [18]–[21], because for U. linza normally a lack of branching is a diagnostic character [22], [23]. Furthermore, our detailed phylogenetic analyses of the LPP clade using a more resolved DNA marker, the 5S rDNA spacer, have suggested that two genetically separate entities could be included within the clade [17]. Although such taxonomic confusion in reference to the Qingdao strains occurs, it has not yet been examined whether the sequence-based species of the LPP clade contains two biological species or not. Thus, we conducted culture and hybridization experiments under controlled laboratory conditions in order to assess the species differentiation in genetic morphology and reproductive compatibility between the two entities within the LPP clade. Furthermore, the specific status of the Qingdao strains was evaluated and discussed.
Materials and Methods
Ulva strains
Collection data and characteristics of strains in the present study are listed in Table 1. Attached samples of U. linza and U. prolifera were collected from marine littoral habitats and brackish river mouths, respectively. Life histories of all strains of U. linza were firstly examined in the present study, while life histories of U. prolifera strains, except UPK, have been reported previously [24]. Collection of floating samples from the bloom was made at the third Bathing Beach, Qingdao, Shandong Province, China, where vast accumulations of filamentous Ulva thalli occurred in July, 2008 (Fig. 1). The floating samples were composed of numerous thallus fragments with dense branches, the morphology of which is consistent with that described in previous reports [19]–[21]. Two fragments were cultured for preliminary investigation. As they were indistinguishable in terms of developmental morphology of thallus branching and types of zoids, one of the two fragments was used for the crossing experiments.
Culture and crossing experiments
The culture method and crossing experiments were described previously [24], [25]. Each strain was isolated by a unialgal technique using phototaxis of zoids. Life history of the strain was determined by microscopic observations of released zoids through at least two generations. When the original strain did not release gametes but zoospores, these could not be used directly for crossing tests. Therefore, the zoospores were cultured up to their gametophyte stage which can discharge gametes. Gametes from each strain were concentrated via positive phototaxis in autoclaved seawater. Small droplets of the dense gamete suspension from two different strains were mixed on a glass slide under a photomicroscope to observe any conjugation between them. As crossing ability of gametes swimming in autoclaved seawater became lower several hours after their release, crossing tests between different species or taxa were performed immediately after checking each for self-compatibility. Success or failure of copulation was confirmed by more than two replicate crossing tests. In successful copulation the hybrid zygotes that formed were isolated via negative phototaxis and cultured up to their reproductive stage, the sporophyte. Formation and release of zoospores in the sporophytes was also observed under a photomicroscope. Morphology of the cultured thalli of ≥20 mm long was observed to determine the absence or presence of branches.
DNA phylogenetic analyses
DNA extraction, PCR amplification and automated sequencing were carried out for all strains, as described previously [17]. Sequences of ULO15, ULS, ULT, ULA and UPK, shown in Table 1, were newly determined since our previous reports [16], [17]. Phylogenetic analyses were performed using the maximum likelihood (ML) algorithm available in the computer program PAUP* 4.0 b10 with JC69 model [26]. Bootstrap values based on 100 re-samplings in ML of the dataset were calculated (TBR, full heuristic search option) [27].
Results
The morphology and life histories of samples, determined by culture experiments, are shown in Table 1. Through several generations each strain repeatedly displayed the same type of thallus morphology in laboratory culture. Two morphotypes of the absence or presence of branches were distinguishable and taxonomically identified as U. linza and U. prolifera, respectively. The morphology of cultured thalli of the Qingdao strains was highly branching. The Qingdao strains had more branches than U. prolifera strains. Young cultured thalli had 4–6 branches per 1 mm in the Qingdao strains and 1–3 branches per 1 mm in U. prolifera strains (Fig. 2). The U. prolifera group included three different life histories being sexual, asexual with biflagellate zoids, and asexual with quadriflagellate zoids. Two filamentous thallus samples collected from the Qingdao bloom released the same types of quadriflagellate zoospores, which developed into gametophytes producing gametes. Therefore, it was determined that the Qingdao samples have a sexual life history with an alternation of isomorphic male and female gametophytes and sporophytes. In the U. linza group life histories of sexual and asexual with quadriflagellate zoids were found. The life histories of the LPP samples are placed on diagrams of ITS and 5S rDNA spacer regions (Figs. 3 and 4). Although division of the U. linza group and U. prolifera group was unclear on the ITS tree, the 5S rDNA spacer analyses clearly and consistently divided them into two groupings. Sexual and asexual strains intermingled within each group and did not fall into subgroups.
(A) Ulva linza (ULC631). (B) U. prolifera (UPE21). (C) The Qingdao strain. The scale bars represent 5 µm.
Initial letters UL and UP in the strain codes stand for U. linza and U. prolifera, respectively. Detailed information of each strain is given in Table 1. Numerals at internal nodes are bootstrap values >50% for 100 replicates in ML analysis.
Numerals at internal nodes are bootstrap values >50% for 100 replicates in ML analysis. As in Figure 3, red lines indicate a sexual type, blue lines an asexual quadriflagellate type and green lines an asexual biflagellate type.
Hybridizations were conducted among representatives of U. prolifera and U. linza and the Qingdao strains (Fig. 5). The results of U. prolifera×U. linza showed an unusual mating activity. Female gametes of U. prolifera successfully copulated with male gametes of U. linza, producing hybrid zygotes which can normally develop into sporophytes. In the successful copulation, after male and female gametes were mixed together, they immediately adhered to one another, the result being the formation of numerous clumps and zygotes by fusion of male and female gametes. However, in U. prolifera male×U. linza female no clumping occurred, indicating a complete gamete incompatibility in this combination. On the other hand, no sterility barrier between the Qingdao strains and U. prolifera strains was found in the laboratory experiments. Their hybrid zygotes normally developed into sporophytes which can produce zoospores. In the Qingdao strains×U. linza strains, no formation of clumps was observed in any combination of male/female gametes.
+ = success of copulation, development of the isolated hybrid zygotes, and production of zoospores in the hybrid sporophytes, − = no observation of clumps or aggregations of mixed gametes.
Discussion
To suppress the considerable plasticity of Ulva thalli in response to changing environmental conditions, morphological characterizations of unialgal strains cultured under controlled laboratory conditions have been successfully conducted in previous taxonomic works of Ulva [22], [23], [28]. Also in the present study the culture experiments demonstrated the existence of at least two genetic morphotypes, U. linza and U. prolifera, within the LPP clade. Although the Qingdao strains with branches are categorized into U. prolifera, the degree of branches was denser than that of other strains of U. prolifera. The result suggests that the Qingdao strains may have different types of genes associated with branching patterns.
Only asexual strains have been previously reported in U. linza [23], [28]. As we first discovered sexual strains of U. linza in this study, crossing experiments could be carried out between U. linza and U. prolifera. The partial (asymmetric) reproductive isolation between U. prolifera and U. linza suggests a recent speciation of the two entities, that is consistently supported by the genetic separation, shown by an approximately 10 times more variable molecular marker, the 5S rDNA spacer, than the ITS region (Figs. 3 and 4) [17]. As contrasted with the asymmetric reproductive isolation between U. prolifera and U. linza, a complete reproductive barrier exists between the Qingdao strains and U. linza strains (Fig. 5). This implies that the U. prolifera group includes two types of sexual strains distinguishable by crossing affinity to U. linza. The results of hybridization experiments suggest that the Qingdao strains completed the speciation from U. linza via intermediate U. prolifera.
It has been reported that U. intestinalis widely occurs in the Baltic Sea with low salinities, while the distribution of U. compressa is restricted to salinities higher than 15 ppt and the limited distribution of the latter was explained by a lack of tolerance to low salinities [29]. Similarly, strains of U. linza are marine, not growing in oligohaline to mesohaline environments [23], [30], while those of U. prolifera are found commonly in estuaries and brackish waters [24], [30], having a wide range of salinity tolerance from oligohaline to euhaline conditions [22]. In our previous survey of U. linza and U. prolifera samples from a total of 53 populations from Japan, the two species had obviously distinct distributions, which are only marine seashore areas or mainly brackish river mouth areas, respectively [17]. Furthermore, the small proportion of U. prolifera that do inhabit marine seashores do not co-occur with U. linza [17]. Considering their habitat, it is thought that the salinity gradient may act as an isolating mechanism and cause the partial or complete reproductive isolation between U. linza and U. prolifera observed in our hybridizations.
Differences between U. linza and U. prolifera can be summarized as follows: (1) their genetic morphology was distinct by the absence or presence of branches, (2) they phylogenetically divide into two discrete clusters based on the 5S rDNA, (3) a partial or complete reproductive boundary exists between them, (4) they have different distributions. Thus, we conclude that U. linza and U. prolifera in the LPP clade should be regarded as two separate species. Although the Qingdao strains fall into U. prolifera, the different characteristics from the other brackish U. prolifera are recognized in the high degree of branches, the lack of crossing affinity to U. linza, and the marine habitat. Therefore, we consider that the Qingdao strains should be taxonomically dealt in infraspecific categories such as subspecies, varietas or forma. To establish the taxonomic status of the Qingdao strains, a more detailed comparison of diagnostic features, such as branching patterns, would be needed.
Sexual and asexual life history types have been reported in U. prolifera (as Enteromorpha prolifera) [24], [28]. In the present study we first found the Qingdao strains have the same sexual type of life history including an alternation of isomorphic sporophytes and male and female gametophytes, and a sexual mating process by male and female gametes. On the other hand, the asexual life histories reproducing by biflagellate or quadriflagellate zoids specialized for settlement have been cytologically proved to be quite independent and not to be the unusually prolonged phases of the sexual life history [24]. The asexual strains with quadriflagellate zoids of U. prolifera in our study can be taxonomically regarded as U. procera because U. procera has a branching morphology and an asexual life history by means of quadriflagellate zoids as key diagnostic characters [14], [22], [28]. Necessary nomenclatural changes resulting from this study will be formally addressed in a following paper.
Recently, ITS2, which is a part of the ITS region, has been suggested to be a molecular key to the level of biological species; in clades of eukaryotes, the identity of ITS2 predicts meaningful intercrossing ability, while a difference of even one compensatory base change in this region predicts total failure of crossing [31], [32]. Such a correlation of the molecular marker with reproductive barriers has also been predicted in Ulva taxonomy [33]. However, although all strains used in the present crossing experiments are identical in the ITS2 sequence, the crossing ability among them shows three different patterns which are successful crossing, partial or complete gamete incompatibility (Fig. 5). This high variation in mating ability within the LPP clade demonstrates that their mate recognition traits are evolving more rapidly than the ITS sequences. Many genes that mediate sexual reproduction, such as those involved in gamete recognition, diverge rapidly in eukaryotic organisms including the unicellular green alga Chlamydomonas [34], [35]. The mating process in Ulva is essentially similar to that in Chlamydomonas [36], [37]. In both algae, male and female gametes have two flagellae and the initial contact between mating gametes involves the flagellar tips of both mating types, which leads to the sexual agglutination and finally cell fusion [38]. Interestingly, female gametes of U. prolifera can copulate with both male gametes of U. prolifera and male gametes of U. linza and furthermore male gametes of U. linza can also copulate with both female gametes of U. prolifera and female gametes of U. linza. The crossing compatibility between these Ulva strains provides an important clue to the gamete recognition system, which directly correlates with speciation and reproductive isolation, particularly with pre-zygotic isolation. The molecular genetic evidence in Ulva must be detected on the model of Chlamydomonas, of which the mating system has been the most thoroughly studied [39].
Clades based on grouping procedures using ITS sequences are deductively considered to reflect genetically isolated entities, providing good resolution for identification at and below the species-level in a wide range of eukaryotic organisms [15] and also in Ulva [13], [14]. Therefore, tracking the algal origin of the green tide is being conducted by reliance on the ITS sequences [19]–[21]. However, we showed that a single clade based on the ITS sequences contains at least two separate biological species and their derivatively independent asexual strains. If the ITS markers are excessively relied upon as criteria to identify species, there is a risk of misunderstanding the ecophysiological characteristics of green tides by mistaking the actual causative strains.
Acknowledgments
We thank Dr. Gabrielle Prendergast, the Shimanto City Board of Education, and Dr. Chris Norman, Japan Scientific Texts, for kindly correcting the English manuscript.
Author Contributions
Conceived and designed the experiments: MH WZ JM. Performed the experiments: MH KI SS. Analyzed the data: MH KI SS. Contributed reagents/materials/analysis tools: MH KI WZ JM SS. Wrote the paper: MH.
References
- 1.
Fletcher RL (1996) The occurrence of “green tides” – a review. In: Schramm W, Nienhuis PH, editors. Marine benthic vegetation, recent changes and the effects of eutrophication, Ecological Studies 123. Berlin: Springer. pp. 7–43.
- 2. Morand P, Merceron M (2005) Macroalgal population and sustainability. J Coastal Res 21: 1009–1020.
- 3. Liu D, Keesing JK, Xing Q, Shi P (2009) World's largest macroalgal bloom caused by expansion of seaweed aquaculture in China. Mar Pollut Bull 58: 888–895.
- 4. Leliaert F, Zhang X, Ye N, Malta E-J, Engelen AH, et al. (2009) Identity of the Qingdao algal bloom. Phycol Res 57: 147–151.
- 5. Liu D, Keesing JK, Dong Z, Zhen Y, Di B, et al. (2010) Recurrence of the world' largest green-tide in 2009 in Yellow Sea, China: Porphyra yezoensis aquaculture rafts confirmed as nursery for macroalgal blooms. Mar Pollut Bull 60: 1423–1432.
- 6. Hu C, Li D, Chen C, Ge J, Muller-Karger FE, et al. (2010) On the recurrent Ulva prolifera blooms in the Yellow Sea and East China Sea. J Geophys Res 115: C05017.
- 7. Taylor R, Fletcher RL, Raven JA (2001) Preliminary studies on the growth of selected ‘green tide’ algae in laboratory culture: effects of irradiance, temperature, salinity and nutrients on growth rate. Bot Mar 44: 327–336.
- 8. Raven JA, Taylor R (2003) Macroalgal growth in nutrient-enriched estuaries: a biogeochemical and evolutionary perspective. Water Air Soil Poll 3: 7–26.
- 9. Raffaelli DG, Raven JA, Poole L (1998) Ecological impact of green macroalgal blooms. Oceanogr Mar Biol 36: 97–125.
- 10. Nelson TA, Nelson AV, Tjoelker M (2003) Seasonal and spatial patterns of “green tides” (ulvoid algal blooms) and related water quality parameters in the coastal waters of Washington state, USA. Bot Mar 46: 263–275.
- 11. Nelson TA, Haberlin K, Nelson AV, Ribarich H, Hotchkiss R, et al. (2008) Ecological and physiological controls of species composition in green macroalgal blooms. Ecology 89: 1287–1298.
- 12. Malta E-J, Draisma SA, Kamermans P (1999) Free-floating Ulva in the southwest Netherlands: species or morphotypes? A morphological, molecular and ecological comparison. Eur J Phycol 34: 443–454.
- 13. Hayden HS, Waaland JR (2004) A molecular systematic study of Ulva (Ulvaceae, Ulvales) from the northeast Pacific. Phycologia 43: 364–382.
- 14.
Maggs CA, Blomster J, Mineur F, Kelly J (2007) Ulva. In: Brodie J, Maggs CA, John DM, editors. Green seaweeds of Britain and Ireland. London: British Phycological Society. pp. 80–103.
- 15. Leliaert F, Verbruggen H, Wysor B, De Clerck O (2009) DNA taxonomy in morphologically plastic taxa: Algorithmic species delimitation in the Boodlea complex (Chlorophyta: Cladophorales). Mol Phylogenet Evol 53: 122–133.
- 16. Shimada S, Nagano M, Hiraoka M, Ichihara K, Mineur F, et al. (2010) Phylogeographic analysis of the genus Ulva (Ulvales, Chlorophyta), including bloom sample in Qingdao, China. Coast Mar Sci 34: 117–122.
- 17. Shimada S, Yokoyama N, Arai S, Hiraoka M (2008) Phylogeography of the genus Ulva (Ulvophyceae, Chlorophyta), with special reference to the Japanese freshwater and brackish taxa. J Appl Phycol 20: 979–989.
- 18. Zhang X, Wang H, Mao Y, Liang C, Zhuang Z, et al. (2010) Somatic cells serve as a potential propagule bank of Enteromorpha prolifera forming a green tide in the Yellow Sea, China. J Appl Phycol 22: 173–180.
- 19. Pang SJ, Liu F, Shan TF, Xu N, Zhang ZH, et al. (2010) Tracking the algal origin of the Ulva bloom in the Yellow Sea by a combination of molecular, morphological and physiological analyses. Mar Environ Res 69: 207–215.
- 20. Wang J, Jiang P, Cui Y, Li N, Wang M, et al. (2010) Molecular analysis of green-tide-forming macroalgae in the Yellow Sea. Aquat Bot 93: 25–31.
- 21. Liu F, Pang SJ, Chopin T, Xu N, Shan TF, et al. (2010) The dominant Ulva strain of the 2008 green algal bloom in the Yellow Sea was not detected in the coastal waters of Qingdao in the following winter. J Appl Phycol 22: 531–540.
- 22. Koeman RPT, van den Hoek C (1982) The taxonomy of Enteromorpha Link, 1820, (Chlorophyceae) in the Netherlands. II. The section Proliferae. Cryptogamie Algol 3: 37–70.
- 23. Koeman RPT, van den Hoek C (1984) The taxonomy of Enteromorpha Link, 1820, (Chlorophyceae) in the Netherlands. III. The sections Flexuosae and Clathratae and an addition to the section Proliferae. Cryptogamie Algol 5: 21–61.
- 24. Hiraoka M, Dan A, Shimada S, Hagihira M, Migita M, et al. (2003) Different life histories of Enteromorpha prolifera (Ulvales, Chlorophyta) from four rivers on Shikoku Island, Japan. Phycologia 42: 275–284.
- 25. Hiraoka M, Ohno M, Kawaguchi S, Yoshida G (2004) Crossing test among floating Ulva thalli forming ‘green tide’ in Japan. Hydrobiologia 512: 239–245.
- 26.
Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony (*and other methods), Version 4. Sunderland, Massachusetts: Sinauer Associates, Inc.
- 27. Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791.
- 28. Bliding C (1963) A critical survey of European taxa in Ulvales. Part I. Capsosiphon, Percursaria, Blidingia, Enteromorpha. Opera Bot 8: 1–160.
- 29. Leskinen E, Alström-Rapaport C, Pamilo P (2004) Phylogeographical structure, distribution and genetic variation of the green algae Ulva intestinalis and U. compressa (Chlorophyta) in the Baltic Sea area. Mol Ecol 13: 2257–2265.
- 30.
Burrow EM (1991) Seaweeds of the British Isles, Volume 2 Chlorophyta. London: British Museum (Natural History). 238 p.
- 31. Müller T, Philippi N, Dandekar T, Schultz J, Wolf M (2007) Distinguishing species. RNA 13: 1469–1472.
- 32. Coleman AW (2009) Is there a molecular key to the level of “biological species” in eukaryotes? A DNA guide. Mol Phylogenet Evol 50: 197–203.
- 33. O'Kelly CJ, Kurihara A, Shipley TC, Sherwood AR (2010) Molecular assessment of Ulva spp. (Ulvophyceae, Chlorophyta) in the Hawaiian Islands. J Phycol 46: 728–735.
- 34. Swanson WJ, Vacquier VD (2002) The rapid evolution of reproductive proteins. Nat Rev Genet 3: 137–144.
- 35. Ferris PJ, Pavlovic C, Fabry S, Goodenough UW (1997) Rapid evolution of sex-related genes in Chlamydomonas. Proc Natl Acad Sci U S A 94: 8634–8639.
- 36. Bråten T (1971) The ultrastructure of fertilization and zygote formation in the green alga Ulva mutabilis Føyn. J Cell Sci 9: 621–635.
- 37. Løvlie A, Bryhni E (1976) Signal for cell fusion. Nature 263: 779–781.
- 38. Miyamura S (2004) Visualization of the rapid fertilization process of the marine green alga, Ulva arasakii Chihara (Ulvophyceae, Chlorophyta) using high-speed video microscopy. Cytologia 69: 197–201.
- 39. Ferris PJ, Waffenschmidt S, Umen JG, Lin H, Lee J-H, et al. (2005) Plus and minus sexual agglutinins from Chlamydomonas reinhardtii. The Plant Cell 17: 597–615.