Chemoattraction of male gametes by a pheromone produced by ...

Proc. Natl. Acad. Sci. USA Vol. 92, pp. 641-645, January 1995 Plant Biology

Chemoattraction of male gametes by a pheromone produced by female gametes of Chlamydomonas RICHARD C. STARR*t, FRANz JOSEPH MARNERt, AND LOTHAR JAENICKEt *Department of Botany, University of Texas at Austin, Austin, TX 78713; and tInstitut fiir Biochemie,

Universitat Koln, Koln, Germany D-50678

Contributed by Richard C. Starr, October 13, 1994

The isogamous strains of C. eugametos, C. moewusii, and C. reinhardtii serve as model systems for the study of the physiology and biochemistry of the sexual process in the green algae. In these two species, clumping and pairing of compatible gametes occur by chance when the flagella bearing agglutinins come in contact with each other. In contrast, gametes in the brown algae are brought together via chemoattraction when the female gamete secretes a pheromone to which the male gametes respond. The only report of chemoattraction in Chlamydomonas was that of Tsubo (2, 3) in which he demonstrated the presence of volatile compounds of unknown identity produced by one strain of C. eugametos that would attract cells of the opposite mating type as well as one or both strains of other pairs of mating strains. Tsubo used small capillary tubes containing the attractant to demonstrate this attraction, but attraction of cells by single cells of the opposite mating type was not observed as one sees in the various genera of the brown algae where chemoattraction is so well documented (4). Harris (5) does not mention the excellent account of chemoattraction of male gametes by single female gametes in Chlamydomonas paupera (6) inasmuch as the present day circumscription of the genus Chlamydomonas no longer includes in the genus those species without pyrenoids (starch centers) in the chloroplast. These pyrenoid-less species of Chlamydomonas, including Pascher's C. paupera, are now placed in the genus Chloromonas (7). Dried soil samples are routinely used as a means by which algae from distant collecting sites can be transported to the laboratory and, even after a period of years, serve as a source of living natural populations of algae from which unialgal cultures can be established (8). When dried soil is placed in a Petri dish and glass-distilled water is added, many dormant algal spores (usually zygotes) may germinate within 48 hr after illumination. If heterotrophic species are being sought, a boiled garden pea (Pisum sativum) is added at the time the soil is wet; the rotting pea provides organic compounds needed by the heterotrophic algae. In March 1992 when a number of soil samples were being screened for certain heterotropic genera, a population of what seemed to be Chloromonas (like Chlamydomonas but without pyrenoids) appeared in great abundance in a soil sample (designated LCH) that had been collected near Lemon Cove, CA, in 1969. Three days after the appearance of the flagellated cells, examination of the population under the compound microscope showed groups of small cells whirling around individual large cells, a phenomenon characteristic of chemoattraction in the brown algae. Clonal axenic cultures were established. Some formed small sperm while others served as foci for the whirling groups of sperm when introduced into a suspension of sperm. Surprisingly, when the algae were grown in the axenic medium, cells regularly formed pyrenoids in their chloroplasts, especially in older cultures, identifying the algae with the genus Chlamydomonas rather than Chloromonas, which lacks pyrenoids. In the latest taxonomic treatment of the genus Chlamydomonas, Ettl (9) recognizes 421 species of the genus living in

In isogamous species of Chlamydomonas, ABSTRACT such as Chlamydomonas reinhardtii and Chlamydomonas eugametos, the sexual process involves the use of flagella agglutinins by which the gametes of compatible strains adhere through chance encounter and ultimately pair and fuse to form zygotes. In a newly described heterogamous species, Chlamydomonas allensworthii, the sexual process is initiated by the chemoattraction of small sperm to a sexually competent female gamete, which continues to secrete the pheromone until it has fused with one of the sperm so attracted. From bacteria-free female strains of C. aUlensworthii, the chemoattractant has been isolated and identified as a pentosylated hydroquinone (Mr = 532) whose spectral, chemical, and physical properties are in accord with the structure of a 2,3-dimethyl-5-(triprenylcarboxymethyl)-1,4-benzohydroquinone-1-(f3-xyloside). A rapid bioassay of the pheromone uses DEAE-Toyopearl 650M beads to which the pheromone adsorbs. When such activated beads are placed in a suspension of sperm, they act as surrogate females and attract the small motile sperm. The purified pheromone shows activity at a concentration as low as 1 pM.

A vital step in the sexual process in animals and plants is the strategy by which the two gametes of opposite mating type find each other. The strategies employed vary, depending on whether the organism producing the gametes, the structures producing the gametes, or the gametes themselves are to be put in close juxtaposition. In seed plants, the pollen grain (the immature male gametophyte) is transported to the pistil (in which one finds the female gametophyte forming) through various agents of pollination, such as the wind, insects, bats, or some special construction of the flower. In those fungi where motile gametes are not present (e.g., Achlya), the hormonally directed growth of the plant body ensures that the reproductive structures are in close proximity. Those plants having one or both gametes flagellated have devised mechanisms by which the gametes themselves are brought together. In oogamous plants such as the ferns or certain algae (e.g., Fucus), chemoattractants direct the swimming of the sperm to the surface of the egg whether it is enclosed within a flask of sterile tissue (the archegonium in the fern) or released into the surrounding liquid medium (as in Fucus). In those algae having flagellated gametes of both mating types, chemoattraction and agglutination, either alone or in combination, serve to accomplish the desired result of getting the two gametes together. The extensive reports by Franz Moewus on the sexuality and genetics of Chlamydomonas eugametos during the 1930s and 1940s stimulated the interest of many scientists, especially in the U.S., although there was much skepticism at the time as to their reliability. This has not changed in the ensuing years (1), but the interest so generated has resulted in Chlamydomonas becoming the alga of choice for much research. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

tTo whom reprint 641

requests

should be addressed.

642

Plant

Proc. Natl. Acad Sc USA 92 (1995)

Biology: Starr et al.

freshwater. The Lemon Cove isolates cannot be identified with any of them, and in view of the unique phenomenon of chemoattraction as a part of the sexual process, the new species Chlamydomonas allensworthii, named after John H. Allensworth, is being described.

SYSTEMATICS Chlamydomonas ailensworthii, sp. nov. (Fig. 1). Cells ellipsoidal to almost spherical, 15 x 11 ,um. Cell wall thin without papilla. Chloroplast cup-shaped with single pyrenoid in lateral position. Two or three pyrenoids sometimes present in largest cells. Elliptical stigma in anterior part of cell. Two flagella of body length. Four contractile vacuoles at base of flagella. Asexual reproduction by two to eight zoospores. Sexual reproduction by anisogametes. Sperm teardrop shaped, 6 x 4 ,um and larger. Female gametes varying in size from slightly larger than sperm to the size of vegetative cells. Chemoattraction of sperm by larger female gametes. Zygotes with a broad, hyaline wall at first, with crenulations forming on surface later. Isolated from Lemon Cove, CA. [Cellulae ellipsoides vel subsphaericae, 15 x 11 ,um; membrana delicata, sine papilla. Chloroplastus olliformis, pyrenoide singulari laterali, pyrenoidibus duobus vel tribus in cellulis maximis aliquando praesentibus. Stigma ellipticum in parte laterali cellulae. Flagella duo, longitudine corpori aequantia, vacuolis pulsantibus quatuor basi flagellorum. Propagatio asexualis per 2-8 zoospora; propagatio sexualis per anisogametis. Spermi lacrimiformes, 6 x 4 ,um vel majores. Gametae femineae statura variantes, minimis spermo paulo longioribus et maximis cellulis vegetativis aequantibus, gametis femineis majoribus spermos attractantibus per chemicam. Zygotae initio pariete latis hyalinisque, crinis superficiaibus postea formantibus. E loco dicto "Lemon Cove, CA" segregata.] Holotype. (Fig. 1). Unlike C. eugametos and C. reinhardtii, C. allensworthii shows a variety of sexual potentials among various clones of identical morphology originating from the same soil sample. Strains of the first two species are always of one of two mating types; Moewus (10) reported so-called subheterothallic strains, but this has never been confirmed. In C. allensworthii, various clonal isolates exhibit sexual potentials ranging from strict heterothallism to homothallism of varying degrees. Intercrossing will occur, but the genetic basis for sexual type is as yet unknown. A similar situation of subheterothallism has been described in Chlamydomonas zimbabwiensis by Heimke and Starr (11). Sexual reproduction is heterogamous, but in young female cultures the female gametes are only slightly larger than the small sperm cells formed by the male strains, and in this case sexual reproduction would appear to be almost isogamous. However, female gametes retain their sexual nature as they enlarge, and thus the heterogamous.nature of the sexual act becomes more and more evident. Sperm cells are special small cells produced by the male strain. Sperm are produced when

C

FIG. 1. C. allensworthii. (A) Median optical section of vegetative cell. (B) Anterior view of vegetative cell showing characteristic four contractile vacuoles. (C) Sperm cell. (X1665.)

the level of the nitrogen in the medium becomes low, and although the female strain may produce active female gametes in medium before the nitrogen level drops, the most active populations are those from medium with lowered nitrogen. Evidence of the presence of a chemoattractant can be demonstrated in several ways; the female cells themselves can be used to show attraction, or the medium in which the females are growing can be used to attract sperm. To demonstrate chemotaxis with the female cells, it is important to use female cells that have been washed by gentle centrifugation to remove the chemoattractant from the medium in which they have been growing. The medium from female cultures often has sufficient attractant to immobilize sperm if they are added directly to it. Washed females continue to secrete the attractant, and thus gradients are set up around each cell to which the sperm cells respond. After a female gamete has been fertilized, it stops producing the chemoattractant, but if incompatible males and females are placed together, the females continue to produce the chemoattractant, and the whirling groups of sperm around the female cells can be observed for more than an hour. The strains from the LCH soil sample and those from another sample (LCN) are identical morphologically, and the gametes show mutual attraction; however, no fusions occur between them. The presence of the chemoattractant in the medium of the female culture can be indicated in several ways. If a fine glass needle is dipped into the medium of a female culture and then dragged through a sperm suspension, a mass attraction of sperm occurs in its path. A more effective and even quantifiable way of assaying the attractant in female cultures is through the use of Toyopearl-DEAE beads. The chemoattractant adsorbs to the beads, which may then be washed to remove excess attractant. Such treated beads, when placed in sperm suspensions, become the centers of attraction for large numbers of sperm.

METHODS Culture Medium. All chemicals were of analytical grade; the solvents were freshly distilled. Quartz-distilled water was used throughout. The culture medium was a modification of the Volvox medium designed by Provasoli and Pintner (12). It was prepared as follows: To 978 ml of glass-distilled water, add the following in the amounts indicated: Conc., Volume, ml Stock solution g/100 ml of H20 1 25.0 NaNO3 2.5 1 CaCl2 2H20 4.0 1 MgSO4 7H20 1 6.0 Disodium glycerophosphate pentahydrate 1 5.0 KCl 10 5.0 Glycylglycine 1 25.0 x 10-6 Biotin 1 15.0 x 10-6 Vitamin B12 1 0.1 Thiamine 6 PIV metal solution

Add 2 g of sodium acetate (anhydrous) and adjust to pH 6.6 using 1 M NaOH. The PIV metal solution was prepared as follows. To 1000 ml of glass-distilled water, add 0.750 g of Na2EDTA and dissolve completely before adding the following salts in order: FeC13 6H20, 97 mg; MnCl2A4H20, 41 mg; ZnCl2, 5 mg; CoCl2-6H20, 2 mg; Na2MoO4-2H20, 4 mg. Store at room temperature in the dark. Stocks were maintained on agar in 20 x 100 mm Petri dishes. The culture medium contained 3 times the concentration of NaNO3 in the basic medium. It is important that the medium

Plant

Biology: Starr et al.

be solidified with no more than 0.75% agar inasmuch as the cells tend to flatten out on the surface of medium with higher percentages. Stocks were transferred every 3 weeks; 1 ml of liquid medium was placed on the agar and a heavy transfer of algal cells was mixed with the medium and spread over the plate with a bent glass Pasteur pipette. Stock plates were maintained on a 16-hr light/8-hr dark cycle under cool white fluorescent lights, giving an intensity of '2000 lux at the level of the plates; the temperature varied between 20 and 25°C. For production of sperm by the male strain, the basic medium with only one-seventh the normal amount of NaNO3 (N/7 medium) was used. For large quantities of sperm, 500-ml Erlenmeyer flasks containing 400 ml of the N/7 medium were used. The flasks were stoppered with cotton wound around a 9-inch disposable glass Pasteur pipette plugged with cotton so that aeration with an air stream from a Silent Giant (Aquarium Pump Supply, Prescott, AZ) aquarium pump is possible after inoculation. The inoculated flasks are placed on a shelf under a 16-hr light/8-hr dark cycle of illumination provided by two power groove cool white fluorescent tubes providing an intensity of "9000 lux at the level of the shelf. Within 4-5 days, the population in the flask is composed mostly of small sperm cells. The flasks were then moved to another shelf receiving -1/10th the illumination and kept at a temperature of 20°C. Each morning the flask of sperm was swirled vigorously to resuspend the cells, which soon returned to the bottom of the flask where the unilateral light was concentrated. Under such conditions, the sperm remained active and responsive to the chemoattractant for a week to 10 days. For production of sexually active females, the N/7 medium was also used. The female strain was sexually active after 1-2 days and can be used at this time to demonstrate the attraction phenomenon. For the large amounts of filtrate from female cultures needed for extraction of the chemoattractant, flasks containing 150 ml of the basic medium were grown for 4 days and then transferred to larger flasks with nitrate-free medium for another 8 days. Bubbling with air was essential for good growth, and the level of illumination was best when kept at an intensity of -3000 lux. Bioassay. The bioassay for the attractant was performed as follows: A 10-fold dilution series (1.0 ml) of the sample was prepared in 10 x 100 mm test tubes. To each tube, 2.5 ,ul of a suspension of DEAE-polyacrylamide beads (ToyopearlDEAE 650M; TosoHaas, Montgomeryville, PA) was added; the mixture was shaken vigorously and allowed to stand for 30 min. Four milliliters of water was added, and the beads were spun down by centrifugation (1000 x g). Adsorption to the DEAE beads was quantitative; the supernatant solution was inactive in a second adsorption cycle. On the other hand, the attractant was not adsorbed to CM-Toyopearl or to plain polyacrylamide beads. In this case, the supernatants were bioactive, and treatment with DEAE beads would yield active beads. A sample of the beads was given with a platinum loop (1-mm diameter) to a droplet of suspended male gametes on a microscope slide and mixed well. The result was read under the microscope at 60- to 100-fold magnification after a minimum of 1 min in orange-red light to avoid disturbance by phototaxis. A positive reaction was indicated by the formation of a halo of sperm cells around the mock female. The higher the concentration of the attractant in the bead and hence in the sample, the denser the halo and the quicker the motility of the male gametes. Very high concentrations and certain (unknown) impurities tend to immobilize the male cells. The threshold was taken as that dilution at which just a visible steady-state accumulation of cells around the bead occurred. The starting media usually had a threshold of about 104 dilution. Isolation of the Attractant. The cells were removed from the female cultures by centrifugation (Sorvall GSA rotor, 10,000 rpm, 30 min), and the slightly slimy, yellowish supernatant was

Proc. NatL Acad ScL USA 92 (1995)

643

concentrated by flash evaporation at 40°C and 35-45 mPa to 1/100th of its volume. The viscous concentrate was mixed with an equal volume of methanol and twice its volume of chloroform and shaked thoroughly for 30 min. The suspension was centrifuged (Sorvall SS34 rotor, 12,000 rpm, 10 min), and the brownish water/methanol phase was saved. It was concentrated to a volume of 10 ml and slowly poured through a small (10 x 30 mm) column of phenyl-RP material. The column was washed several times with its volume of distilled water. The dark brown zone held back at the top was washed down with 3 ml of methanol. All insolubles were removed by centrifugation (Eppendorf centrifuge, 13,000 rpm, 3 min), and the clear brown supernatant was loaded on a Lichrosorb RP18 preparative HPLC column. A water/methanol gradient (50-100% methanol within 20 min) was applied by which the active compound was eluted as a sharp band between 85% and 86% methanol. This fraction was rechromatographed in the same way on a smaller RP18 HPLC column to remove slower and faster moving minor contaminants. The yield of the -95% pure compound was about 0.1 mg/liter of starting mediumi.e., roughly calculating, between 30% and 60% of the active factor in the starting medium (the bioassay does not allow for more exact calculations). After evaporation of the solvent under reduced pressure, a brownish glassy residue remains that did not crystallize despite several attempts with different solvents and cooling cycles. The attractant is soluble in methanol, ethanol, and all other alcohols and in water. It is only sparingly soluble in chloroform and insoluble in cyclohexane. It is slightly acidic and forms easily dissociated alkali salts. The attractant has a UV spectrum of low intensity in the aromatic range with its maximum at 282 nm: E1%cWater) = 50; e = 2650. It quenches fluorescence on indicator-doped silica gel TLC plates and shows a slight blue fluorescence on plain silica gel and RP18 plates. Rf values on silica gel TLC: methanol/water, 1:1 (vol/vol), 0.80; 2-methyl-1-propanol/methanol/water, 8:8:1 (vol/vol), 0.63; 2-methyl-1-propanol/methanol/water, 18:1:1 (vol/vol), 0.52. Rf values on RP18 plates (dried at 100°C): methanol/ water, 60:40 (vol/vol), 0.58. Structure Determination of the Attractant. Mass spectroscopy (solid phase; electron impact, 70 eV) yielded m/z (%) 532 (M+, 0.25), 414 (2), 400 (30), 189 (100), 151 (70), 43 (85). The 13C NMR spectrum (methanol-d4) showed 30 carbon atoms, which could be assigned as follows: 1 carbonyl, 8 C==, 4 CH=, 1 O-CH-0, 3 CH-0, 1 CH2-O, 7 CH2, 5 CH3. This would add up to a plausible sum formula of C30H4408, Mr = 532.06, in accordance with the molecular mass found by mass spectrometry. Of the nine unsaturations calculated from the formula, seven are contained in the carbonyl group and the C==C double bonds, respectively. Thus, two ring systems should be present (one of them aromatic). The 1H NMR spectrum showed 1 aromatic H, 3 olefinic H, 1 O-CH-, several CH-0, 2 benzylic CH3, 3 allylic CH3. The C/H pairings were deduced from a C/H correlated spectroscopy spectrum. Particularly valuable was the longrange correlation 2JCH and 3JCH showing the placement of the substituents relative to the aromatic CH. This gave evidence for the pentosylated and modified plastohydroquinone shown in Compound 1 in which only the dissociable protons have to be supplemented. D-xylose as the substituting sugar has been proven directly as a f3-glycoside by 1H NMR spectroscopy and after acid hydrolysis of the sugar either with HCl or with formic acid and trimethylsilylation followed by different work-ups. In the first, the compound was treated with 2 M HCl at 100°C for 2 hr. After ether extraction and drying of the water phase, the residue was trimethylsilylated with 100 ,ul of bis(trimethlysilyl)trifluoracetamide and analyzed by GC/MS. Comparison with similarly treated standards (D-xylose, D-ribose) showed

644

Plant

Biology: Starr et al.

tris(trimethylsilyl)D-xylose as the only product. The ether extract, analyzed by HPLC in the procedure described above, contained no educt but a biologically inactive compound, less polar though with the same chromophore. For the second procedure, the reduced pressure-dried attractant (E282 = 0.280) was hydrolyzed wih 10 ,ul of concentrated formic acid (50°C, 12 hr). Evaporation of the mixture to dryness and extraction with ether yielded a residue, which was derivatized with bis(trimethylsilyl)trifluoroacetamide and reduced with NaBH4 to the tetrakis(trimethylsilyl)glycitol, which was analyzed by GC together with pentose and hexose standards. The only product is the xylose derivative. The ether extract was dried and evaporated. The residue is more difficultly soluble in water than the adduct; it has a similar spectrum but slightly shifted to shorter wave length: Am. = 227 nm (El cm = 0.340). It is biologically inactive. A full account of the structural identification of the attractant will be published elsewhere. The spectral, chemical and physical properties of the compound isolated are in accord with the structure of a 2,3dimethyl-5-(triprenylcarboxymethyl)-1 ,4-benzohydroquinone-1-(f3-xyloside) (Compound 1).

Compound 1

DISCUSSION In those sexual systems of algae and other organisms where the female gametes are nonmotile, the production of pheromones, which would attract the motile sperm is of obvious significance. Even in algae where the gametes of both sexes are flagellated, a system of chemoattraction is useful if the gametes of one sex settle down and become nonmotile as a first step in the sexual process. One sees this in the brown algae such as Ectocarpus with its isogametes or Cutleria with its large flagellated female gametes. The same settling phenomenon followed by chemoattraction was also described by Pascher for the green flagellate Chloromonas (Chlamydomonas) paupera. Unlike these examples of the brown algae and Chloromonas, the female gametes in C. allensworthii described in this account do not settle down and become nonmotile but rather secrete the sexual pheromone while swimming through the medium. This is especially evident when female gametes are washed prior to introduction into a suspension of sperm. Such females continue to secrete the pheromone and set up gradients around themselves to which the sperm are attracted. The female gametes will often try to swim away with a trail of sperm following them. In contrast, if medium from a female culture is introduced along with the female gametes into a suspension of sperm, the excess pheromone may immobilize temporarily the sperm, and the only effect observed in this introduction is the formation of small jiggling groups of gametes not unlike the agglutination clumps one sees in the typical Chlamydomonas such as C. reinhardtii and C. eugametos.

Recent isolates from the Lemon Cove site (three species of Chlamydomonas and Provasoliella) exhibited sexual systems with large female gametes and small male gametes, both motile. Unfortunately, in all isolates both types of gametes were produced in the same clone (homothallism),

Proc. Natl. Acad. Sci. USA 92 (1995)

thereby removing the opportunity to carry out investigations as have been described in C. allensworthii. However, observation of the sexual process in such homothallic populations shows the obvious attraction of the male gametes to the female gametes. Thus, the chemoattraction of male gametes in heterogamous systems may well be the rule rather than the exception. Attempts to find other strains of Chlamydomonas showing this chemoattraction have already yielded isolates from Austin, TX, and Lismore, Australia, which are morphologically identical with the C. allensworthii from Lemon Cove, CA. The sperm from Lemon Cove are attracted to the females of Austin and Lismore strains, but no fusions between them can be observed. This is not surprising inasmuch as three different sets of males and female strains from Lemon Cove are also sexually incompatible even though the chemoattraction is mutual. In this latter instance, all three sets were isolated from a single sample of dried soil collected in the summer of 1992. Chemical signaling in protists and chemoattraction of motile cells has been suggested often but rarely demonstrated with certainty. Chemical communication between unicellular organisms has been the topic of a recent review (13). Spermatozoid chemotaxis by sugars and plant acids was the classical paradigm by which this phenomenon had been discovered in ferns and mosses by Wilhelm Pfeffer more than a century ago, and the chemotaxis of white blood cells by small peptides has been well studied. The cell biologically best studied cases are the chemotaxis of the male gametes of the water mold Allomyces by the carene derivative sirenin, emitted by the female gametes into the medium, and the cyclic adenylate emission of recruiting cells of the slime mold Dictyostelium to aggregate the slugs. Chemotaxis in algae has been reported at first in the phaeophycean mating process. In the course of these studies, a series of straight-chain or cyclic, multiply stereospecifically unsaturated hydrocarbons, biologically derived from linolenic acid by oxidizing and cyclizing fragmentations, have been identified. Some of them, particularly those with specific trigger functions in the release of the spermatozoids, have an epoxi (oxirane) functionalization. The compound reported here acting in C. allensworthii as lure belongs to an entirely different class of compounds, apparently derived from plastohydroquinone by incomplete oxidative degradation of its polyprenyl side chain and site specific glycosylation. From present studies, this derivatization seems to be essential for

biological functioning. In both cases, however, the ritualizing signal compound is produced from metabolic wastes of specific minor cell constituents by constructively reutilizing the less digestible moiety. The new species C. allensworthii is named in honor of and in memory of Captain John H. Allensworth, who was working with us on this research problem at the time of his death in July 1993 as the result of an accident during his annual training period in the Texas National Guard. The authors wish to thank Dr. V. L. Wray (Gesellschaft fur Biologische Forschung, Braunschweig) for determination of D-xylose as the 13-glycoside using 1H NMR, Dr. S. Waffenschmidt (Institut fur Biochemie, Universitat Koln) for the pentose and hexose standards used in the GC, and Dr. Richard Moe (University of California, Berkeley) for writing the Latin diagnosis. 1. Sapp, J. (1990) Where the Truth Lies, Franz Moewus and the Origins ofMolecularBiology (Cambridge Univ. Press, Cambridge, U.K.), p. 340. 2. Tsubo, Y. (1957) Bot. Mag. 70, 327-334. 3. Tsubo, Y. (1961) J. Protozool. 8, 114-121. 4. Maier, I. & Muller, D. G. (1986) Bio. Bull. 170, 145-175. 5. Harris, E. H. (1989) The Chlamydomonas Sourcebook (Academic, New York), p. 780. 6. Pascher, A. (1932) Jahrb. Wiss. Bot. 57, 551-580.

Plant Biology: Starr et al. 7. Ettl, H. (1970) Beih. Z. Nova Hedwigia 34, 1-283. 8. Starr, R. C. (1973) in Handbook of Phycological Methods, ed. Stein, J. R. (Cambridge Univ. Press, Cambridge, U.K.), Vol. 1, pp. 161-167. 9. Ettl, H. (1983) Chlorophyta I, Phytomonadina (Susswasserflora von Mitteleuropa, Bd. 9) (Gustav Fischer, Stuttgart), p. 807. 10. Moewus, F. (1934) Arch. Protistenkd. 83, 98-109.

Proc. Natl. Acad ScL USA 92 (1995)

645

11. Heimke, J. W. & Starr, R. C. (1979) Arch. Protistenkd. 122, 20-42. 12. Provasoli, L. & Pintner, I. J. (1959) in The Ecology ofAlgae, eds. Tryon, C. A. & Hartman, R. T. (Pymatuning Lab. of Field Biology, Univ. of Pittsburgh), Spec. Publ. No. 2, pp. 84-96. 13. Jaenicke, L. (1991) in Progress in Botany, eds. Behnke, H., Esser, K., Kubitzki, K., Runge, M. & Ziegler, H. (Springer, Berlin), Vol. 52, pp. 138-189.