rudall to semente evolucao

8/2/2019 Rudall to Semente Evolucao

1/13

1581

American Journal of Botany 96(9): 15811593. 2009.

Recent years have seen considerable progress in identifica-tion of the earliest lineages of flowering plants. Phylogeniesbased on molecular data indicate that although monocots andeudicots represent two entirely distinct major angiospermclades, a few taxa do not belong in either lineage and are either

early divergent on the angiosperm tree or placed in an unre-solved polytomy with monocots and eudicots (Fig. 1). Thoughrelatively species-poor, the early-divergent angiosperm lin-eages assume disproportional significance in comparative stud-ies of angiosperm evolution. In most analyses, the basalangiosperm grade (termed the ANA grade, formerly the ANITAgrade), consists of three lineages: Amborellaceae (one genus:Amborella), Nymphaeales (about eight genera: Barclaya,Bra-senia, Cabomba, Euryale, Nymphaea [including Ondinea],Nuphar, Trithuria, Victoria) and Austrobaileyales (five genera:Austrobaileya, Illicium, Kadsura, Schisandra, Trimenia s.l.).Most recently, Saarela et al. (2007) reassigned the small aquaticfamily Hydatellaceae to the waterlily clade, Nymphaeales, fromits former placement close to the grasses in the monocot order

Poales. Following detailed systematic study (Sokoloff et al.2008a), Hydatellaceae now consist of a single genus, Trithuriacontaining 12 species: 10 from Australia, one from India, andone from New Zealand.

Improved resolution of early-angiosperm relationships has

enhanced our understanding of key angiosperm features suchas the early evolution of the flower and double fertilization. Un-derstanding structural homologies and developmental pathwaysin early-divergent angiosperms is important in recognizing howthey evolved from nonflowering ancestors among the gymnosperms. Furthermore, embryological characters can be impor-tant for understanding systematics among plants in which thesporophyte is highly reduced, including Hydatellaceae. For ex-ample, embryological and ovule characters of Hydatellaceae(pendulous-anatropous ovule, presence of a starchy perispermand cellular endosperm development) were Hamanns (1976primary reasons for segregating them as a family distinct fromCentrolepidaceae. Embryological characters in early-divergenangiosperms have been the subject of numerous recent studies(e.g., Batygina et al., 1980, 1982; Battaglia, 1986; Shamrov and

Winter, 1991; Winter and Shamrov, 1991a, b; Shamrov, 1998Batygina and Vasilyeva, 2002; Williams and Friedman, 20022004; Friedman and Williams, 2003; Friedman et al., 2003Friedman, 2006, 2008; Tobe et al., 2007; Rudall et al., 2008Friedman and Ryerson, 2009; Williams, 2008, 2009). Friedmanand Williams (2003) hypothesized that a four-celled, four-nu-cleate (single-module) gametophyte represents the plesiomor-phic condition in angiosperms and gave rise to the morecommon seven-celled, eight-nucleate (double-module) condition that occurs in more than 80% of angiosperms (Palser1975). The four-nucleate condition characterizes all families ofNymphaeales (including Hydatellaceae) and Austrobaileyales

1 Manuscript received 28 January 2009; revision accepted 14 April 2009.The authors thank R. Bateman for critically reading the manuscript and

T. Macfarlane for help with fieldwork in Australia. F. Marone and M.Stampanoni (Tomcat Beamline, Swiss Light Source, Paul ScherrerInstitute) and S. Joomun (Royal Holloway, University of London) providedassistance with synchrotron x-ray tomographic microscopy, and the SwissLight Source and EU provided time and funding to work there. The researchwas partly supported by a 2007 CoSyst grant.

6 Author for correspondence (e-mail: [email protected])7 Present address: Museum of Paleontology, University of Michigan,

1109 Geddes Road, Ann Arbor, Michigan 48109 USA

doi:10.3732/ajb.0900033

SEEDFERTILIZATION, DEVELOPMENT, ANDGERMINATION

IN HYDATELLACEAE (NYMPHAEALES): IMPLICATIONSFOR

ENDOSPERMEVOLUTIONINEARLYANGIOSPERMS1

Paula J. Rudall,2,6Tilly Eldridge,2Julia Tratt,2Margaret M. Ramsay,2

Renee E. Tuckett,3Selena Y. Smith,4,7Margaret E. Collinson,4

Margarita V. Remizowa,5and Dmitry D. Sokoloff5

2Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AB, UK; 3The University of Western Australia, Crawley, WA 6009,Australia; 4Department of Earth Sciences, Royal Holloway University of London, Egham, Surrey, TW20 0EX, UK; and

5Department of Higher Plants, Biological Faculty, Moscow State University 119991, Moscow, Russia

New data on endosperm development in the early-divergent angiosperm Trithuria (Hydatellaceae) indicate that double fertiliza-tion results in formation of cellularized micropylar and unicellular chalazal domains with contrasting ontogenetic trajectories, asin waterlilies. The micropylar domain ultimately forms the cellular endosperm in the dispersed seed. The chalazal domain formsa single-celled haustorium with a large nucleus; this haustorium ultimately degenerates to form a space in the dispersed seed, simi-lar to the chalazal endosperm haustorium of waterlilies. The endosperm condition in Trithuria and waterlilies resembles the helo-bial condition that characterizes some monocots, but contrasts withAmborella andIllicium, in which most of the mature endospermis formed from the chalazal domain. The precise location of the primary endosperm nucleus governs the relative sizes of the cha-

lazal and micropylar domains, but not their subsequent developmental trajectories. The unusual tissue layer surrounding the bi-lobed cotyledonary sheath in seedlings of some species ofTrithuria is a belt of persistent endosperm, comparable with that ofsome other early-divergent angiosperms with a well-developed perisperm, such as Saururaceae and Piperaceae. The endosperm ofTrithuria is limited in size and storage capacity but relatively persistent.

Key words: angiosperm evolution; embryo; endosperm; Hydatellaceae; seed development; synchrotron; Trithuria;waterlilies.


2/13

1582 American Journal of Botany [Vol. 96

were carried out on T. submersa (HK). Seeds of Nymphaea nouchalii Burm.f. (Fig. 7G) were obtained from the Kew living collections, and seeds ofN. lotusL. (Fig. 10) were obtained from the spirit (fixed material) collection at K.

MethodsFor germination studies, seeds ofT. submersa were germinated inthe Conservation Biotechnology Unit at Kew. The seeds were sterilized with 0.5%SDICN (sodium dichloroisocyanurate) for 20 min, rinsed in distilled water, thensoaked in 5 ppm (14.4 M) gibberelic acid for 24 h. They were pricked out into

Petri dishes containing 1/2 Murashige and Skoog medium (Murashige and Skoog,1962) in a five by five grid, sealed with parafilm, and placed in an incubator at13C. Seeds were collected prior to soaking, then at the following intervals: 1, 2,7, 14, 21, 28, and 31 d after the initial sterilization of the seeds. Seeds collectedbefore 21 d were soaked in concentrated HCl for 2 h before fixation, to erode theseed coat and allow penetration of the fixative to structures within the seed.

Material for LM or SEM examination was fixed in FAA (40% formaldehydesolutionglacial acetic acid70% ethanol, 10 : 5 : 85) at room temperature for36 h, then transferred to 70% ethanol. Material for TEM examination was fixedin Karnovskys fixative (2% parafomaldehyde2.5% gluteraldehyde in 0.05 Mphosphate buffer, pH 7.2) at 4C for 24 h.

For light microscopy (LM), material was embedded in Histo-Technovit(Heraeus-Kulzer, Wehrheim, Germany) 7100 resin and sectioned using a Leica(Wetzlar, Germany) RM 2155 rotary microtome fitted with a tungstencarbideknife. Sections were stained in toluidine blue and mounted in DPX resin (amixture of distyrene, a plasticizer, and xylene). Optical sections were photo-graphed using a Leitz Diaplan photomicroscope fitted with a Zeiss (Gttingen,

Germany) Axiocam digital camera, in some cases using differential interferencecontrast microscopy (DIC). Some images were merged using Adobe (San Jose,California, USA) Photoshop. Drawings were made by sketching over photo-graphed images in Powerpoint (Microsoft, Redmond, Washington, USA).

For SEM, material was dissected in 70% ethanol, then dehydrated throughabsolute ethanol and critical-point dried using a Autosamdri-815B CPD (Tousi-mis Research, Rockville, Maryland, USA). Material was mounted onto speci-men stubs using double-sided tape, coated with platinum using an Emitech(Hailsham, UK) K550 sputter coater, and examined using a Hitachi (Tokyo,Japan) cold-field emission SEM S-4700-II at 1 kV.

For TEM, seeds fixed in Karnovskys fixative were transferred into 0.05 Mphosphate buffer, pH 7.2, further fixed in 1% osmium tetroxide in 0.05 M phos-phate buffer, dehydrated in an ethanol series, and embedded in LR Whiteacrylic resin (London Resin Co., Basingstoke, UK). The resin was polymerizedin a vacuum oven at 60C for 20 h. The seeds were sectioned using a Reichert-Jung Ultracut ultramicrotome fitted with glass knives. Semithin sections (0.5m) were stained with toluidine blue and mounted onto glass slides using DPX.

Ultrathin sections were mounted onto formvar-coated slot grids, stained usingan LKB Bromma 2168 Ultrostainer (Leica, Deerfield, Illinois, USA) with leadcitrate and uranyl acetate, examined using a JEOL (Tokyo, Japan) JEM-1210TEM, and photographed using a Kodak (Rochester, New York, USA) Megaview III soft-imaging system.

For synchrotron radiation x-ray microtomography (SRXTM), fruit sampleswere mounted onto brass stubs using polyvinyl acetate glue. Samples were im-aged at the TOMCAT beamline, Swiss Light Source, Paul Scherrer Institute,Villigen, Switzerland. Data were acquired using the 10 objectives on an x-raymicroscope, and an exposure time of 400420 ms at 9.9 keV. A total of 15012048 projections was acquired over 180. Projection data were processed, andcorrected sinograms were then used for segment reconstruction. Reconstructedimages were processed using Avizo 5.0 (Mercury Computer Systems, Chelms-ford, Massachusetts, USA).

RESULTS

PrefertilizationEmbryology before fertilization was de-scribed in detail by Rudall et al. (2008) and Friedman (2008).Mature embryo sacs ofT. submersa are illustrated here (Fig. 2)to show the micropylar position of the egg apparatus and thechalazal position of the central cell nucleus.

Stages before seed dispersalEarly development of the en-dosperm and embryo was observed for Trithuria cowieana (Figs.3, 4), and some stages were observed in T. australis, T. bibrac-teata, T. filamentosa, T. lanterna (Fig. 5), and T. submersa. InT. cowieana, very early postfertilization stages were seen in

in contrast to the monosporic, seven-celled, eight-nucleate con-dition that is very common in other angiosperms, and the some-what divergent eight-celled, nine-nucleate condition that occursinAmborella (Friedman, 2008; Friedman and Ryerson, 2009).

Following detailed comparative studies of flower, megag-ametophyte, and seedling development in Hydatellaceae andNymphaeaceae (Rudall et al., 2007, 2008, 2009; Sokoloff et al.,2008b), our aim in this paper is to continue our current morpho-logical characterization of Hydatellaceae by examining em-bryo, endosperm, and seed development, including seedgermination. In particular, we address whether endosperm de-

velopment is entirely cellular in Hydatellaceae, as suggested byHamann (1976) and Hamann et al. (1979), and whether an en-dosperm haustorium is formed, as in some waterlilies (Cook,1902, 1906, 1909; Seaton, 1908; Swamy and Parameswaran,1962; Schneider, 1978; Floyd and Friedman, 2001), and ex-plore whether this new information can shed light on the evolu-tionary origin of endosperm. A further goal is to examinewhether the unusual tissue layer surrounding the bilobed coty-ledonary sheath in some species ofTrithuria is a belt of persis-tent endosperm, as suggested by Sokoloff et al. (2008b), orrepresents an expanded tegmen layer (i.e., derived from the in-ner integument), as interpreted by Tillich et al. (2007).

MATERIALS AND METHODS

MaterialSeveral stages of early postfertilization development were ob-served in Trithuria cowieana D.D.Sokoloff, Remizowa, T.D.Macfarl. & Rudall(fixed material collected by Macfarlane et al., Northern Territory, Australia,2008), and some stages in T. lanterna (fixed material collected by Macfarlaneet al., Northern Territory, Australia, 2008), and Trithuria submersa Hook.f. HKindicates material grown in the Conservation Biotechnology Unit at the RoyalBotanic Gardens, Kew (K), from seeds collected by Tuckett at Mersa Road swamp,Western Australia 2006. Other species examined for comparison were T. lanternaD.A.Cooke (K: 47115; Dunlop 4740A, Northern Territory, Australia, 1978);Trithuria australis (Diels) D.D. Sokoloff, Remizowa, T.D.Macfarl. & Rudall(Macfarlane 3357 and Hearn, Western Australia, approx. 50 km E of Manjimup,1999; vouchers at NSW and PERTH); T. filamentosa Rodway (K: 28269;de Malahide s.n., Lake Dobson, Australia, 1966). Seed germination experiments

Fig. 1. Diagram showing relationships of early-divergent angiospermlineages, based on recent molecular analyses (e.g., Qiu et al., 2000; Soltiset al., 2000; Saarela et al., 2007).


3/13

1583September 2009] Rudall et al.Endosperm development in Hydatellaceae

clearly cellularized at later stages (Figs. 5A, C, D), suggesting thadivisions are cellular from the onset in the micropylar domain.

Dispersed seedsIn dispersed seeds of T. submersa (Fig6AD), the embryo consists of a single-celled suspensor and asmall, globular proembryo that is flattened at one pole, but withno differentiation of tissues or development of cotyledons. Theembryo is surrounded by cellular endosperm. At this stagethere is normally a space between the endosperm and perispermin all species (Figs. 6, 7), though the endosperm maintains con-

tact with the perisperm at the edge of the seed.Perisperm is formed before fertilization and occupies more

than 80% of the seed capacity, as also noted by previous au-thors (Hamann, 1976; Friedman, 2008; Rudall et al., 2008)Perisperm is composed of large cells packed with starch grainsApart from the peripheral layer, most perisperm cells are ar-ranged in groups; within each group, the cell walls start to breakdown and the nuclei clump together (Fig. 5F, G). Hydatellaceaeshare this character (perisperm) in common with other Nymphaeales (Khanna, 1964, 1965; Schneider, 1978, 1983; Schneiderand Ford, 1978; Rudall et al., 2008), other early-divergent angiosperms (e.g., Piperaceae: Johnson, 1900, 1902; Saururaceae

reproductive units (RUs) whose stalks were still extending(Fig. 3G); subsequently, RUs stalks extend farther to bear fruitsabove the tips of the leaves. Pollen tubes were observed penetrat-ing the micropyle and extending into the embryo sac in T. cow-ieana (Fig. 3A, C, D). Soon after fertilization, the innermost layerof the inner integument becomes rapidly thickened in the micro-pylar region to form a tegmen (compare Fig. 3C with Fig. 3D ).The zygote (Figs. 3AD, 5E), initially forms two prominentnucleoli, then divides to form a two-celled embryo, each cellpossessing two prominent nucleoli (Figs. 4, 5A, B). The proem-

bryo sometimes possesses a clear suspensor cell (Fig. 5C, D).The primary endosperm nucleus (Fig. 3A) lies in the center

of the embryo sac and possesses two nucleoli. It undergoes mito-sis and a transverse cell wall is formed, resulting in two cleardomainschalazal and micropylar. The chalazal haustorial do-main contains a single large nucleus (Fig. 3BF) that grows in sizeuntil it is extremely large (Fig. 4); the haustorial domain eventu-ally degenerates to form a space (Figs. 6, 7). In the micropylardomain, the single initial nucleus (Fig. 3C) subsequently under-goes several mitoses. At least, the first division of the micropylardomain is cellular (Figs. 3B, DF, 5E); cell walls are subsequentlysometimes difficult to distinguish (Fig. 4), but endosperm is

Fig. 2. Trithuria submersa, unfertilized four-nucleate embryo sacs. c = central cell, ea = egg apparatus (nuclei not visible), eg = egg cell, m = micropyle, st = stigmatic hair, sy = synergid. Scale bars = 20 m in A, 10 m in BD.


4/13



5/13


Fig. 3. Trithuria cowieana, embryo sacs at earliest postfertilization stages. (A) Pollen tube present, with undivided primary endosperm nucleus andzygote. Red square outlines region where an optical section in a slight different focal plane has been overlaid to show pollen tube. (B) Micropylarendosperm cell has divided again to form two cells; chalazal (haustorial) nucleus remains undivided. (C) Penetrated pollen tube still present in micropyleand tegmen not formed. The embryo sac could be fertilized, in which case the endosperm has undergone one cellular division into chalazal and micropyladomains. However, the tegmen is not yet formed, as in Fig. 1A, so another possible interpretation of this image is that this is an unfertilized double embryo sac of the type that is commonly found in Trithuria (described by Rudall et al., 2008), in which case the lowermost nucleus is another partly formedunfertilized embryo sac, rather than a chalazal endosperm haustorium. (D) Pollen tube still present in embryo sac, micropyle compressed and tegmen forming; micropylar endosperm cell has divided again to form two cells; chalazal (haustorial) nucleus remains undivided. (E, F) Stage slightly later than in Dzygote (proembryo) with two prominent nucleoli. (G) Longitudinal section of top of a single plant showing three reproductive units (RUs) at different de-velopmental stages (youngest at bottom left); the uppermost ovule in the largest RU (arrow) is recently fertilized. e = endosperm nucleus in micropylardomain, em = embryo, hn = chalazal (haustorial) nucleus, p = pollen tube, pe = primary endosperm nucleus, ttl = thickened tegmen layer, z = zygote. Scalebars = 10 m in AD, F, 20 m in E, 100 m in G.

Fig. 4. (AD) Trithuria cowieana, early postfertilization with diagrams. e = endosperm nucleus in micropylar domain, em = embryo, hn = chalaza(haustorial) nucleus. In diagrams, embryo is in red, micropylar endosperm domain is in yellow, chalazal (haustorial) endosperm domain is in green. Scalebars = 10 m.


6/13


1980). At the chalazal pole, there is a group of densely packed,thick-walled cells.

Seed sowing and germinationSeven days after sowing, theoperculum remains intact, and the seed has not enlarged, thoughthe embryo has grown slightly (Fig. 8A). After 14 d, the embryo

Seaton, 1908), and some monocots (Rudall, 1997), though theperisperm ofAcorus, the putative sister to all other monocots, isnonhomologous (Rudall and Furness, 1997).

At the micropylar end of the seed is a conspicuously thick-ened layer of tegmen (inner integument) cells beneath a conicaltestal structure (the operculum), as in waterliles (e.g., Collinson,

Fig. 5. (AE) Trithuria lanterna, (E) zygote and (AD) early proembryo development. (F, G) T. cowieana, perisperm, showing clumped groups ofnuclei. e = endosperm nucleus in micropylar domain, em = embryo, h = haustorium nucleus, z = zygote. Scale bars = 10 m.


7/13


Fig. 6. Trithuria submersa, sown, dispersed seeds at (AE) 7 days after sowing, (F, G) 14 d after sowing, (H) 21 d after sowing. e = endospermem = embryo, h = haustorial space, o = operculum, p = perisperm, ttl = thickened tegmen layer. Scale bars = 20 m in A, D, E; 50 m in B, C, FH.


8/13


free sperm nuclei were observed, and synergids (or degeneratedsynergids) are at best difficult to recognize in Trithuria becauseof their small size (Friedman, 2008; Rudall et al., 2008). How-ever, the primary endosperm nucleus (Fig. 3A) is relativelylarge and bears two nucleoli, consistent with a double set ofchromosomes (Williams, 2009), the expected (diploid) condi-tion after fusion of a haploid sperm with a haploid central cellnucleus. Diploid endosperm is the common condition in specieswith a four-nucleate embryo sac, including Nymphaeales andAustrobaileyales (e.g., Williams and Friedman, 2002), and afour-nucleate embryo sac has been observed in several speciesof Trithuria (Rudall et al., 2008). Triploid endosperm is themore common condition in angiosperms, but this condition re-sults from fusion between a sperm and a diploid fusion nucleus(or two polar nuclei) of a seven-celled embryo sac.

Embryo and seed developmentThere are clear similaritiesbetween mature seeds ofTrithuria submersa and those of some

has clearly proliferated, causing it to rupture the seed coat anddisplace the operculum (Figs. 6F, G, 8B, C); the thickened layerof the tegmen protrudes like a ruff (Fig. 8C). The space resultingfrom degeneration of the chalazal haustorium between theperisperm and the endosperm is reduced in size (Fig. 6F, G) andis eventually obliterated (Fig. 6H) by embryo enlargement. After21 d, the operculum becomes detached from the seed, and theembryo emerges from the micropylar end (Figs. 6H, 8C). From28 d after sowing, there is clear polarization of the root and shoottissues (Figs. 8D, 9AC). The seedling haustorium interfaceswith the perisperm via a distinct transfer layer, but a small amountof persistent endosperm tissue remains (Fig. 9D).

DISCUSSION

Double fertilizationOur observations of pollen tubes inthe embryo sac ofT. cowieana (Fig. 3A, C, D) are a good indi-cator that double fertilization occurs in Trithuria. No obvious

Fig. 7. Synchrotron radiation x-ray microtomographic (SRXTM) images (A, C, EG, longitudinal sections; B, D, transverse sections). (A, B) Trithuriasubmersa. (C, D) T. occidentalis. (E, F) T. bibractaea. (G)Nymphaea nouchalii. e = endosperm, em = embryo; f = fruit wall, hs = haustorial space; o = operculum,p = perisperm, s = seed coat, ttl = thickened tegmen layer.


9/13


rather similar to some of our images of T. lanterna (Fig. 5AB). However, it seems likely that the chalazal haustorium ispresent in all species but is sometimes inconspicuous and dif-ficult to detect at early stages; we found a large (probably haustorial) nucleus in a few specimens ofT. lanterna (Fig. 5E). Asingle cell located at the chalazal pole in T. inconspicua (Friedman, 2008) could be an inconspicuous chalazal haustoriumThe presence of a space between the endosperm and perisperm

in dispersed seeds of several species of Trithuria examinedhere (both in histological sections and x-ray optical sections ofnonfixed dispersed viable seeds: Figs. 6, 7) indicates a haustorium. The shape and location of the space conforms to the cha-lazal haustorium, which presumably degenerates at an earlystage. SRXTM images show this space in species ofTrithuriaandNymphaea (Fig. 7), and in seeds of other species with anendosperm haustorium, such as Saururus (S. Y. Smith, unpublished data). Thus, the space is potentially useful as a finger-print for haustorial endosperm, though in dispersed seeds ofNymphaea the haustorium is highly variable in shape andsometimes absent (Fig. 10); in many species it extends into theperisperm (e.g., Valtzeva and Savich, 1965; Shamrov, 1998).

Many (possibly all) waterlilies possess a chalazal endospermhaustorium, includingBrasenia, Cabomba,Nymphaea (includ

ing Castalia), and Nuphar(Cook, 1906, 1909; Khanna, 1965Valtzeva and Savich, 1965; Batygina et al., 1982; Shamrov1998; Floyd and Friedman, 2000, 2001). In waterlilies, the pri-mary endosperm mitosis is transverse and divides the en-dosperm into micropylar and chalazal domains. The micropyladomain undergoes division and becomes cellularized, and thechalazal domain remains undivided and acts as a haustoriumsometimes subsequently extending into the maternally derivedperisperm tissue (e.g., inBrasenia: Khanna, 1965;NymphaeaValtzeva and Savich, 1965; Cabomba: Floyd and Friedman2000). Thus, in all Nymphaeales, the mature cellularized en-dosperm is formed entirely from the micropylar domain.

waterlilies, including the presence of a well-developed tegmenat the micropylar end of the seed, a highly characteristic opercu-lum, a large starchy perisperm that occupies most of the seedvolume, and a small endosperm that persists around the embryoafter germination (e.g., Schneider, 1978; Schneider and Ford,1978; Collinson, 1980; present study). This combination offeaturesfour-nucleate embryo sac, operculate seeds withperisperm and haustorial endospermis probably unique to the

waterlily clade (Nymphaeales, including Hydatellaceae) amongangiosperms. These features are therefore potentially a usefulidentification tool for well-preserved fossil material, though ad-mittedly the embryo sac is rarely preserved, even in permineral-ized fossil material where perisperm is documented(Cevallos-Ferriz and Stockey, 1989). The minimal degree ofembryo development in the dispersed seed of Trithuria resem-bles the condition in some waterlilies (Fig. 10), though manyother waterlilies, such as Nuphar luteum, Nymphaea spp. (in-cluding Ondinea), Euryale ferox, and Barclaya longifolia pos-sess a well-differentiated embryo in the dispersed seed (e.g.,Meyer, 1960; Valtzeva and Savich, 1965; Schneider, 1978;Schneider and Ford, 1978; Shamrov, 1998), indicating some varia-tion within the Nymphaeales in timing of developmental stages.

A chalazal endosperm haustorium occurs in Hydatel-laceaeWe have observed a chalazal haustorium at earlystages in numerous specimens ofT. cowieana (Figs. 3, 4). Thiscondition compares closely with the condition reported in otherNymphaeales, which possess a chalazal haustorium (discussedlater). Previous studies on the apomictic New Zealand speciesT. inconspicua (formerly Hydatella inconspicua) (Hamannet al., 1979; Friedman, 2008) and the Indian species, Trithuriakonkanensis (Gaikwad and Yadav, 2003) did not report a haus-torium. Sections ofT. inconspicua illustrated by Hamann et al.(1979) and Friedman (2008) show a few-celled proembryo sur-rounded by some cellular endosperm adjacent to perisperm,

Fig. 8. Trithuria submersa, SEM sown seeds. (A) Dispersed seed after 24-h soak in gibberelic acid, (B) 14 d after sowing, (C) 21 d after sowing, (D)28 d after sowing. op = operculum. Scale bars = 50 m.


10/13


Fig. 9. Trithuria submersa, longitudinal sections of germinated seeds, 28 d after sowing (AC, LM; D, TEM). (A) End view (with seedling axis incross section), (B, C) lateral views (with seedling axis in longitudinal section), (D) lateral view with detail inset. cs = cotyledonary sheath, e = endosperm,fp = first leaf of seedling plumule; p = perisperm, r = primary root, rttl = remains of thickened tegmen layer, sh = seedling haustorium, tl = transfer layer ofseedling haustorium. Scale bars = 50 m.


11/13


senia) vs. cellularized early divisions in other waterlilies. Ourobservation of the cellularized condition in Trithuria suggeststhat the condition in Cabombaceae represents the loss of earlymicropylar cellularization.

The distinction between cellular and helobial endosperm is often unclear, making it difficult to source comparative information, especially in Nymphaeales. Swamy and Parameswaran

(1962) narrowly redefined the helobial type as representing onlycases in monocots in which the micropylar cell is the larger of thetwo daughter cells that result from the primary unequal en-dosperm division, and the smaller chalazal cell often remainsnoncellularized (coenocytic) and haustorial. Both cellular andhelobial types have been reported in Nymphaea (Cook, 19021906, 1909; Seaton, 1908). However, the helobial records werereinterpreted as cellular by Swamy and Parameswaran (1962)because either the first division is more or less equal, or the cha-lazal (rather than micropylar) cell is the larger of the two. Despitethis, Valtzeva and Savich (1965) concluded that endosperm de-velopment inNymphaea resembles the helobial type. In both cel-lular and helobial types, the primary mitosis is coupled withcytokinesis, but the resulting micropylar and chalazal daughtercells produce domains with differing morphology and develop-mental trajectories. The early endosperm condition in Trithuriaand waterlilies closely resembles the helobial condition that characterizes many monocots. This condition contrasts with that ofAmborella and Illicium, in which the primary partitioning celwall is oblique, and all or most of the mature endosperm is formedfrom the chalazal domain (Floyd and Friedman, 2000).

The precise location of the primary endosperm nucleus gov-erns the relative sizes of the chalazal and micropylar domainsbut not their subsequent developmental trajectories. Vijayraghavan and Prabhakar (1984) suggested that in monocotswith helobial endosperm, the chalazal domain is often smallerbecause of the location of the primary endosperm nucleus closeto the antipodals. The central cell nucleus in prefertilized em-bryo sacs is chalazal (Fig. 2). The primary endosperm nucleus

in T. cowieana lies in an almost central position or slightlycloser to the antipodals (Fig. 3A), and consequently the primarychalazal cell is only slightly smaller than the micropylar one(Fig. 3C). Thus, there is a clear correlation between locationand size.

However, there is no clear correlation between location andsubsequent cellularization. InAmborella, the primary endospermnucleus is located in the chalazal region, close to the antipodalsand there is no haustorium, both domains ultimately becomingcellularized (Floyd and Friedman, 2001). The role of the chalazahaustorium in Nymphaeales may be taken over by the antipodalsin some lineages with an eight-nucleate embryo sac, thus releas-ing the chalazal endosperm domain for other roles. The functionof the antipodals is obscure; postulated roles include a haustoriaone (to transfer nutrients from the nucellus/perisperm) or to se

crete growth substances that regulate endosperm developmen(Willemse and Van Went, 1984). If the four-nucleate embryo sacin Nymphaeales (lacking antipodals) is the plesiomorphic condition in extant angiosperms, as plausibly proposed by Friedmanand Williams (2003), then the same could be true for a chalazalendosperm haustorium, perhaps associated with a perispermMore comparative data are needed on early endosperm develop-ment in Austrobaileyales, all of which apparently lack antipodals(e.g., Tobe et al., 2007). However, Illicium (Austrobaileyalesdoes not fit this scenario; a haustorium is absent, and endospermcellularization is complex and not restricted to one domain(Floyd and Friedman, 2001). Similarly, a different set of

The chalazal haustorial domain in Trithuria cowieana containsa single large nucleus that is probably polytene (Fig. 4). The largesize of this nucleus probably results from chromosome endoredu-

plication, which is perhaps the commonest form of endospermpolyploidy, producing giant (polytene) nuclei filled with chromo-somes in a prophase-like condition (DAmato, 1984). Althoughit does not extend into the perisperm like some waterlily hausto-ria, it could function similarly to aid diffusion of nutrients fromthe perisperm to the endosperm and embryo. Another possiblerole is regulation of early endosperm development.

Endosperm typology and diversificationEndosperm is tra-ditionally classified into three types: nuclear, cellular, and helo-bial (Vijayraghavan and Prabhakar, 1984), but as many authorshave shown (e.g., Floyd and Friedman, 2000, 2001), this typol-ogy is unsatisfactory because it masks a complex suite ofcharacters.

The nuclear endosperm type has evolved several times in an-

giosperms, including more than once in early-divergent angio-sperms (e.g., in some Piperaceae: Lei et al., 2002). Loss ofcellularization is probably generated by activation of a programfor cell-cycle arrest in dividing nuclei (Raghavan, 2006). In an-giosperms with nuclear endosperm, including the archetypalmodel organismArabidopsis (e.g., Brown et al., 1999), the en-dosperm is a noncellularized coenocyte during early develop-mental stages, becoming cellularized later in development.

Onset of cellularization can also differ between the micropy-lar and chalazal domains. Floyd and Friedman (2001) high-lighted the distinction between noncellularized early divisionsin the micropylar domain in Cabombaceae (Cabomba andBra-

Fig. 10. Nymphaea lotus, longitudinal section of seed (haustorium ab-sent). e = endosperm, em = embryo, p = perisperm, ttl = thickened tegmenlayer. Scale bar = 100 m.


12/13


constraints must operate in monocots, some of which possessboth a chalazal endosperm haustorium and antipodals (Swamyand Parameswaran, 1962), though Acorus (sister to all othermonocots in most recent phylogenies) possesses large antipodals(Rudall and Furness, 1997) and lacks a chalazal endosperm haus-torium (Floyd and Friedman, 2000).

Persistence and role of cellularized endospermOur re-sults demonstrate that the unusual tissue layer surrounding thebilobed cotyledonary sheath in seedlings of some species ofTrithuria (including T. submersa) is a belt of persistent en-dosperm, as suggested by Sokoloff et al. (2008b), and not anexpanded tegmen layer (i.e., derived from the inner integument)as suggested by Tillich et al. (2007). Interestingly, a closelycomparable (possibly homologous) layer of endosperm thatpersists around the seedling also occurs in some other early-di-vergent angiosperms with a well-developed perisperm, such asSaururaceae and Piperaceae (Takhtajan, 1988). Thus, the en-dosperm ofTrithuria, though limited in size and storage capac-ity, is relatively persistent. Its primary role could be as aregulatory tissue to transfer nutrients from the perisperm to the

embryo, though this role is ultimately taken over by the seed-ling haustorium.

Endosperm originEndosperm is a key feature of angio-sperms, yet its homologies and evolutionary origin remainenigmatic. Assessments of endosperm evolution have sufferedfrom its heavily typological classification and from the difficul-ties in homologizing endosperm with structures in gymno-sperms. There are two main hypotheses for the evolutionaryorigin of endosperm (discussed by Favre-Duchartre, 1984;Friedman, 2001; Baroux et al., 2002): (1) as the cellular-growthphase of the female gametophyte (heterochronically delayeduntil after fertilization) or (2) as a monstrous proembryo thatfails to develop into a plant. The latter (proembryo) hypothesis

is currently widely preferred, but both hypotheses require fur-ther review. Our observations of a haustorium in Trithuria, to-gether with records of similar structures in other ANA-gradeangiosperms, could support the proembryo hypothesis. One im-portant similarity between the angiosperm embryo and en-dosperm is that, in both cases, the primary mitosis often resultsin a micropylar and chalazal cell that ultimately pattern the re-sulting structures. In the embryo, divisions of the chalazal cellproduce most of the embryo. The micropylar cell gives rise tothe stalk (suspensor) that attaches it to the seed coat, or some-times to both the suspensor and a portion of the embryo proper(Natesh and Rau, 1984). As discussed earlier, a chalazal en-dosperm haustorium, perhaps associated with a perisperm,could represent the plesiomorphic condition in angiosperms,though this requires further testing. If so, a likely role of the

chalazal haustorium, in addition to facilitating transfer of nutri-ents from the perisperm to the embryo, is to regulate early en-dosperm development of the micropylar cellularized region. Inturn, the micropylar cellularized region, which is relativelysmall but persistent, may act as a nutrient transfer layer fromthe perisperm to the embryo, until this role is taken over by theepidermis of the seedling haustorium itself.

LITERATURE CITED

Baroux,C.,C.Spillane, andU.Grossniklaus. 2002. Evolutionary ori-gins of the endosperm in flowering plants. Genome Biology 3: 15.

Battaglia,E. 1986. Embryological questions: 7. Do new types of em-bryo sac occur in Schisandra?Annals of Botany 44: 6992.

Batygina,T. B.,T. I.Kravtsova, andI. I.Shamrov. 1980. The compar-ative embryology of some representatives of the orders Nymphaealesand Nelumbonales.Botanicheskii Zhurnal 65: 10711086.

Batygina,T. B.,I. I.Shamrov, andG. E.Kolesova. 1982. Embryologyof the Nymphaeales and Nelumbonales. II. The developmentof the female embryonic structures. Botanicheskii Zhurnal 67:

11791195.Batygina,T. B., andV. E.Vasilyeva. 2002. Oenothera-type of embryo

sac development. In T. B. Batygina [ed.], Embryology of floweringplants: Terminology and concepts. I. Generative organs of flower,173176. Science Publishers, Plymouth, Massachusetts, USA.

Brown, R. C., B. E. Lemmon, H. Nguyen, and O. A. Olsen.1999. Development of endosperm in Arabidopsis thaliana. SexualPlant Reproduction 12: 3242.

Cevallos Ferriz,S. R. S., andR. A.Stockey. 1989. Permineralizedfruits and seeds from the Princeton Chert (Middle Eocene) of BritishColumbia: Nymphaeaceae.Botanical Gazette150: 207217.

Collinson,M. E. 1980. Recent and Tertiary seeds of the Nymphaeaceaesensu lato with a revision of Brasenia ovula (Brong.) Reid andChandler.Annals of Botany 46: 603632.

Cook, M. T. 1902. Development of the embryo-sac and embryo ofCastalia odorata and Nymphaea advena. Bulletin of the Torrey

Botanical Club 29: 211220.Cook, M. T. 1906. The embryogeny of some Cuban Nymphaeaceae.

Botanical Gazette 42: 376392.Cook, M. T. 1909. Notes on the embryology of the Nymphaeaceae.

Botanical Gazette 48: 5660.DAmato,F. 1984. Role of polyploidy in reproductive organs and tissues.

In B. M. Johri [ed.], Embryology of angiosperms, 519566. Springer-Verlag, Berlin, Germany.

Favre Duchartre,M. 1984. Homologies and phylogeny.In B. M. Johri[ed.], Embryology of angiosperms, 697734. Springer-Verlag, Berlin,Germany.

Floyd,S. K., andW. E.Friedman. 2000. Evolution of endosperm de-velopmental patterns among basal flowering plants. International

Journal of Plant Sciences 161: S57S81.Floyd, S. K., and W. E. Friedman. 2001. Developmental evolu-

tion of endosperm in basal angiosperms: Evidence from Amborella

(Amborellaceae),Nuphar(Nymphaeaceae), andIllicium (Illiciaceae).Plant Systematics and Evolution 228: 153169.

Friedman,W. E. 2001. Developmental and evolutionary hypotheses forthe origin of double fertilization and endosperm. Comptes Rendus du

Academie Science, Paris 324: 559567.Friedman,W. E. 2006. Embryological evidence for developmental labil-

ity during early angiosperm evolution.Nature 441: 337340.Friedman,W. E. 2008. Hydatellaceae are water lilies with gymnosper-

mous tendencies.Nature 453: 9497.Friedman, W. E., W. N. Gallup, and J. H. Williams. 2003. Female

gametophyte development in Kadsura: Implications for Schisandraceae,Austrobaileyales, and the early evolution of flowering plants.International

Journal of Plant Sciences 164 (Supplement): S293S305.Friedman,W. E., andK. C.Ryerson. 2009. Reconstructing the ances-

tral female gametophyte of angiosperms: Insights from Amborellaand other ancient lineages of flowering plants.American Journal of

Botany 96: 129143.Friedman,W. E., andJ. H.Williams. 2003. Modularity of the angio-

sperm female gametophyte and its bearing on the early evolution ofendosperm in flowering plants.Evolution 57: 216230.

Gaikwad,S. P., andS. R.YadavSr. 2003. Further morphotaxonomicalcontribution to the understanding of family Hydatellaceae.Journal ofthe Swamy Botanical Club 20: 110.

Hamann,U. 1976. HydatellaceaeA new family of Monocotyledoneae.New Zealand Journal of Botany 14: 193196.

Hamann, U., K. Kaplan, and T. Rbsamen. 1979. ber die Samen-schalenstruktur der Hydatellaceae (Monocotyledoneae) und diesystematische Stellung von Hydatella filamentosa. Botanische

Jahrbcher100: 555563.


13/13


Johnson,D. S. 1900. On the endosperm and embryo ofPeperomia pel-lucida.Botanical Gazette 30: 111.

Johnson,D. S. 1902. On the development of certain Piperaceae.BotanicalGazette 34: 321340.

Khanna, P. 1964. Morphological and embryological studies inNymphaeaceae. I. Euryale ferox Salisb. Proceedings of the Indian

Academy of Sciences 59B: 237243.Khanna, P. 1965. Morphological and embryological studies in

Nymphaeaceae. II.Brasenia schreberei Gmel. andNelumbo nuciferaGaertn.Australian Journal of Botany 13: 379387.

Lei,L. G.,Z. Y.Wu, andH. X.Liang. 2002. Embryology ofZippeliabegoniaefolia (Piperaceae) and its systematic relationships.Botanical

Journal of the Linnean Society 140: 4964.Meyer,K. I. 1960. On the embryology ofNuphar luteum Sm.Bulletin of

Moscow Society of Naturalists.Biological Series 65: 4860.Murashige,T., andF.Skoog. 1962. A revised medium for rapid growth

and bioassays with tobacco tissue cultures. Physiologia Plantarum15: 473497.

Natesh,S., andM. A.Rau. 1984. The embryo.In B. M. Johri [ed.],Embryology of angiosperms, 377444. Springer-Verlag, Berlin,Germany.

Palser,B. F. 1975. The bases of angiosperm phylogeny: Embryology.Annals of the Missouri Botanical Garden 62: 621646.

Qiu,Y. L.,J.Lee,F.Bernasconi-Quadroni,D. E.Soltis,P. S.Soltis,M.

Zanis,E. A.Zimmer,et al. 2000. Phylogeny of basal angiosperms:Analyses of five genes from three genomes.International Journal ofPlant Sciences 161 (6 Supplement): S3S27.

Raghavan,V. 2006. Double fertilization: Embryo and endosperm devel-opment in flowering plants. Springer, Berlin, Germany.

Rudall, P. J. 1997. The nucellus and chalaza in monocotyledons:Structure and systematics.Botanical Review 63: 140184.

Rudall, P. J., and C. A. Furness. 1997. Systematics ofAcorus:Ovule and anther. International Journal of Plant Sciences 158:640651.

Rudall,P. J.,M. V.Remizowa,A.Beer,E.Bradshaw,D. W.Stevenson,T. D.Macfarlane,R. E. et al. 2008. Comparative ovule and mega-gametophyte development in Hydatellaceae and water lilies reveal amosaic of features among the earliest angiosperms. Annals of Botany101: 941956.

Rudall, P. J., M. V. Remizowa, G. Prenner, C. A. Prychid, R.

Tuckett, and D. D. Sokoloff. 2009. Non-flowers near the baseof extant angiosperms? Spatiotemporal arrangement of organs in re-productive units of Hydatellaceae, and its bearing on the origin of theflower.American Journal of Botany 96: 6782.

Rudall,P. J.,D. D.Sokoloff,M. V.Remizowa,J. G.Conran,J. I.Davis,T. D.Macfarlane, andD. W.Stevenson. 2007. Morphology ofHydatellaceae, an anomalous aquatic family recently recognized asan early-divergent angiosperm lineage. American Journal of Botany94: 10731092.

Saarela,J. M.,H. S.Rai,J. A.Doyle,P. K.Endress,S.Mathews,A. D.Marchant,B. G.Briggs, andS. W.Graham. 2007. A newbranch emerges near the root of angiosperm phylogeny.Nature 446:312315.

Schneider,E. L. 1978. Morphological studies on the Nymphaeaceae. IX.The seed ofBarclaya longifolia Wall.Botanical Gazette 139: 223230.

Schneider,E. L. 1983. Gross morphology and floral biology ofOndinea

purpurea den Hartog.Australian Journal of Botany 31: 371382.Schneider, E. L., and E. G. Ford. 1978. Morphological studies on

the Nymphaeaceae. X. The seed of Ondinea purpurea den Hartog.Bulletin of the Torrey Botanical Club 105: 192200.

Seaton,S. 1908. The development of the embryo-sac ofNymphaea advena.Bulletin of the Torrey Botanical Club 35: 283290.

Shamrov, I. I. 1998. Formation of hypostase, podium and postamenin the ovule of Nuphar lutea (Nymphaeaceae) and Ribes aureum(Grossulariaceae).Botanicheskii Zhurnal 83: 314.

Shamrov, I. I., and A. N. Winter. 1991. Ovule development in representatives of the genera Nymphaea and Victoria (Nymphaeaceae)

Botanicheskii Zhurnal 76: 10721083.

Sokoloff, D. D., M. V. Remizowa, T. D. Macfarlane, and P. JRudall. 2008a. Classification of the early-divergent angiospermfamily Hydatellaceae: One genus instead of two, four new speciesand sexual dimorphism in dioecious taxa. Taxon 57: 179200.

Sokoloff, D. D., M. V. Remizowa, T. D. Macfarlane, R. ETuckett, M. M. Ramsay, A. S. Beer, S. R. Yadav, and P. JRudall. 2008b. Seedling diversity in Hydatellaceae: Implicationfor the evolution of angiosperm cotyledons.Annals of Botany 101153164.

Soltis,D. E.,P. S.Soltis,M. W.Chase,M. E.Mort,D. C.AlbachM. Zanis, V. Savolainen, et al. 2000. Angiosperm phylogenyinferred from 18S rDNA, rbcL and atpB sequences.Botanical Journaof the Linnean Society 133: 381461.

Swamy, B. G. L., and N. Parameswaran. 1962. The helobial endosperm. Biological Reviews of the Cambridge PhilosophicaSociety 38: 150.

Takhtajan,A. L. [ed.] 1988. Anatomia seminum comparativa, vol. 2Nauka, Leningrad, CCCP.

Tillich,H. J.,R. E.Tuckett, andE.Facher. 2007. Do Hydatellaceaebelong to the monocotyledons or basal angiosperms? Evidence fromseedling morphology. Willdenowia 37: 399406.

Tobe, H., Y. Kimoto, and N. Prakash. 2007. Development andstructure of the female gametophyte in Austrobaileya scanden(Austrobaileyaceae).Journal of Plant Research 120: 431436.

Valtzeva,O. V., andE. I.Savich. 1965. Development of embryo in Nymphaea candida Presl. and N. tetragona Georgi. BotanicheskiZhurnal 50: 13231326.

Vijayraghavan,M. R., andK.Prabhakar. 1984. The endosperm.InB. M. Johri [ed.], Embryology of angiosperms, 319376. SpringerVerlag, Berlin, Germany.

Willemse,M. T. M., andJ. L.Van Went. 1984. The female gametophyte. In B. M. Johri, [ed.], Embryology of angiosperms, 159196

Springer-Verlag, Berlin, Germany.Williams,J. H. 2008. Novelties of the flowering plant pollen tube un

derlie diversification of a key life history stage. Proceedings of theNational Academy of Sciences, USA 105: 1125911263.

Williams, J. H. 2009. Amborella trichopoda (Amborellaceae) and theevolutionary developmental origins of the angiosperm progamicphase.American Journal of Botany 96: 144165.

Williams, J. H., and W. E.Friedman. 2002. Identification of diploidendosperm in an early angiosperm lineage.Nature 415: 522526.

Williams, J. H., and W. E. Friedman. 2004. The four-celled femalegametophyte ofIllicium (Illiciaceae; Austrobaileyales): Implicationfor understanding the origin and early evolution of monocots, eumagnoliids, and eudicots.American Journal of Botany 91: 332351.

Winter,A. N., andI. I.Shamrov. 1991a. The development of the ovuleand embryo sac in Nuphar lutea (Nymphaeaceae). Botanicheski

Zhurnal 76: 378390.

Winter, A. N., and I. I. Shamrov. 1991b. Megasporogenesiand embryo sac development in representatives of the generaNymphaea and Victoria (Nymphaeaceae). Botanicheskii Zhurna76: 17161728.

rudall to semente evolucao

Documents