sign in sign up
<<
Home , Cell fusion, Egg cell, Gamete

Proc. Nati. Acad. Sci. USA Vol. 82, pp. 6129-6132, September 1985 Preferential fertilization in Plumbago: Ultrastructural evidence for -level recognition in an angiosperm (cytoplasnic heritable organelles/male gamete/sexual ) SCOTT D. RUSSELL Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019 Communicated by John G. Torrey, May 13, 1985

ABSTRACT Gametic fusion patterns in the angiosperm can be followed with transmission electron microscopy. The Plumbago zeylanica were determined by using cytoplasmically present report uses the dimorphic distribution of dimorphic sperm cells differing in and plastid plastids to trace the pattern of fertilization in this , content and then identifying paternal organelles through their demonstrating that such fusion is preferential. ultrastructural characteristics within the maternal at the time of fertilization. The virtual absence of plastids MATERIALS AND METHODS within the sperm that is physically associated with the vegetative nucleus allows paternal plastids to be used to trace Specimen Preparation. of P. zeylanica L. (Palmen- the fate of the two male after fusion. Such paternal garten, Frankfort, FRG) were grown at 18-230C with 16-hr plastids were present in the in >94% of the observed cases, days in growth chambers at the University of Oklahoma. indicating the preferential fusion of the plastid-rich, mitochon- were emasculated prior to anthesis and were artifi- drion-poor sperm cell with the egg. In only one instance did the cially pollinated. were collected 81/2 to 9 hr after opposite pattern occur. Since the possibility of this result artificial (following the time table published in ref. occurring as the consequence ofchance in random fusions is <1 7), dissected, and fixed at room temperature in 3% glutaral- in 7000, this represents strong evidence for the presence of a dehyde in 0.067 M phosphate buffer (pH 6.8) for 6-8 hr. final putative recognition event occurring at the gametic level. was rinsed briefly and fixed in 2% osmium tetroxide in the same buffer, dehydrated in ethanol followed by is a phenomenon characteristic of flow- propylene oxide, and embedded in Spurr's low-viscosity ering plants involving two separate fusion events that incor- resin. Material was stained with uranyl acetate and lead porate both sperm cells into developmentally different lin- citrate by using techniques described elsewhere (7) and eages, both of which are biologically significant. Although observed with a Zeiss 10a transmission electron . numerous tubes compete for the opportunity to fer- Interpretation. To accurately discriminate the fate of the tilize the , if successful, the plastid-rich sperm cell, the following served as minimal delivers only two male gametes and both fuse with separate requirements that were met for each included data set: (i) the female reproductive cells. In contrast with and female gametophyte had to appear normal and show the non- plants, the angiosperms display a remarkable de- presence of male nuclei in both the egg and central cell, gree of conservation of gametic material at late stages of indicating the normal occurrence of double fertilization; (ii) . Of the two sperm cells, one fuses with two populations of plastids had to be readily identifiable the to produce the . Nearly simultaneously, within either the fertilized egg, central cell, or both; (iOi) the the second sperm cell fuses with the central cell in a second, ratio of male to female plastids could not exceed 1:5, the separate fusion producing the nutritive . Whether upper limit of the expected ratio; (iv) at least three paternal the gametic fusion that initiates this process is a random event plastids (as recognized by characteristics given in Table 1 and or one that results in a directed outcome has remained among the text) had to be present in either the egg or central cell, but the most refractory problems in angiosperm (1, not both. The plants used in the study displayed about 95% 2) and one that is addressed in the present study. seed set, with no evidence of apomictic reproduction. Since sperm cells are outwardly similar cells derived from a single mitotic division of the precursor generative cell, the RESULTS male gametes of angiosperms have traditionally been as- identical with neither cell Sperm Structure. The microgametophyte of P. zeylanica is sumed to be essentially (3), pos- tricellular and heterospermic, containing two dimorphic sessing a preferential tendency to fuse with the egg. Evidence sperm cells distinguished by whether or not the cell is contradicting this assumption has recently been reported in physically associated with the vegetative nucleus (4, 5). The Plumbago zeylanica (4, 5), indicating cytoplasmic hetero- sperm cell physically associated with the vegetative nucleus morphism in size, shape, organelle content, and physical is designated Syn, and the sperm cell not associated with the association with the vegetative nucleus. Although details of vegetative nucleus is designated Sua. The two sperm cells the fusion of gametes are concealed by nucellar cell layers differ in size, shape, nuclear dimensions, and organelle surrounding the egg, ultrastructural studies have revealed content. The S, is smaller, has a smaller nucleus, and that sperm cytoplasmic organelles in Plumbago are transmit- contains numerous plastids (average, 24.2 plastids) and few ted (6, 7) and can be identified by ultrastructural differences mitochondria (average, 39.8; ref. 5). In contrast, the larger in the fertilized female gametophyte (7). Differences present Svn infrequently contains plastids (average, 0.45) but contains in organelle content within the sperm cells can therefore be numerous mitochondria (average, 256; ref. 5) and an exten- employed as an ultrastructurally observable probe, which sive cellular projection that is physically associated with the

The publication costs of this article were defrayed in part by page charge Abbreviations: Sw,,, sperm cell physically associated with the vege- payment. This article must therefore be hereby marked "advertisement" tative nucleus; Sua, sperm cell not associated with the vegetative in accordance with 18 U.S.C. §1734 solely to indicate this fact. nucleus; TEM, transmission electron micrograph.

6129 Downloaded by guest on September 30, 2021 6130 Botany: Russell Proc. Natl. Acad. Sci. USA 82 (1985) Table 1. Differences in organellar morphology between plastids Table 2. Frequency of plastids present in the Sv,, based on a of sperm origin and those of the megagametophyte Poisson distribution of absolute organelle counts of 11 sperm Characteristic Sperm plastid Egg plastid cells (5) Size Up to 1.5',m Up to 2.5 ,um No. of plastids Frequency Average sec- 0 0.6347 tional area 0.518 ,tm 0.893 g±m 1 0.2885 Shape Ellipsoidal, elongate Pleiomorphic 2 0.0658 Stroma Denser than cytoplasm Equally electron dense 3 9.9351 x 1O-3 Lamellae Inflated Constricted 4 1.1290 x 1O-3 Plastoglobuli Aggregated Randomly distributed 5 or more 1.1095 x 1o-4

vegetative nucleus. These differences in organelle content used, with a male nucleus observed either approach- alone are significant enough to distinguish between the two ing or in the process of fusing with the nucleus of the female sperm cell types, with P < 0.001 for both mitochondrial and reproductive cell that it entered. In the remaining ovules, (i) plastid content (5). the required number of plastids was not observed, (ii) plastid Cytoplasmic Transmission. Mitochondria and plastids of populations could not be distinguished, (iii) reproductive sperm origin can both be identified within the female abnormalities were observed, or (iv) fixation was inadequate. gametophyte following fertilization (6). Differences in plastid The region where gametic fusion occurs was often evident structure being greater, these were used as the primary in the egg/ and central cell/endosperm (Fig. 3) located criteria for determining sperm cell fusion patterns in the near the terminal aperture of the pollen tube (6, 7) within present study. Sperm plastids may be distinguished from 20-30 ,um of the summit of the egg (8). Sperm mitochondria maternally originating plastids by the characteristics enumer- can sometimes be identified by size and structural differ- ated in Table 1 (see also ref. 7) and illustrated in Fig. 1. Since ences, as described in a previous paper (7). Near the site of one sperm lacks the presence and cell essentially plastids, unfused, bodies of sperm cyto- location of a number of sperm plastids in female cytoplasm gametic fusion, enucleated should be sufficient to identify which sperm cell fused with plasmic material were frequently seen between the zygote the egg. However, the infrequent presence of plastids in the and endosperm (Figs. 3 and 4). These enucleated bodies opposite sperm cell (Svn) implies that a certain number of (Figs. 3 and 4) arise in part through severance from the paternal plastids must be identified before this judgment can nucleated portion of the sperm cell during its expulsion from be made. Therefore, it is necessary to calculate the expected the pollen tube (7) or possibly through diminution of the cell, frequency of sperm plastids in the S, (Table 2), which was as has been suggested in barley (9). Such bodies contain only accomplished by using data generated from a previous study mitochondria of dimensions similar to those located in the (5). As the presence of three or more paternal plastids in the sperm cell (Fig. 4) and were not observed to contain plastids. cytoplasm of either the zygote or endosperm should accu- These observations indicate a probable origin of enucleated rately establish the fate of the plastid-rich Sua in all but 1% of cytoplasmic bodies from the Svn. A previous study indicates the cases (Table 2), this was selected as a requirement for that such bodies can remain between the egg and central cell each included data set. The actual margin of error is much unfused for up to several days after fertilization before less than 1%, because the female gametophyte was not degenerating (7). serially sectioned and therefore numerous paternal plastids present in the female gametophyte were not sampled. Further, in all female in which the pattern of fusion could be determined, the ratio between male- and female-originating plastids had to meet expected values (<1:5, male:female) and the two populations had to be easily distinguishable. Regions near the sperm nuclei were ob- served with particular scrutiny as these are preferential locations of sperm plastids following gametic fusion. Sperm Cell Fusion. Although >100 ovules were examined in detail during the course of this study, only 17 provided unambiguous data concerning the fate of the male cytoplasm. X t'2' 9 8 4 >nt '4 Of these, 16 demonstrated transmission of three or more paternal plastids into the cytoplasm of the egg or zygote. Fig. ~~~~. ~ X 2 illustrates one such case. In only 1 of the 17 examined W41"K ovules were more than three paternal plastids observed in the central cell. In none of the data sets used were paternal plastids observed in both female reproductive cells. Fertil- ization appeared to be proceeding normally in each of the

FIG. 2. TEM illustrating presence of both sperm plastids (Ps) and FIG. 1. (A) Transmission electron micrograph (TEM) of female egg plastids (Pr) in the zygote following gametic fusion. The egg gametophyte plastids (Pr). (B) TEM of sperm plastids (P.) within nucleus (EN) and sperm nucleus (SN) have fused as evidenced in pollen tube near time of fertilization. (x 12,640.) other TEMs (not shown). (x6570.)

Downloaded by guest on September 30, 2021 ;2eit;a Botany: Russell Proc. Natl. Acad. Sci. USA 82 (1985) 6131 association with the vegetative nucleus. Such extreme dif- ferences as those described in Plumbago are likely to be uncommon in angiosperms, if only for the reason that a minority of the flowering plants has both plastids and mito- chondria in their sperm cells (11). Cytoplasmic heterosper- my, as seen to date, originates with the presence of a physical association between the vegetative nucleus and one sperm cell. Examples of physical associations between a sperm cell and the vegetative nucleus have been described elsewhere in Brassica oleracea (12), Brassica campestris (13), Hip- peastrum vitatum (14), and Spinacia oleracea (15) and have been shown but not discussed between the vegetative nucleus and generative cell in Gossypium hirsutum (16), Nicotiana FIG. 3. TEM of region of zygote (Z) and endosperm (En) where the Sn fused with the central cell. Cytoplasmic body (unlabeled tabacum (17), and Prunus avium (17). In only the Poaceae, to larger arrowhead) is present amid the discharged contents of the date, has the absence of such as association been docu- pollen tube (PD). Small sperm mitochondria (smaller arrowheads) mented, specifically in Alopecurus pratensis (10), Hordeum are evident in the region attributed to the Sn cytoplasm. (x8280.) vulgare (9), and Triticale (18). The extent of the occurrence of cytoplasmic heterospermy cannot be estimated at this time but may occur in numerous angiosperms. DISCUSSION Nuclear heterospermy is characterized by genetically transmissible nuclear differences in the sperm cells and its Clearly, sexual reproduction in P. zeylanica results in the occurrence is much rarer. Since such nuclear differences preferential transmission of the plastid-rich, mitochondrion- have to arise either during or after generative , poor sperm cell (Sua) into the egg cell to participate in the circumstances under which nuclear heterospermy may fertilization; the plastid-poor, mitochondrion-rich sperm cell occur are limited. Nuclear heterospermy has been demon- (Svn) participates in fusion with the central cell to form the strated in only one plant, maize, in which the sperm cells may endosperm. This pattern of gametic fusion results in trans- differ in the presence or absence of B (19). The mission of paternal plastids into the embryo, in which they apparent mechanism for this difference is a of may, if allowed to remain viable, fulfill their developmental B chromosomes at generative cell division, such that only potential to become . Conversely, paternal mi- one sperm cell receives B chromosomes (19). Other mecha- tochondria are preferentially transmitted into the endosperm, nisms for the generation of genetic differences in sperm cells where by shear numbers alone, they may stimulate increased may exist but have yet to be described. respiration and overall metabolic activity of the endosperm Preferential Fusion. The indication that transmission of during its early development. The occurrence of such dimor- paternal plastids into the egg occurred in 16 of 17 cases phism coupled with a pattern of preferential fusion could represents a statistically compelling case for preferential have important implications for our understanding of fertil- fusion. Were such fusions random, the expected frequency of ization (10). the obtained result would be 1.297 x 1O' (or <1 in 7000), as Sperm Heteromorphism. The sperm of flowering plants determined from the binomial distribution. Accordingly, one displaying consistent differences in morphology or transmis- may conclude that the two dimorphic sperm cells differ very sible cytoplasmic differences, as in P. zeylanica, are defined significantly in their ability to fertilize the egg and that this herein as examples of cytoplasmic heterospermy. In Plum- pattern is consistent with cytoplasmic organelle content and bago, such differences extend to size, shape, nuclear dimen- a number of other sperm cell characteristics (5). Most of the content, and physical resulting are therefore products of fusion with the sions, mitochondrion content, plastid plastid-rich, mitochondrion-poor sperm cell, Sua. It has not yet been demonstrated whether sperm cells giving the oppo- site pattern of transmission will or will not be successful in forming a new plant. Given the normal abortion rate of -5%, this might represent one possible source of seed abortion, occurring even in plants grown under optimal conditions. The only other report of preferential fertilization in angiosperms is that of the nuclear dimorphic sperm cells of maize (19), in which the sperm containing the nondisjunct B chromosomes are preferentially transmitted into the embryo. It was determined that 76.8% of the nondisjunct B4 chromo- somes in the TB-4a strain and 67.1% of the nondisjunct B9 chromosomes in the TB-9b strain were transmitted into the egg cell to form hyperploid embryos. Roman (19) therefore concluded that sperm cells containing an extra B chromo- some may participate in "directed fertilization," in his terminology. He states two hypotheses for this behavior: (i) genetic influences from the B affected the pattern of fusion or (ii) the sperm cells are polarized in organization or behavior. Since preferential fertilization was demonstrated in hyperploid sperm cells of B chromosomes with two different chromosomal origins, Roman discards the first hypothesis, stating that "it is doubtful that the deficiency FIG. 4. TEM of cytoplasmic body originating from the Svn with or duplication ofA-chromatin" material in the "gametes is an arrowheads indicating points of contact between the zygote (Z) and essential factor in determining directed fertilization" (19). endosperm (En), located amid degenerating pollen cytoplasm (PD). Indeed, the chances of selecting two very different B chro- (x28,620.) mosomes (namely, chromosomes 4 and 9) carrying similar Downloaded by guest on September 30, 2021 6132 Botany: Russell Proc. Natl. Acad. Sci. USA 82 (1985)

genetic tendencies favoring their own preferential transmis- both the egg and central cell simultaneously prior to fusion (7, sion are remote. 21), there is an opportunity for cellular recognition as the However, Roman's hypothesis that preferential fusion male and female gametic membranes contact one another and resulted from a preferred pattern of sperm cell arrival-the this may result in the discrimination of even subtle differ- first sperm cell having an advantage in selecting the pattern ences in the two cells. Whether such recognition involves a of fusion (19)-appears to be an oversimplification. If pollen specific receptor, whether it involves the cooperation of tube discharge is as rapid in vivo as it is in vitro, the advantage multiple receptors, or whether differences in membrane ofa sperm cell arriving in the female gametophyte first would charge and surface area alone might provide the mechanism be -0.2 sec, presumably too little to confer specificity. Even for discrimination is central to further elucidation ofthe basis in the grasses, the sperm cells remain closely associated (10). of sperm specificity during double fertilization. In more tightly linked sperm cells, the temporal advantage could be even less. An alternative hypothesis is that directed I thank Susan M. Heinrichs for excellent technical assistance in fertilization in maize is the result of preferential behavior of this study and Diana Secor for reading the manuscript prior to the mitotic spindle resulting in the accumulation of nonpair- publication. This research was supported in part by National Science ing chromosomes into the cell that is developmentally more Foundation Grant PCM-8208466, the University of Oklahoma Re- search Council, and a College of Arts and Sciences Research likely to fuse with the egg cell. The nondisjunction of the B Fellowship. Use of facilities of the Samuel Roberts Noble Electron chromosome at generative cell division therefore may allow Microscopy Laboratory is gratefully acknowledged. differences in the fertilization capacity of the sperm cells to be seen without itself directly predisposing the sperm cells to 1. Jensen, W. A. (1974) in Dynamic Aspects ofPlant Ultrastruc- preferential fertilization. The capacity for preferential fertil- ture, ed. Robards, A. W. (McGraw-Hill, New York), pp. ization apparently may already be present in these sperm 481-503. cells from their inception whether or not such differences are 2. Dumas, C., Knox, R. B., McConchie, C. A. & Russell, S. D. readily evident. (1984) What's New Plant Physiol. 15, 17-20. Gametic Recognition. The preferential fusion of a specific 3. Jensen, W. A. & Fisher, D. B. (1968) Protoplasma 65, sperm morphotype is unambiguous evidence for the presence 277-286. 4. Russell, S. D. & Cass, D. D. (1981) Protoplasma 107, 85-107. of a final putative recognition event occurring at the gametic 5. Russell, S. D. (1984) Planta 162, 385-391. level. Presumably such recognition, mediated by the same 6. Russell, S. D. (1980) Science 210, 200-201. mechanisms influencing recognition events in other systems 7. Russell, S. D. (1983) Am. J. Bot. 70, 416-434. (20), is present in the sperm of some and perhaps many 8. Russell, S. D. (1982) Can. J. Bot. 60, 2219-2230. angiosperms. If a gametic system of recognition is present, 9. Mogensen, H. L. & Ruschd, M. L. (1985) Protoplasma, in however, its function is clearly different than the well- press. described sporophytic and gametophytic compatibility sys- 10. Heslop-Harrison, J. & Heslop-Harrison, Y. (1984) Acta Bot. tems that exercise their effects on the pollen tube. Once a Neerl. 33, 131-134. pollen tube arrives at the female the success of 11. Kirk, J. T. 0. & Tilney-Bassett, R. A. E. (1978) The Plastids: gametophyte, Their Chemistry, Structure, Growth and Inheritance (Elsevier its two sperm cells is essentially assured. Gametic recogni- Biomedical, Amsterdam). tion, if present, may serve a different purpose: first, to 12. Dumas, C., Knox, R. B. & Gaude, T. (1985) Protoplasma 124, identify the two sperm cells as candidates for cellular fusion 168-174. within the female gametophyte, and second, to provide the 13. McConchie, C. A., Jobson, S. & Knox, R. B. (1984) in Polli- means by which one sperm cell is specifically targeted for nation '84, eds. Williams, E. & Knox, R. B. (School of fusion with the egg. Botany, University of Melbourne, Melbourne, Australia), pp. One ultrastructurally evident characteristic of gametic 26-29. fusion that may reflect on the mechanism ofrecognition is the 14. Mogensen, H. L. (1985) in Pollen Biotechnology and Ecology, presence ofenucleated sperm cytoplasmic bodies that remain eds. Mulcahy, D. L. & Ottaviano, E. (Springer Verlag, New unfused between the egg and central cell (Figs. 3 and 4). Their York), in press. continued presence after fertilization (7) suggests that a 15. Wilms, H. J. & van Aelst, A. C. (1983) in Fertilization and threshold effect occurs during recognition, whereby only the Embryogenesis in Ovulated Plants, ed. ErdelskA, 0. (Veda, main sperm cell body can effect gametic fusion. Such a Bratislava, Czechoslovakia), pp. 105-112. 16. Jensen, W. A. & Fisher, D. B. (1970) Protoplasma 69, mechanism would prevent the fusion of small enucleated 215-235. bodies, which might inhibit further fusion with the nucleated 17. Cresti, M., Ciampolini, F. & Kapil, R. N. (1984) J. sperm cell. Submicrosc. Cytol. 16, 317-326. The discrimination between sperm cells that differ in size, 18. Schroder, M. B. (1983) in Fertilization and Embryogenesis in volume, surface area, and organelle content may use any Ovulated Plants, ed. Erdelskd, 0. (Veda, Bratislava, Czecho- combination of these features as a means of conferring slovakia), pp. 101-104. preferential fusion, but it seems most likely that the deter- 19. Roman, H. (1948) Proc. Natl. Acad. Sci. USA 34, 36-42. mining feature is reflected and perhaps caused by specific 20. Clarke, A. & Knox, R. B. (1978) Q. Rev. Biol. 53, 3-28. differences in the cell surface. As each sperm cell contacts 21. Mogensen, H. L. (1982) Carlsberg Res. Commun. 47, 313-354. Downloaded by guest on September 30, 2021