Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Early Development in Fern Gametophytes: Interpreting the Transition to Prothallial Architecture in Terms of Coordinated Photosynthate Production and Osmotic Ion Uptake

Early Development in Fern Gametophytes: Interpreting the Transition to Prothallial Architecture... Abstract Gametophytes of Onoclea sensiblis L. were grown under various light and media‐ion conditions to gain a better understanding of the source/sink relationships between photosynthetic and ion‐absorbing cells. There was a clear interdependency between green cell and rhizoid functions, such that the growth and development of the rhizoids was completely dependent on the internal delivery of photosynthates from green cells, and conversion of the one‐dimensional filament into the two‐dimensional prothallus required monovalent cations that could only be provided by rhizoid uptake. The need for monovalent cations was related to osmotic demands of dividing and expanding cells; prothallial development was blocked by monovalent cation deficiency, and the system resorted to Na+ uptake to support cell expansion when K+ was absent. Surgical excisions of filament cells further demonstrated the high degree of coordinated growth between the light‐absorbing and ion‐absorbing regions. It was also learned that excised sub‐apical cells of the protonemata, like the intensively studied apical cell, were capable of remodelling remnants of the filament into a normal prothallus. Key words: Fern gametophyte, Onoclea sensibilis L., development, ion uptake, photosynthate. Received: 20 August 2001; Returned for revision: 9 October 2001; Accepted: 12 November 2001. INTRODUCTION The fern gametophyte is a small plant that exists as a prolonged intermediate in the fern life cycle, between the germination of a spore and the mature sporophyte. Following its emergence from a spore, it grows from two cells into a distinctively shaped structure containing several hundred cells. In the course of development one end of the structure becomes a flattened, single‐layer of chloroplast‐containing cells, resembling a leaf. The other end manifests an array of long, filamentous rhizoids that have the appearance, and apparently serve the same function, as root hairs. The rhizoids are differentiated directly from basal green cells (as a consequence of an unusual, highly asymmetric cell division); thus there are few intervening cells and no vascular tissue between the photosynthetic zone and the region designated for ion and water uptake. General reviews of fern gametophyte anatomy and development include Dyer (1979), Raghavan (1989) and Banks (1999). Both germination and development of later form are influenced by the intensity and spectral quality of light and, in the laboratory setting, gametophytes of many ferns can be induced to exhibit a considerable range of morphologies, from long, uniseriate, tip‐growing filaments (generally achieved in extended periods of low‐intensity light, or darkness) to prothalli with various length‐to‐width ratios (produced by adjusting light intensity or photoperiod). Regardless of the experimentally induced form of the green‐celled axis, however, it is nearly always observed that exposure to light and/or photoperiod conditions that approximate those in the natural environment leads to new rounds of cell divisions and the emergence of the ‘true’ prothallial shape. Utilizing a variety of fern species, researchers have shown that an early, pivotal event in the formation of the prothallus involves the re‐orientation of a cell plate (from a transverse to a longitudinal alignment) during division of the apical chlorocyte (reviewed in Dyer and King, 1979). This step has the effect of altering the pattern of growth from extension in one dimension (sustaining a filamentous architecture) to expansion in two dimensions (producing a planar surface). Much of the research over the past 50‐plus years has understandably focused on the nature of photoreception and the mechanisms that govern the timing and positioning of divisions in this cell. While these undertakings have been most useful in furnishing insights about the nature and cellular location [reviewed in Furuya (1983) and Raghavan (1989)] of the pigments that trigger the fern photoresponses, intriguing observations concerning planes of minimum surface area (Cooke and Paolillo, 1980a), cytoskeletal proteins (Stetler and DeMaggio, 1972), localized ion gradients (Racusenet al., 1988), bioelectrical fields (Racusenet al., 1988) and the frequencies of plasmodesmata (Tilneyet al., 1990) have thus far failed to yield a coherent framework of events that explains how the transition from one‐dimensional to two‐dimensional growth is accomplished. Moreover, since division of the apical cell is considered to be the gateway to the emergence of thallial architecture in the fern gametophyte, we are presently unable to derive potentially useful insights from this comparatively simple system which might help guide the study of analogous, but more complex, morphogenetic phenomena (such as leaf development) in vascular plants. Here I re‐examine the formation of the prothallus in a fern gametophyte in the context of a ‘whole‐plant’ developmental response. In particular, the focus of this study was to examine how the allocation of resources between photosynthetic and non‐photosynthetic regions was related to attainment of the overall form of the organism. Embedded within this undertaking were the following specific aims: (1) to relate, quantitatively, the amount of photosynthetically active light to the changes in prothallial and rhizoid biomass; (2) to examine the correlation in growth rates between the emerging prothallus (presumed photosynthetic source) and the proliferation of rhizoids (presumed sink); (3) to determine how the length of the interstitial region between source and sink influenced the delivery of photosynthates and ions during development; and (4) to establish the rhizoids as the primary site for nutrient ion uptake to satisfy the osmotic requirements of both non‐green and green cells. MATERIALS AND METHODS Gametophyte culture Fertile fronds of Onoclea sensibilis L. were collected in January near the upper branch of the Patapsco River in Carroll county, Maryland, USA. They were packed into zip‐closure plastic bags and stored at –35 °C until needed. A supply of spores adequate for approx. 6 months of experiments could be obtained from ten fronds, and the procedure to release them began by surface‐sterilizing fronds in 15 % laundry bleach, and then drying them overnight on aluminium foil in a clean, but non‐operating, laminar flow hood. Spore cases were removed by hand from the dried fronds into a mortar, and then ground to a fine powder with the pestle. This material, containing spores and debris, was placed into a screw‐top plastic centrifuge tube and stored at –35 °C until needed for an experiment. To start a culture of gametophytes, spore‐containing powder was shaken through a small sieve that had been made from 41 µm nylon mesh. Approx. 0·5 ml of this screened spore powder was collected in a 15 ml plastic centrifuge tube, and then wetted, with agitation, in a solution of 0·1 % triton X‐100 for 5 min. A clinical centrifuge operating at 300 g for 15 s was used to pellet the spores, and the wetting solution was discarded. The spores in the pellet were dispersed and surface‐sterilized in a solution of 0·1 % triton X‐100 and 15 % liquid laundry bleach for 4 min. All the remaining steps were carried out using a laminar flow hood and sterile technique. The tube was next centrifuged for 15 s at 300 g, the sterilizing solution decanted and the spores washed twice with sterile water, using the same techniques. After the second wash, the spores were diluted to a desired inoculation density (usually by adding 10 ml of sterilized water) of approx. 1000 spores ml–1. Surface sterilized spores were transferred to 1 % agar or agarose medium in 15 cm plastic Petri plates. The control medium for gametophyte growth was based on the APW‐6 recipe provided by Cooke and Racusen (1982). It consisted of 1 mm 2‐morpholinoethanesulfonic acid (adjusted to pH 6), 0·1 mm NaCl, 0·1 mm KCl and 0·1 mm CaCl2, solidified with 1 % Bacto‐agar (Difco Laboratories, Detroit, MI, USA) or low‐salt agarose (Ultra pure agarose 5510UA, BRL/Life Technologies, Rockville, MD, USA). As described in the Results, gametophytes grew equally well on a Bacto‐agar substrate with no added salts, apparently because this general purpose agar had sufficient ion content to support the osmotic and nutritional needs of these plants. Using values provided by the manufacturer (expressed in salt and ash content), it was calculated that a 1 % agar mixture contained, in the bound and unbound state, approx. 100 µm Na+, 5 µm K+ and 1 µm Ca2+. For ion deficiency experiments, therefore, certain salts were eliminated and the solidifying agent was low‐salt agarose, an agar‐type matrix in which the relevant cations above were present in concentrations at least five‐times lower. During the initial phases of this study, sucrose was present in the agar media at concentrations of 0·1 to 1 %. As comparisons of gametophyte growth with and without sucrose in darkness or various light conditions showed no significant differences, this ingredient was not included in the media in any of the experiments for which data are presented (see also Results and Discussion). Petri dishes were sealed with strips of Parafilm and placed in clear plastic, zip‐closure bags to inhibit drying and contamination. These bags were then placed in continuous cool white light at 25 µm m–2 s–1 for 2 d to stimulate germination. At this point plates of 2‐d‐old gametophytes were transferred to darkness or to a variety of light and photoperiod regimes as described below. Terminology During development of the gametophyte the organism undergoes dramatic changes in morphology, and various terms are needed to distinguish newly differentiated cells and changing tissue infrastructure. The protonemata refers to the filamentous stage of gametophyte development. This structure may be quite elongated in dark‐grown organisms and comprises a tip‐growing apical cell and other sub‐apical, cylindrical cells. The chloroplast‐containing cells of the protonemata, or filament, may be referred to as chlorocytes, and these green cells also make up the prothallus, the planar two‐dimensional structure that develops from a protonematal initial in light. Rhizoids are filamentous, single‐celled, non‐green cells that elongate by tip growth. The initial, primary rhizoid emerges from the spore during germination. Secondary (2°) rhizoids are those that emerge from other cells of the protonemata or the prothallus. The term ‘spore end’ is used in this paper to designate the original site of emergence of the protonemata, which, during the extension of the filament, represents the basal end of the filament opposite to the tip cell. Light and photoperiod treatments All light treatments in these studies were derived from adjustments to the timing or intensity of the cool white fluorescent source described above. To expose gametophytes to particular photoperiodic conditions, the bagged Petri plates were placed in a light‐tight box over which was installed the fluorescent source. Room air was circulated through the box with a fan and the temperature range over the lights‐on/lights‐off periods was 24–28 °C. The light source and fan were connected to a 24 h timer. For experiments in which gametophytes were exposed to lower intensities of light, individual Petri plates were wrapped in aluminium foil, and a rectangular window was created by removing a section of the foil covering the lid. This opening was then covered with one to six layers of neutral‐density filters made from a sheet of developed X‐ray film. The filters were fastened to the foil‐covered lid with opaque tape. The films could be empirically adjusted to different OD (optical density) values by exposing them for various lengths of time to a red darkroom safelight. The filters in these experiments were made from a film that exhibited OD490 of 0·585 and OD640 of 0·565, as measured in a scanning spectrophotometer. The average of these values, 0·575, was used to calculate intensities of white light produced by the following combinations of filters (all in µm m–2 s–1): 6·75 (one filter), 1·90 (two filters), 0·52 (four filters) and less than 0·01 (six filters). Experimental set‐up and photo‐documentation All data for these experiments were collected by photographing individual developing gametophytes every day or two, over a period of approx. 30 d. Since these transitions in form are complex and do not readily lend themselves to quantification by simple parameters such as numbers of cells or dimensions of structures, much of the Results section consists of photographic sequences that capture the progressive changes arising from cell divisions, cell expansion and cell differentiation. Each of the photomontages, therefore, is the record of a single, representative gametophyte; in every instance I obtained complete photographic records of comparable development in at least six replicate organisms. Working with a 25× dissecting microscope that was placed in a laminar flow hood, individual gametophytes were lifted from the agar surface on the tip of a number 11 scalpel and promptly transferred to the agar surface of a fresh plate. To counteract the desiccating environment of the laminar flow hood, an aluminium foil shield was taped behind the stage of the microscope and, with practice, one could transfer the gametophyte with some adherent agar, which, following transfer, was removed from the new agar surface. To improve visualization of gametophytes on an agar surface, a technique was developed for floating a small (approx. 3 mm × 3 mm) ‘chip’ of clear cellophane (cat. No. SE542; Hoefer Scientific, San Francisco, CA, USA) over each gametophyte. The chips were sterilized in 95 % alcohol for 30 min, followed by two 10 min washes in sterile water. Small, pointed strips of Whatman number 1 paper were used to wick away some of the water between the chip and the gametophyte so that the chip did not drift away but at the same time did not press too closely to the organism. Considerable practice was needed to consistently achieve the desired apposition between the cellophane and the organism. These miniature coverslips were sufficiently porous to allow gas and nutrient exchange; unimpaired growth of gametopytes under these conditions was observed for periods exceeding 60 d. In certain experiments, localized changes in media ion composition were accomplished by adding droplets of media or smaller, ion pre‐loaded chips to the surface of the cellophane overlay. Earlier studies of diffusion through the cellophane were made by preparing two 5 cm × 5 cm layers of hydrated cellophane in the following manner: the lower layer was saturated with a 0·3 g ml–1 solution of bromophenol blue at pH 5, and placed on a glass plate. Bromophenol blue is yellow at this pH. A second layer of cellophane saturated with pH 5 buffer was laid on top of the dye‐containing layer and small KOH pellets were placed on the surface of the second layer. This entire preparation was then covered with a glass Petri dish lid and the edges sealed with water. Bromophenol blue (molecular weight = 670) does not penetrate the cellophane, but KOH penetrated the upper layer instantly, raising the pH and changing the dye colour to blue. The migration of the alkalinizing band could thus be measured over time; the rate of diffusion of ions (D) through cellophane appeared little different to that predicted for diffusion in unstirred water (e.g. for K+, D = 1·9 × 10–4 cm–2 s–1) A knowledge of the approximate rate of lateral diffusion in the cellophane was important in the design of experiments in which a concentrated solution of KCl was added to rhizoid and prothallus ends of the gametophyte grown in cation‐depleted conditions (see Results and Fig. 7). In these experiments, gametophytes that were blocked during prothallial expansion were lifted from the agar surface and placed on a (8 mm × 8 mm) section of plastic, ‘food wrap’ film, resting on the agar. The plastic is hydrophobic and thus impermeable to ions and water. The rhizoid and prothallial ends of the organism were then separately covered with cellophane chips that had been imbibed with sterile water, such that a 50 µm gap remained between them. Given the hydrophobic nature of the underlying plastic film, this increment was sufficient to keep water from establishing continuity, and thus a path for ion diffusion between the chips. The external surfaces of the gametophyte could, of course, serve as a path for diffusion, but this would be expected to be lower than the bulk flows in surrounding water, and, in any case, would not be artificially introduced, since such diffusion would be attributable to inherent features of the organism. Based on studies of KOH migration, it was possible to generate a basal (rhizoid‐end) or acropetal (prothallus‐end) gradient of K+, that was sustained for several days, by placing an additional 1 mm × 1 mm chip, preloaded with 10 mm KCl, on the surface of the cellophane, approx. 2 mm back from the distal ends of the rhizoid or prothallial cells. Observations were made with a Zeiss inverted microscope and photographs were taken with a conventional 35 mm camera, using T max (300 ASA) film. During observations and photography, gametophyte cells were typically exposed to the more intense light from the microscope condenser for less than 5 min; no significant differences in the light‐mediated development of gametophytes were observed that might be attributed to the frequency of examination under the microscope. Photographic negatives were scanned and converted into digital format. Software was used to adjust image brightness, contrast and sharpness to levels which revealed the greatest amount of cellular detail. Microsurgery Portions of filamentous gametophytes were excised with a fresh number 11 scalpel, working under a dissecting microscope at 25× in a laminar flow hood. To prevent the blade from pushing the gametophyte into the agar, the organism was first transferred with the scalpel tip to a 1 cm × 1 cm sheet of sterile, washed cellophane that was placed on a region of the agar surface that was free of gametophytes. The cuts were made with a ‘meat cleaver’ motion and the excised portion was transferred with the scalpel tip back onto the nearby agar. The desired explant was then picked up on a cellophane chip gripped by forceps, transferred to a fresh plate and covered with the chip, as described above. Survival rate for these procedures was quite high; more than 95 % of the surgically removed portions of gametophytes resumed cell division and produced prothalli. RESULTS Light fluence and the conversion of filamentous to planar form Gametophytes grown from germinated spores under a variety of light conditions displayed a consistent relationship between the form attained after 20 d and the amount of white light received in a 24 h period. Organisms grown for 20 d under different total fluences, achieved by adjusting the intensity of light with neutral density filters (Fig. 1A–D), displayed the characteristic range of forms, from filamentous (in complete darkness) to fully expanded prothalli (in unfiltered light), with the filamentous architecture giving way to two‐dimensional forms in organisms grown under less than two layers of neutral density filters (Fig. 1C, D). As described in the Materials and Methods, the light intensity reaching ferns under two layers of filters was 1·9 µm m–2 s–1, giving a 24 h fluence of approx. 165 000 µm m–2. Filaments raised under these light conditions had shortened cells and higher densities of chloroplasts, but rhizoid growth, like that seen in filaments produced under lower light intensities, was minimal. Given the resolution obtainable with layers of identical neutral density filters, there was a clear demarcation between the responses of gametophytes grown under one (Fig. 1C) or two (Fig. 1B) filters. When grown under one filter (6·75 µm m–2 s–1) only prothalli were formed and, except for some extension of the lower (spore end) axis, these were very close in appearance to prothalli produced in unfiltered light. Organisms grown for 20 d in different‐length photoperiods of unfiltered white light were morphologically similar to those obtained under various intensities of continuous white light (Fig. 1E–H). The 24D (continuous darkness) and 24L (continuous light) regimes are, of course, identical to the dark and unfiltered light treatments in the previous experiments. In the photoperiod studies, however, the threshold for maintaining gametophytes in a filamentous state was 2L22D (2 h light and 22 h darkness; Fig. 1F), and these organisms appeared outwardly similar to those grown under two neutral density filters (Fig. 1B). The calculated total fluence for the 2L22D photoperiod of 180 000 µm m–1 compares favourably with that determined above for the two layers of neutral density filters, indicating that these light‐mediated developmental responses are consistent with the predictions for light dose/duration reciprocity. Only planar growth was observed in organisms grown in 4L20D (Fig. 1G), and similar to prothalli produced under one neutral density filter (Fig. 1C), the cordate lamina was somewhat more elongated than that seen in 24L organisms. This extension of the longitudinal axis was apparently due to the presence of several conspicuously elongated cells between the middle and the base of the planar surface. Correlation of prothallial expansion and rhizoid proliferation A consistent feature of every experiment that was undertaken in this study was that a sufficient quantity of white light (>2 µm m–2 s–1) was an absolute requirement for both prothallial and rhizoid development. Figure 2 compares the change in the calculated volume of the prothallus and the aggregate volume of rhizoids in 10 d, dark‐grown organisms as they responded to continuous white light. The rate of prothallial expansion exceeded the rate of rhizoid production and elongation such that there was a consistent six‐ to eight‐fold difference in tissue volumes over a 10 d period. As indicated in the Materials and Methods, the presence of 0·1–1 % glucose or sucrose in the media had no significant effect on any aspect of filamentous or two‐dimensional growth of gametophytes grown in darkness, in filtered light or in unfiltered light (data not shown). Over extended periods of darkness (up to 40 d) gametophytes grew as increasingly longer filaments, but in these organisms there was no extension of the germination rhizoid and no further rhizoids were produced (e.g. see Fig. 5A). At intensities just below the threshold of total fluence which triggered two‐dimensional growth of the apical cell, chloroplast density increased markedly within the filament cells (fx2, Fig. 1B; 2L22D, Fig. 1F), but still no rhizoid growth occurred. On the other hand, total fluences that resulted in the early cell divisions of two‐dimensional growth also promoted rhizoid proliferation (fx1, Fig. 1C; 4L20D, Fig. 1G). These results were always seen, regardless of the age or length of the filament, or the absence of spore end or other filament cells, in surgically severed organisms. Placement of the planar and rhizoid poles exhibits a threshold dependence on axis length In these experiments the range of gametophyte morphologies produced under different light conditions was used to examine how the apparent temporal coordination of two‐dimensional and rhizoid growth might be influenced by the distance between the initial rhizoid pole near the spore, and the apical cell. This distance was generally less than 25 µm (one cell) in full‐light grown organisms and could be extended to nearly 1000 µm in dark‐grown gametophytes (e.g. Fig. 5A). Figure 3 shows the positions and relative growth of 2° rhizoids during the development of selected 2L22D photoperiod‐grown gametophytes with initial filament lengths of 100 µm (A), 200 µm (B) and 250 µm (C). In cases where the distance between the filament apical cell and the spore end of the filament was less than 100 µm, it was consistently observed that all subapical cells of the filament served as sites for 2° rhizoid production, and thus became incorporated into the maturing prothallus over 16 d (Fig. 3A). In organisms that were selected with lengths of approx. 200 µm, the zone of 2° rhizoid growth was extended, over the 16 d observation period, from the spore end towards the cells of the emerging prothallus (Fig. 3B). The series of time‐lapse photographs in Fig. 4 shows that this relocation of 2° rhizoid production occurred as a progressive migration toward the expanding mass of cells at the former filament apex. This phenomenon appeared to depend solely on distance, irrespective of the number of cells that comprised this zone of the filament. For example, the organism in Fig. 4 shows this temporal migration of 2° rhizoids over two cells; the organism in Fig. 7A–C shows this process occurring over the same distance, but involving three to four cells. In situations where the apical cell and spore wall were separated by distances greater than 200 µm (e.g. 250 µm), 2° rhizoids emerged only from cells near the spore wall and those intimately associated with the expanding prothallus (Fig. 3C). Over the 16 d observation period, this relocation of the site of rhizoid emergence established the base for prothallial emergence at the distal end of the filament, and still‐attached cells of the subapical filament and the spore end were not incorporated into the prothallial architecture. In many cases where very long filaments were grown in periods of extended darkness, transfer to full light resulted in the production of two prothalli, one derived from the filament apical cell and the other emanating from the spore end, both attached to the remnant strand of the filament (Fig. 5A–C). Regeneration of gametophytes from excised filament cells Further exploration of the positioning of the rhizoid pole in relation to the emerging photosynthetic surface was undertaken by surgically separating regions of dark‐grown, or 2L22D photoperiod‐grown organisms. When the apical end of the filament was excised from the basal end (including the remnant of the spore), with the explants being exposed to full light, new rounds of cell divisions, starting in the apical cell, gave rise to a prothallus within 30 d (Fig. 5D–F). Prothalli were also formed from the spore end explants in light, with the spore region, as opposed to any of the remaining filament cells, serving as the site for renewed cell divisions (data not shown). Similar filamentous explants, maintained for over 24 h in darkness or under light conditions which did not support photosynthesis, were unresponsive when exposed to full light and were assumed to have died from the interruption of metabolite flow. Regardless of the length of the filament, the number of filament cells (from a single intact apical cell to more than four cells), or the location of excision, only the cells at the tip or base extremities were involved in prothallial regeneration. As was the case with rhizoid emergence in filaments over 200 µm long (described above), new rhizoids were positioned near the base of the expanding cell cluster and their numbers and rate of elongation were positively correlated with the rate of two dimensional expansion. The formation of prothalli at the apex and base of severed filaments was reminiscent of the bipolar formation of prothalli often observed in intact, long filament gametophytes (see Fig. 5A–C), suggesting at first that the intervening, generally elongated cells of the filament did not have regeneration capabilities. It was therefore somewhat surprising to find that excised segments of the mid‐filament were also capable of renewed cell division and ultimately prothallus formation (Fig. 5G–I). In all cases where such regeneration was observed, new rhizoids appeared from one of the filament cells within 2 d, followed by bulging of one of the side walls in the adjacent cell. This bulging was followed by cell divisions which further distorted the original symmetry of the filament, giving rise to a trajectory of pre‐prothallial growth that was oriented at a right angle to the original filament axis. Minimum ion requirements for gametophyte development Regeneration of prothalli from filament cells that were separated from the spore end indicated that the loss of access to organic and mineral nutrients within the spore was readily surmounted if even one intact chlorocyte remained in contact with the basic salt medium and continued to receive sufficient light. To determine how ions in the media might influence regeneration from filament explants, spore end‐excised fragments from 10‐d‐old, dark‐grown or 2L22D‐grown organisms were plated on sucrose‐free, agarose media lacking either K+ or Ca2+ and exposed to full light. Surprisingly, the elimination of added K+ or Ca2+ in the medium had no significant effect on the timing or extent of prothallial regeneration (data not shown; results comparable with those shown in Fig. 5D–F). Since the removal of either K+ or Ca2+ had no effect on gametophyte development, further experiments were undertaken to determine how these organisms were capable of normal development under significantly lower levels of two major nutrient cations. First, by removing both K+ and Ca2+ from the recipe, leaving only NaCl (and the MES buffer) in the agarose‐solidified medium, it was observed that prothalli were still able to form, but the progression was slower and resulted in smaller prothalli (Na+/ose, Fig. 6E–H). Production of prothalli was, however, strongly curtailed by the removal of K+ and Na+ (Ca2+ could be present or absent) from agarose‐based media (Z/ose, Fig. 6I–L). Similar symptoms were obtained in organisms germinated and grown on –K+, –Na+ media and those transferred to this ion‐deficient media 2 d after germination on control media. This inhibition of development could be reversed by adding K+ (or Na+, but not Ca2+) to the organism’s environment; this was done either by transfer to a +K+ plate or by adding a few drops of 0·1 mm KCl to the cellophane chip overlying the organism. Robust rhizoid proliferation and laminar expansion were evident within several days (+K+, d35, Fig. 6L). In the course of these ion‐removal studies, it was also learned that monovalent cation deficiency symptoms could only be elicited on media solidified with low salt agarose; Fig. 6A–D shows the normal development of the prothallus when Bacto‐agar was used as a solidifying agent in a medium that had no added cations. Comparison of the resident cation content in agarose and Bacto‐agar powders indicated that the concentration of each of the major cations is at least five times higher in standard Bacto‐agar. It is not known what proportion of these contaminating ions are unbound and become part of the ionic composition of the media but, judging from these results, it is clear that gametophytes are capable of extracting sufficient ions for growth from unsupplemented Bacto‐agar. Establishing rhizoids as the primary site for ion uptake The monovalent cation deficiency symptoms produced by growing gametophytes on agarose without added K+ or Na+ were used as a starting point for experiments to learn more about the site(s) of osmotic ion uptake in the developing fern gametophyte. As shown in Fig. 7E–H and I–L, deficiency symptoms, as manifested by slower growth, were evident within 10 d. After 23 d, there was no further progression of prothallial development, and indications of general ill health were evidenced by clumping of chloroplasts and a reduction in the number and length of normal‐looking rhizoids. When K+ was added (indicated by +) to alternate sides of a pair of cellophane chips overlying the base (spore end) and the developmentally arrested prothallus, re‐initiation of prothallial expansion was evident only when K+ was able to reach the rhizoids (Fig. 7M–O). Potassium reaching the prothallial portion, but not the rhizoids, did not alleviate the ion deficiency‐induced block on development (Fig. 7P–R). DISCUSSION The mature, two‐dimensional fern gametophyte has the essential structural attributes of many, more familiar, terrestrial autotrophs. The green prothallial surface serves as a photosynthetic solar panel, analogous to a leaf. Rhizoids, with their high surface area to volume ratios, seem the logical domain for ion and perhaps water uptake, and play a similar role to root hairs. The results of this study confirm the predicted functional status of these regions and, furthermore, indicate that there is a strict interdependence between the photosynthate‐producing cells and the zone of osmotic ion uptake. In every experimental circumstance, rhizoid growth required that the laminar surface receive quantities of light sufficient for net carbon fixation, and the enlargement of the photosynthetic portion of the prothallus required a supply of monovalent cations sufficient for the osmotically driven expansion of dividing green cells. The level of interdigitation of these functions was such that following removal of media K+ the organism responded by utilizing sodium ions, taken up by rhizoids, in an effort to meet the osmotic demands of cell expansion in the prothallus. The main contribution to our understanding of plant development to emerge from this study is that the structural alterations which occur during the specification of the dedicated light‐ and ion‐absorbing regions in these fern gametophytes are primarily influenced by cell–cell interactions which, in turn, appear to be inseparable from the organism’s requirements for metabolites and osmotic ions. Specifically, the growth of rhizoids and green cells was never observed to be de‐coupled, and, by repositioning the site of 2° rhizoid emergence, the interstitial zone between these regions was maintained within a span of less than 100 µm in light. Taken together, these observations suggest that ultimate dimensions of the longitudinal axis were subject to control of the developing organism according to its physiological requirements for bidirectional transport of photosynthates and nutrient ions. The driving forces for these transport events were not explored in this study, but given that the distances were in the range of one to four cells and that the translocated species were ions/small molecules, it is not unreasonable to suggest that the nutrient streams are diffusion‐based, perhaps augmented by cytoplasmic streaming. Accordingly, my analysis of the range of morphologies that may be elicited in Onoclea gametophytes proceeds from the standpoint that the controlling parameters are the rates of delivery from photosynthetic and ion‐uptake source regions, and the magnitude of the demands these same regions exert on photosynthate and ion supplies in their corresponding roles as sinks. Each of these is discussed, in turn, below. Analysis of the photosynthetic source–sink relationship The cells of the dark‐grown filament, which develops into the prothallial portion of the gametophyte, quickly increased their numbers of chloroplasts in response to even short exposures of light. These cells of the expanding prothallus were thus assumed to function as the photosynthetic source, and the non‐green rhizoids, which were able to grow only when light struck the prothallus, clearly behaved as a sink for some of the photosynthate produced. As shown graphically in Fig. 2 and photographically in most of the other figures, growth of the photosynthetic and non‐photosynthetic portions of the gametophyte, from the earliest cell expansion to the complete prothallus, appeared to be highly coordinated. Even in surgically excised filament cells, subsequent rhizoid emergence and growth was always correlated with the initial swelling and intense packing of chloroplasts in a green cell. The development of gametophytes in light displayed an all‐or‐nothing type of response with a total fluence threshold of approx. 175 000 µm m–2 per 24 h period, which, administered as continuous irradiation, corresponded to a fluence rate of approx. 2 µm m–2 s–1. This level is below the light compensation point reported for many C3 and C4 terrestrial plants, but ferns are remarkably well adapted to low light conditions and this value compares rather well with other determinations of the photosynthetic compensation point in fern gametophytes (Friend, 1975). As reported in earlier studies, it was noted that expansion of the Onoclea filament tip cell (as well as other filament cells) and dramatic increases in chloroplast density were triggered by total fluences well below the apparent photosynthetic threshold for this organism (e.g. 2L22D photoperiod, Fig. 1F). By convention, therefore, these initial events may be categorized as photomorphogenetic phenomena, and there is a considerable body of evidence which shows that phytochrome and a blue light photoreceptor residing in the plasma membrane and/or membranes associated with the nucleus (reviewed in Furuya, 1983) are responsible for these changes in morphology. In Onoclea, however, the photomorphogenetic effects were clearly separable from the more profound changes in form that required the involvement of photosynthetic, chloroplast‐based machinery. It was observed in this study that filaments with somewhat expanded cells and dense populations of chloroplasts could be held in this state by maintaining the organisms in total fluences below the apparent photosynthetic compensation point. All further processes of cell expansion, cell division and cell differentiation (e.g. 2° rhizoid production) leading to prothallial form, required quantities of light sufficient for net photosynthesis. With regard to earlier studies that reported the presence of media sucrose was necessary to sustain, or enhance, the growth of fern gametophytes in very low light conditions (Miller and Miller, 1961; Kato, 1967), development of Onoclea gametophytes in the present experiments was tied strictly to light conditions that supported photosynthesis, whether or not sucrose was present in the medium. It was also never observed that a supplementary carbon source could induce these ferns to develop as heterotrophs. Growth in darkness was apparently sustained by mobilization of reserves present in the spore; surgical removal of the spore end resulted in the rapid death of organisms in darkness, even with sucrose‐supplemented media, but spore‐less explants developed into prothalli on sucrose‐free media in the light. Furthermore, in considering these earlier studies, is not clear why it was apparently believed that an exogenously applied carbon source might be a viable substitute for the internally produced photosynthates. The green portion of the gametophyte has a waxy cuticle (Wada and Staehelin, 1981), which would be expected to impede uptake, and the rhizoids, by all accounts, show little growth in darkness, irrespective of sugar availability in the media. Reports of species‐specific, or experimentally induced, rhizoid or prothallial growth in low intensity light (Howland and Boyd, 1974) or darkness (Miller and Miller, 1970) must mean that there are circumstances under which the transition to two‐dimensional growth can preferentially utilize prepackaged photosynthates in the spore. Establishing causal links between an organism’s structure and its functions is always desirable; however, pursuing the apparent relationships between the changes in gametophyte structure and the dynamics of photosynthate allocation is encumbered by some deficits in our knowledge which can not be circumvented. First, we do not know the identity of the primary transportable species of photosynthates in ferns. The relevant literature on phloem loading and unloading in higher plants is reasonably unified in the belief that sucrose is the membrane‐transported and vascularly translocated molecule (Ziegler, 1975; Giaquinta, 1983). Secondly, the rate of photosynthesis and photosynthate allocation has never been measured in fern gametophytes. In other systems estimates of net carbon incorporation have been derived from a determination of changes in dry weight, but given the gametophyte’s diminutive size, this would have to be based on a sizeable, developmentally mixed population of organisms. Furthermore, without some independent measure of the rate of carbon uptake, or the rate of consumption from cell divisions and molecular syntheses, these calculations would not bear on the key issue of how domains of the gametophyte detect and then compensate for inequities in supply and demand of fixed carbon. Finally, although the path between source and sink regions is greatly simplified by the absence of vascular tissue, the extent to which symplastic and apoplastic modes of translocation are involved in distribution of photosynthate is not known. Analysis of the osmotic ion source–sink relationship Rhizoids of fern gametophytes are positioned at the base of the expanding prothallus and have the appropriate morphological features of nutrient‐absorbing structures; it is therefore surprising that there is little direct evidence related to their function. Comparing protonema and rhizoid cells on the bases of cell wall composition and the ability to take up vital stains, Smith (1972a) found that rhizoids were much more permeable to water‐soluble dyes, and that the cell behaved as a cation‐exchange medium. Smith (1972b) also showed that marker enzymes associated with phosphate metabolism were present in gametophyte rhizoids, but approaches which established the absorptive capabilities of root hairs, such as ion deficiency experiments or studies with radio‐tracers, have not been applied to fern rhizoids. Because the fern gametophyte grows in direct contact with the medium (in these experiments they were essentially grown in submerged conditions), one cannot safely assume, on the basis of location and appearance, that rhizoids are the sole site for ion uptake in this organism. In fact, the ability of green, thallus‐like surfaces to take up ions directly from media is apparently widespread in the plant kingdom. Examples include ion uptake through the laminar surfaces of the green alga, Ulva (West and Pitman, 1967) and foliar ion uptake exhibited by various terrestrial higher plants (Franke, 1967; Kannan, 1980). Results from the present study suggest that rhizoids appear to be the primary, if not the sole, site of uptake for K+ in Onoclea gametophytes. This is most clearly demonstrated in the monovalent cation ion deficiency experiments in which inhibition of prothallial expansion was alleviated by supplying K+ to the rhizoid end of the organism (Fig. 7M–O). As for most plants, ion uptake systems in fern gametophytes are apparently centred around the accumulation of K+ which, at typical cellular concentrations of 100 mm, is the chief osmotic agent that governs cell turgor and, in turn, turgor‐driven cellular expansion. The internal concentration of K+ in gametophyte cells has not been determined, but measurements of membrane potentials (Racusen and Cooke, 1982), which are highly dependent on outward K+ diffusion, produced values for filament cells that are similar to those seen in a variety of plant cells with known internal concentrations of approx. 100 mm K+ (Findlay and Hope, 1976). The importance of media‐supplied K+ as the osmotic agent that regulates expansion of green cells in the gametophyte is highlighted in experiments where only NaCl is provided in the agarose medium. Although expansion of the prothallus was noticeably restrained, Onoclea gametophytes were still capable of producing the characteristic, heart‐shaped form, apparently because they were able, under K+ deficiency conditions, to substitute Na+ as the osmotic agent. Similar trade‐offs have been documented in other systems, such as the growth of cultured leaf discs (Marschner and Possingham, 1975), and the operation of stomata (Raghavendraet al., 1976). In these cases it appears that the requirement for a monovalent cation species, as part of the regulatory mechanism for cell turgor, has priority over the role that K+ has in stabilizing enzymes (Clarkson and Hanson, 1980). Eliminating both K+ and Na+ from the nutrient solution was the only medium condition which completely (but reversibly) inhibited development of the prothallus. Concentrations of Ca2+ and other nutrient ions were apparently high enough to support growth of a complete prothallus. The effects of monovalent cation deficiency were not, however, immediately obvious; development of the prothallus kept pace with that of control organisms for approx. 10 d. Thereafter, the resulting, roughly triangular structure ceased growth and, over extended time, began to exhibit abnormalities of organelle clumping and physical discontinuity among cells (Fig. 7H, L). The deleterious effects of monovalent cation starvation were ultimately reversed by application of either K+ or Na+; it is possible that the early, unimpeded phase of growth was supported by mobilizing K+ and/or Na+ within the spore. The following calculation suggests that this could indeed be the case. Assuming a combined K+ and Na+ concentration within the spore of 250–300 mm (derived from values in Wayne and Hepler, 1985), and a K+/Na+ concentration inside expanded prothallial cells of 100 mm, it follows that the ion contents of the spore could support a prothallial structure 2·5 to 3 times the volume of the spore. Comparison of the spore and prothallus volumes of gametophytes grown under K+/Na+ deficiency, such as those shown in Figs 6K and 7G, are consistent with this prediction. As is the case with photosynthetic production and dispersal, further analysis of ion source/sink interactions is impeded by a lack of knowledge concerning the rates of ion uptake from solution, and the path of intercellular transport within the developing gametophyte. However, since the ion sink regions accumulate, but cannot metabolically transform the transported entity, it is possible to use changes in volumes of the prothallus and total rhizoids to estimate the demand for K+. For example, assuming that the internal cellular concentration of K+ is approx. 100 mm, then computing the change in total volume of cells (prothallus + rhizoids) between day 4 and day 10 in Fig. 2, and finally coupling these factors with an estimate of the mean surface area of rhizoids (from aggregate length), a net influx for K+ of approx. 5 pm cm–2 s–1 is obtained. This value compares favourably with previous determinations of 1–2·5 pm cm–2 s–1 for K+ influx in freshwater algae (MacRobbie, 1974) and higher plants (Higinbothamet al., 1967). Since the expanding prothallus comprises a volume six to eight times that of the rhizoids, it follows that approx. 85 % of the rhizoid K+ influx is transported to prothallial cells. Determination of polarity and pattern formation coincide with metabolite transport in the gametophyte The light‐induced changes in the position of new cell walls in green apical cells of the fern gametophyte have always held a special prominence for those who utilize this organism as a model system to study plant development, and considerable effort has been directed towards an understanding of the apparently unique physiological (Daviset al., 1974; Racusen and Cooke, 1982), cytoskeletal (Stetler and DeMaggio, 1972; Murata and Wada, 1989), chemical (Raghavan, 1968; Smithet al., 1973) and geometric (Cooke and Paolillo, 1980b) properties that allow it to function as a sort of developmental ‘ground‐cell’ for the organism. In contrast, few previous studies have focused on the ways in which the changes in structure might be related to the overall function of the organism. A prerequisite for development in multicellular organisms is a mechanism for determining the polarity of the main body axis. During germination in ferns, axis polarity is apparently established before the emergence of the rhizoid and protonematal tip cells, indicating that chemical, electrical or structural features of the spore govern the initial placement of the ion‐ and light‐absorbing regions. The regeneration experiments reported here indicate further that remnants of severed filaments are capable, within the span of a few cell divisions, of re‐establishing the poles for a new axis. The determinants of axis polarity in regenerating gametophytes are unknown, but it was consistently observed that the re‐modelled axis comprised two to four green cells that tended not to exceed 200 µm, between the site of 2° rhizoid emergence and the tip cell—a distance comparable with that in other systems in which a diffusible morphogen is involved in the determination of polarity (Wolpert, 1996). As the re‐establishment of polarity in severed gametophyte filaments was dependent on quantities of light sufficient to support photosynthesis, it is conceivable that activation of the cellular processes that increase the supply of available metabolites also establishes, in tandem, the developmental polarity for the gametophyte axis. With increasing transport activity, this initial determination of polarity is reinforced by the biochemical and structural changes attending the production of more nutrient‐absorbing and photosynthetic cells. Non‐vascularized, terrestrial plants with thallus‐type architecture possess inherent structural limitations on their ability to efficiently move photosynthates and ions over long distances. The available escort paths involve either multiple cell‐to‐cell transfers, in this case necessitating coordination of many spatially localized membrane transporters to effect the appropriate direction of transport through the tissue, or movement through extracellular spaces, which, again, would require some mechanism to provide a driving force and to establish direction. In a mature prothallus, the path from cells in the meristematic notch to the rhizoids at the base may consist of as many as eight comparably sized chlorocytes (for example, see Fig. 6D), and, given that the developing prothallus consistently maintains an even shorter distance (two to four cells) between the rhizoids and the expanding two‐dimensional chlorocyte array, this may, in fact, represent the physical limit over which transport is possible in this system. In this context it is noteworthy that in the roots (Drew, 1987), sink tissues (Eschrich, 1989; Woodet al., 1997) and leaves (Philpott, 1953; Giaquinta, 1983) of the much better‐studied vascular plants, it is most common to find cell‐to‐cell transport across fields consisting of fewer than ten parenchymous cells, and it remains an issue of some debate as to the relative involvement of symplastic (plasmodesmatal) and apoplastic (membrane transport) mechanisms (Van Bel, 1993). With this demonstration that the bi‐directional flow of osmotic ions and photosynthates has cause‐and‐effect roles in the development of the prothallus, it seems reasonable to suggest that the orientation of cell divisions, believed to be the main patterning mechanism in the gametophyte, might profitably be examined in light of their potential contribution to the transport of ions or photosynthates. ACKNOWLEDGEMENTS The author gratefully acknowledges Susan Klinedinst for assistance in developing some of the techniques used in this study, Melissa Dooley for help in searching the literature and Dr F. Mark Schiavone for comments during revision. View largeDownload slide Fig. 1. A–D, Development of gametophytes grown under different intensities of continuous white light. Two‐day‐old germinated spores were transferred to Petri dishes with four (f × 4), two (f × 2), one (f × 1) or no (full light; FL) neutral density filters covering the plate lid (see Materials and Methods). E–H, Development of gametophytes grown under different photoperiods of white light. Two‐day‐old germinated spores were exposed to photoperiods of continuous darkness (24D), 2 h light (2L22D), 4 h light (4L20D) and continuous light (24L). All photographs were taken 20 d after the start of the experiment. Bar = 100 µm. View largeDownload slide Fig. 1. A–D, Development of gametophytes grown under different intensities of continuous white light. Two‐day‐old germinated spores were transferred to Petri dishes with four (f × 4), two (f × 2), one (f × 1) or no (full light; FL) neutral density filters covering the plate lid (see Materials and Methods). E–H, Development of gametophytes grown under different photoperiods of white light. Two‐day‐old germinated spores were exposed to photoperiods of continuous darkness (24D), 2 h light (2L22D), 4 h light (4L20D) and continuous light (24L). All photographs were taken 20 d after the start of the experiment. Bar = 100 µm. View largeDownload slide Fig. 2. Changes in aggregate volume of the prothallus (triangles) and rhizoids (circles) during gametophyte development over 10 d. The experiment was started with organisms germinated 2 d earlier in continuous light. Values were calculated by applying simple geometric formulae to measurements of length, width and thickness of representative gametophytes grown in continuous white light. Bars are means ± s.d. for 20 organisms. View largeDownload slide Fig. 2. Changes in aggregate volume of the prothallus (triangles) and rhizoids (circles) during gametophyte development over 10 d. The experiment was started with organisms germinated 2 d earlier in continuous light. Values were calculated by applying simple geometric formulae to measurements of length, width and thickness of representative gametophytes grown in continuous white light. Bars are means ± s.d. for 20 organisms. View largeDownload slide Fig. 3. Position and growth of secondary rhizoids during development of selected gametophytes with 100 µm (A), 200 µm (B) and 250 µm (C) filament lengths. Gametophytes with specific filament lengths were transferred from various‐aged cultures of organisms grown under a 2L22D photoperiod. The x‐axis of each graph represents the length of the filament axes, with the base of the filament at the origin and the apical cell at the highest x‐axis value. For each 25 µm interval along the filament, the lengths of all rhizoids were summed, and this aggregate rhizoid length plotted on the y‐axis. Open bars show rhizoid position and growth after 4 d in full light; hatched bars show rhizoid growth after 20 d in full light. Results are means ± s.d. of eight organisms for each plot. View largeDownload slide Fig. 3. Position and growth of secondary rhizoids during development of selected gametophytes with 100 µm (A), 200 µm (B) and 250 µm (C) filament lengths. Gametophytes with specific filament lengths were transferred from various‐aged cultures of organisms grown under a 2L22D photoperiod. The x‐axis of each graph represents the length of the filament axes, with the base of the filament at the origin and the apical cell at the highest x‐axis value. For each 25 µm interval along the filament, the lengths of all rhizoids were summed, and this aggregate rhizoid length plotted on the y‐axis. Open bars show rhizoid position and growth after 4 d in full light; hatched bars show rhizoid growth after 20 d in full light. Results are means ± s.d. of eight organisms for each plot. View largeDownload slide Fig. 4. Acropetal migration of secondary rhizoids during development of a filamentous gametophyte with an axial length between the apical cell and the spore of approx. 200 µm. The organism at day 0 (d0) had been grown in the dark for 10 d and then transferred to continuous light. Note progression of secondary rhizoid emergence at locations progressively closer to the expanding prothallus. Bar = 100 µm. View largeDownload slide Fig. 4. Acropetal migration of secondary rhizoids during development of a filamentous gametophyte with an axial length between the apical cell and the spore of approx. 200 µm. The organism at day 0 (d0) had been grown in the dark for 10 d and then transferred to continuous light. Note progression of secondary rhizoid emergence at locations progressively closer to the expanding prothallus. Bar = 100 µm. View largeDownload slide Fig. 5. A–C, Prothallial development from a very long (>200 µm) filamentous gametophyte. The organism at d0 had been grown in the dark for 40 d and then transferred to continuous light. D–F, Prothallial regeneration from an excised filament containing the apical cell. Explant at d0 was severed from a 30‐d‐old, dark‐grown gametophyte; the apical cell is to the left and the cut end to the right. From d0 onward, the explant was in continuous light. G–I, Prothallial regeneration from an excised segment of subapical cells; both the apical cell and the spore end have been removed. Explant at d0 was excised from a 30‐d‐old dark‐grown gametophyte; from d0 onward, the explant was in continuous light. Rhizoids were present at d5 onward, but protruded below the structure and are not visible in photographs until d19. Bar = 100 µm. View largeDownload slide Fig. 5. A–C, Prothallial development from a very long (>200 µm) filamentous gametophyte. The organism at d0 had been grown in the dark for 40 d and then transferred to continuous light. D–F, Prothallial regeneration from an excised filament containing the apical cell. Explant at d0 was severed from a 30‐d‐old, dark‐grown gametophyte; the apical cell is to the left and the cut end to the right. From d0 onward, the explant was in continuous light. G–I, Prothallial regeneration from an excised segment of subapical cells; both the apical cell and the spore end have been removed. Explant at d0 was excised from a 30‐d‐old dark‐grown gametophyte; from d0 onward, the explant was in continuous light. Rhizoids were present at d5 onward, but protruded below the structure and are not visible in photographs until d19. Bar = 100 µm. View largeDownload slide Fig. 6. Gametophyte development under conditions where specific media ions were removed. Organisms at d0 had been grown in a 2L22D photoperiod for 4 d, then transferred to continuous light. A–D (Z/agar; ‘zero agar’) shows that normal development occurs when organisms were placed on Bacto‐agar with no added salts. E–H (Na/ose; ‘sodium agarose’) shows that development of the prothallus is slower and smaller when organisms were placed on a low‐salt agarose medium containing only 0·1 mm NaCl. I–L (Z/ose; ‘zero agarose’) shows inhibition of prothallus development at about d7 (J) in organisms placed on low‐salt agarose medium with no added salts. Lower‐diagonal portion of L shows the resumption of prothallus development after a further 8 d (d35), when 0·1 mm K+ was added to the d27, Z/ose organism directly above it. Bar = 100 µm. View largeDownload slide Fig. 6. Gametophyte development under conditions where specific media ions were removed. Organisms at d0 had been grown in a 2L22D photoperiod for 4 d, then transferred to continuous light. A–D (Z/agar; ‘zero agar’) shows that normal development occurs when organisms were placed on Bacto‐agar with no added salts. E–H (Na/ose; ‘sodium agarose’) shows that development of the prothallus is slower and smaller when organisms were placed on a low‐salt agarose medium containing only 0·1 mm NaCl. I–L (Z/ose; ‘zero agarose’) shows inhibition of prothallus development at about d7 (J) in organisms placed on low‐salt agarose medium with no added salts. Lower‐diagonal portion of L shows the resumption of prothallus development after a further 8 d (d35), when 0·1 mm K+ was added to the d27, Z/ose organism directly above it. Bar = 100 µm. View largeDownload slide Fig. 7. Demonstration that the rhizoid end of the gametophyte, and not the prothallial end, is the primary site of ion uptake. A–L shows development of prothalli on control media (agar with all nutrient ions; A–D), on low‐salt agarose with no added salts (Z/ose, E–L). Organisms were grown in a 2L22D photoperiod for 7 d, and then transferred to continuous light at d0. Ion deficiency symptoms are evident in both the Z/ose treatments by d10 (G and K). M–R shows the response of the same two deficient organisms when K+ is added to the cellophane chip overlying the rhizoid end (Z/ose a, M–O), or to the cellophane chip overlying the prothallus end (Z/ose b, P–R). Bar = 100 µm. View largeDownload slide Fig. 7. Demonstration that the rhizoid end of the gametophyte, and not the prothallial end, is the primary site of ion uptake. A–L shows development of prothalli on control media (agar with all nutrient ions; A–D), on low‐salt agarose with no added salts (Z/ose, E–L). Organisms were grown in a 2L22D photoperiod for 7 d, and then transferred to continuous light at d0. Ion deficiency symptoms are evident in both the Z/ose treatments by d10 (G and K). M–R shows the response of the same two deficient organisms when K+ is added to the cellophane chip overlying the rhizoid end (Z/ose a, M–O), or to the cellophane chip overlying the prothallus end (Z/ose b, P–R). Bar = 100 µm. References BanksJA. 1999 Gametophyte development in ferns. Annual Review of Plant Physiology and Plant Molecular Biology  50: 163–186. Google Scholar ClarksonDT, Hanson JB. 1980. The mineral nutrition of higher plants. Annual Review of Plant Physiology  31: 239–298. Google Scholar CookeTJ, Paolillo DJ. 1980a. The control of the orientation of cell divisions in fern gametophytes. American Journal of Botany  67: 1320–1333. Google Scholar CookeTJ, Paolillo DJ. 1980b. Dark growth and the associated phenomenon of age‐dependent photoresponses in fern gametophytes. Annals of Botany  45: 693–702. Google Scholar CookeTJ, Racusen RH. 1982. Cell expansion in the filamentous gametophyte of the fern Onoclea sensibilis L. Planta  155: 449–458. Google Scholar DavisBD, Chen JCW, Philpott M. 1974 The transition from filamentous to two‐dimensional growth in fern gametophytes. IV. Initial events. American Journal of Botany  61: 722–729. Google Scholar DrewMC. 1987 Function of root tissues in nutrient and water transport. In: Gregory PJ, Lake JV, Rose DA, eds. Root development and function. Cambridge: Cambridge University Press, 71–101. Google Scholar DyerAF. 1979. The experimental biology of ferns. New York: Academic Press. Google Scholar DyerAF, King MAL. 1979. Cell division in fern protonemata. In: Dyer AF, ed. The experimental biology of ferns. New York: Academic Press, 307–354. Google Scholar EschrichW. 1989. Phloem unloading of photoassimilates. In: Baker DA, Milburn JA, eds. Transport of photoassimilates. Essex: Longman Scientific and Technical, 206–263. Google Scholar FindlayGP, Hope AB. 1976. Electrical properties of plant cells: methods and findings. In: Lüttge U, Pitman MG, eds. Encyclopedia of plant physiology (new series), Transport in plants, II A. Cells. New York: Springer‐Verlag, 53–92. Google Scholar FrankeW. 1967. Mechanism of foliar penetration of solutions. Annual Review of Plant Physiology  18: 281–300. Google Scholar FriendDJC. 1975. Adaptation and adjustment of photosynthetic characteristics of gametophytes and sporophytes of Hawaiian tree‐fern (Cibotium glaucum) grown at different irradiances. Photosynthetica  9: 157–164. Google Scholar FuruyaM. 1983. Photomorphogenesis in ferns. In: Shropshire W, Mohr H, eds. Encyclopedia of plant physiology (new series), Photomorphogenesis, 16B. New York: Springer‐Verlag, 569–600. Google Scholar GiaquintaRT. 1983 Phloem loading of sucrose. Annual Review of Plant Physiology  34: 347–387. Google Scholar HiginbothamN, Etherton B, Foster RJ. 1967. Mineral ion contents and cell transmembrane electropotentials of pea and oat seedling tissue. Plant Physiology  42: 37–46. Google Scholar HowlandGP, Boyd EL. 1974. Genetic control of photomorphogenesis. Isolation of nonfilamentous mutants after gamma irradiation of Pteridium aquilinum spores. Radiation Botany  14: 281–285. Google Scholar KannanS. 1980. Mechanism of foliar uptake of plant nutrients: Accomplishments and prospects. Journal of Plant Nutrition  2: 717–735. Google Scholar KatoY. 1967. Physiological and morphogenetic studies of fern gametophytes and sporophytes in aseptic culture. VIII. Behavior of Pteris vittata gametophytes on sugar and complex media in red light. Planta  77: 127–134. Google Scholar MacRobbieEAC. 1974. Ion uptake. In: Stewart WDP, ed. Algal physiology. Los Angeles: University of California Press, 676–713. Google Scholar MarschnerH, Possingham JV. 1975. Effect of K+ and Na+ on growth of leaf discs of sugar beet and spinach. Zeitschrift für Pflanzenphysiologie 75: 6–16. Google Scholar MillerJH, Miller PM. 1961 The effect of different light conditions and sucrose on the growth and development of the gametophyte of the fern, Onoclea sensibilis. American Journal of Botany  48: 154–159. Google Scholar MillerJH, Miller PM. 1970. Unusual dark‐growth and antheridial differentiation in some gametophytes of the fern Onoclea sensibilis. American Journal of Botany  57: 1245–1248. Google Scholar MurataT, Wada M. 1989. Organization of cortical microtubules and microfibril deposition in response to blue‐light‐induced apical swelling in a tip‐growing Adiantum protonema cell. Planta  178: 334–341. Google Scholar OotakiT, Furuya M. 1969. Experimentally induced apical dominance in protonemata of Pteris vittata. Embryologia  10: 284–296. Google Scholar PhilpottJ. 1953. A blade tissue study of leaves of forty seven species of Ficus. Botanical Gazette  115: 15–35. Google Scholar RacusenRH, Cooke TJ. 1982. Electrical changes in the apical cell of the fern gametophyte during irradiation with photomorphogenetically active light. Plant Physiology  70: 331–334. Google Scholar RacusenRH, Ketchum KA, Cooke TJ. 1988. Modifications of extracellular electric and ionic gradients preceding the transition from tip growth to isodiametric expansion in the apical cell of the fern gametophyte. Plant Physiology  87: 69–77. Google Scholar RaghavanV. 1968. RNA and protein metabolism in the particulate fractions of the gametophytes of bracken fern during growth in red and blue light. Planta  81: 38–48. Google Scholar RaghavanV. 1989. Developmental biology of fern gametophytes. Cambridge: Cambridge University Press. Google Scholar RaghavendraAS, Rao IM, Das VSR. 1976. Replaceability of potassium by sodium for stomatal opening in epidermal strips of Commelina benghalensis. Zeitschrift für Pflanzenphysiologie  80: 36–42. Google Scholar SmithDL. 1972a. Localization of phosphatases in young gametophytes of Polypodium vulgare L. Protoplasma  74: 133–148. Google Scholar SmithDL. 1972b. Staining and osmotic properties of young gametophytes of Polypodium vulgare L. and their bearing on rhizoid function. Protoplasma  74: 465–479. Google Scholar SmithDL, Robinson PB, Govier RN. 1973. Growth factors produced by gametophytes of Polypodium vulgare L. grown under red and blue light. New Phytologist  72: 1261–1268. Google Scholar StetlerDA, DeMaggio AE. 1972. An ultrastructural study of fern gametophytes during one‐ to two‐dimensional development. American Journal of Botany  59: 1011–1017. Google Scholar TilneyLG, Cooke TJ, Connelly PS, Tilney MS. 1990 The distribution of plasmodesmata and its relationship to morphogenesis in fern gametophytes. Development  110: 1209–1221. Google Scholar Van BelAJE. 1993. Strategies of phloem loading. Annual Review of Plant Physiology and Plant Molecular Biology  44: 253–281. Google Scholar WadaM, Staehelin LA. 1981. Freeze‐fracture observations on the plasma membrane, the cell wall and the cuticle of growing protonemata of Adiantum capillus‐veneris L. Planta  151: 462–468. Google Scholar WayneR, Hepler PK. 1985. The atomic composition of Onoclea sensibilis spores. American Fern Journal  75: 12–18. Google Scholar WestKR, Pitman MG. 1967 Ionic relations and ultrastructure in Ulva lactuca. Australian Journal of Biological Sciences  20: 901–914. Google Scholar WolpertL. 1996. One hundred years of positional information. Trends in Genetics  12: 359–364. Google Scholar WoodRM, Offler CA, Patrick JW. 1997. The cellular pathway of short‐distance transfer of photosynthates and potassium in the elongating stem of Phaseolus vulgaris L. A structural assessment. Annals of Botany  79: 89–100. Google Scholar ZieglerH. 1975. Nature of transported substances. In: Zimmermann MH, Milburn JA, eds. Encyclopedia of plant physiology (new series), Transport in plants, I. Phloem transport. New York: Springer‐Verlag, 59–100. Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Botany Oxford University Press

Early Development in Fern Gametophytes: Interpreting the Transition to Prothallial Architecture in Terms of Coordinated Photosynthate Production and Osmotic Ion Uptake

Annals of Botany , Volume 89 (2) – Feb 1, 2002

Loading next page...
 
/lp/oxford-university-press/early-development-in-fern-gametophytes-interpreting-the-transition-to-7yqxlOMIZ7

References (41)

Publisher
Oxford University Press
ISSN
0305-7364
eISSN
1095-8290
DOI
10.1093/aob/mcf032
Publisher site
See Article on Publisher Site

Abstract

Abstract Gametophytes of Onoclea sensiblis L. were grown under various light and media‐ion conditions to gain a better understanding of the source/sink relationships between photosynthetic and ion‐absorbing cells. There was a clear interdependency between green cell and rhizoid functions, such that the growth and development of the rhizoids was completely dependent on the internal delivery of photosynthates from green cells, and conversion of the one‐dimensional filament into the two‐dimensional prothallus required monovalent cations that could only be provided by rhizoid uptake. The need for monovalent cations was related to osmotic demands of dividing and expanding cells; prothallial development was blocked by monovalent cation deficiency, and the system resorted to Na+ uptake to support cell expansion when K+ was absent. Surgical excisions of filament cells further demonstrated the high degree of coordinated growth between the light‐absorbing and ion‐absorbing regions. It was also learned that excised sub‐apical cells of the protonemata, like the intensively studied apical cell, were capable of remodelling remnants of the filament into a normal prothallus. Key words: Fern gametophyte, Onoclea sensibilis L., development, ion uptake, photosynthate. Received: 20 August 2001; Returned for revision: 9 October 2001; Accepted: 12 November 2001. INTRODUCTION The fern gametophyte is a small plant that exists as a prolonged intermediate in the fern life cycle, between the germination of a spore and the mature sporophyte. Following its emergence from a spore, it grows from two cells into a distinctively shaped structure containing several hundred cells. In the course of development one end of the structure becomes a flattened, single‐layer of chloroplast‐containing cells, resembling a leaf. The other end manifests an array of long, filamentous rhizoids that have the appearance, and apparently serve the same function, as root hairs. The rhizoids are differentiated directly from basal green cells (as a consequence of an unusual, highly asymmetric cell division); thus there are few intervening cells and no vascular tissue between the photosynthetic zone and the region designated for ion and water uptake. General reviews of fern gametophyte anatomy and development include Dyer (1979), Raghavan (1989) and Banks (1999). Both germination and development of later form are influenced by the intensity and spectral quality of light and, in the laboratory setting, gametophytes of many ferns can be induced to exhibit a considerable range of morphologies, from long, uniseriate, tip‐growing filaments (generally achieved in extended periods of low‐intensity light, or darkness) to prothalli with various length‐to‐width ratios (produced by adjusting light intensity or photoperiod). Regardless of the experimentally induced form of the green‐celled axis, however, it is nearly always observed that exposure to light and/or photoperiod conditions that approximate those in the natural environment leads to new rounds of cell divisions and the emergence of the ‘true’ prothallial shape. Utilizing a variety of fern species, researchers have shown that an early, pivotal event in the formation of the prothallus involves the re‐orientation of a cell plate (from a transverse to a longitudinal alignment) during division of the apical chlorocyte (reviewed in Dyer and King, 1979). This step has the effect of altering the pattern of growth from extension in one dimension (sustaining a filamentous architecture) to expansion in two dimensions (producing a planar surface). Much of the research over the past 50‐plus years has understandably focused on the nature of photoreception and the mechanisms that govern the timing and positioning of divisions in this cell. While these undertakings have been most useful in furnishing insights about the nature and cellular location [reviewed in Furuya (1983) and Raghavan (1989)] of the pigments that trigger the fern photoresponses, intriguing observations concerning planes of minimum surface area (Cooke and Paolillo, 1980a), cytoskeletal proteins (Stetler and DeMaggio, 1972), localized ion gradients (Racusenet al., 1988), bioelectrical fields (Racusenet al., 1988) and the frequencies of plasmodesmata (Tilneyet al., 1990) have thus far failed to yield a coherent framework of events that explains how the transition from one‐dimensional to two‐dimensional growth is accomplished. Moreover, since division of the apical cell is considered to be the gateway to the emergence of thallial architecture in the fern gametophyte, we are presently unable to derive potentially useful insights from this comparatively simple system which might help guide the study of analogous, but more complex, morphogenetic phenomena (such as leaf development) in vascular plants. Here I re‐examine the formation of the prothallus in a fern gametophyte in the context of a ‘whole‐plant’ developmental response. In particular, the focus of this study was to examine how the allocation of resources between photosynthetic and non‐photosynthetic regions was related to attainment of the overall form of the organism. Embedded within this undertaking were the following specific aims: (1) to relate, quantitatively, the amount of photosynthetically active light to the changes in prothallial and rhizoid biomass; (2) to examine the correlation in growth rates between the emerging prothallus (presumed photosynthetic source) and the proliferation of rhizoids (presumed sink); (3) to determine how the length of the interstitial region between source and sink influenced the delivery of photosynthates and ions during development; and (4) to establish the rhizoids as the primary site for nutrient ion uptake to satisfy the osmotic requirements of both non‐green and green cells. MATERIALS AND METHODS Gametophyte culture Fertile fronds of Onoclea sensibilis L. were collected in January near the upper branch of the Patapsco River in Carroll county, Maryland, USA. They were packed into zip‐closure plastic bags and stored at –35 °C until needed. A supply of spores adequate for approx. 6 months of experiments could be obtained from ten fronds, and the procedure to release them began by surface‐sterilizing fronds in 15 % laundry bleach, and then drying them overnight on aluminium foil in a clean, but non‐operating, laminar flow hood. Spore cases were removed by hand from the dried fronds into a mortar, and then ground to a fine powder with the pestle. This material, containing spores and debris, was placed into a screw‐top plastic centrifuge tube and stored at –35 °C until needed for an experiment. To start a culture of gametophytes, spore‐containing powder was shaken through a small sieve that had been made from 41 µm nylon mesh. Approx. 0·5 ml of this screened spore powder was collected in a 15 ml plastic centrifuge tube, and then wetted, with agitation, in a solution of 0·1 % triton X‐100 for 5 min. A clinical centrifuge operating at 300 g for 15 s was used to pellet the spores, and the wetting solution was discarded. The spores in the pellet were dispersed and surface‐sterilized in a solution of 0·1 % triton X‐100 and 15 % liquid laundry bleach for 4 min. All the remaining steps were carried out using a laminar flow hood and sterile technique. The tube was next centrifuged for 15 s at 300 g, the sterilizing solution decanted and the spores washed twice with sterile water, using the same techniques. After the second wash, the spores were diluted to a desired inoculation density (usually by adding 10 ml of sterilized water) of approx. 1000 spores ml–1. Surface sterilized spores were transferred to 1 % agar or agarose medium in 15 cm plastic Petri plates. The control medium for gametophyte growth was based on the APW‐6 recipe provided by Cooke and Racusen (1982). It consisted of 1 mm 2‐morpholinoethanesulfonic acid (adjusted to pH 6), 0·1 mm NaCl, 0·1 mm KCl and 0·1 mm CaCl2, solidified with 1 % Bacto‐agar (Difco Laboratories, Detroit, MI, USA) or low‐salt agarose (Ultra pure agarose 5510UA, BRL/Life Technologies, Rockville, MD, USA). As described in the Results, gametophytes grew equally well on a Bacto‐agar substrate with no added salts, apparently because this general purpose agar had sufficient ion content to support the osmotic and nutritional needs of these plants. Using values provided by the manufacturer (expressed in salt and ash content), it was calculated that a 1 % agar mixture contained, in the bound and unbound state, approx. 100 µm Na+, 5 µm K+ and 1 µm Ca2+. For ion deficiency experiments, therefore, certain salts were eliminated and the solidifying agent was low‐salt agarose, an agar‐type matrix in which the relevant cations above were present in concentrations at least five‐times lower. During the initial phases of this study, sucrose was present in the agar media at concentrations of 0·1 to 1 %. As comparisons of gametophyte growth with and without sucrose in darkness or various light conditions showed no significant differences, this ingredient was not included in the media in any of the experiments for which data are presented (see also Results and Discussion). Petri dishes were sealed with strips of Parafilm and placed in clear plastic, zip‐closure bags to inhibit drying and contamination. These bags were then placed in continuous cool white light at 25 µm m–2 s–1 for 2 d to stimulate germination. At this point plates of 2‐d‐old gametophytes were transferred to darkness or to a variety of light and photoperiod regimes as described below. Terminology During development of the gametophyte the organism undergoes dramatic changes in morphology, and various terms are needed to distinguish newly differentiated cells and changing tissue infrastructure. The protonemata refers to the filamentous stage of gametophyte development. This structure may be quite elongated in dark‐grown organisms and comprises a tip‐growing apical cell and other sub‐apical, cylindrical cells. The chloroplast‐containing cells of the protonemata, or filament, may be referred to as chlorocytes, and these green cells also make up the prothallus, the planar two‐dimensional structure that develops from a protonematal initial in light. Rhizoids are filamentous, single‐celled, non‐green cells that elongate by tip growth. The initial, primary rhizoid emerges from the spore during germination. Secondary (2°) rhizoids are those that emerge from other cells of the protonemata or the prothallus. The term ‘spore end’ is used in this paper to designate the original site of emergence of the protonemata, which, during the extension of the filament, represents the basal end of the filament opposite to the tip cell. Light and photoperiod treatments All light treatments in these studies were derived from adjustments to the timing or intensity of the cool white fluorescent source described above. To expose gametophytes to particular photoperiodic conditions, the bagged Petri plates were placed in a light‐tight box over which was installed the fluorescent source. Room air was circulated through the box with a fan and the temperature range over the lights‐on/lights‐off periods was 24–28 °C. The light source and fan were connected to a 24 h timer. For experiments in which gametophytes were exposed to lower intensities of light, individual Petri plates were wrapped in aluminium foil, and a rectangular window was created by removing a section of the foil covering the lid. This opening was then covered with one to six layers of neutral‐density filters made from a sheet of developed X‐ray film. The filters were fastened to the foil‐covered lid with opaque tape. The films could be empirically adjusted to different OD (optical density) values by exposing them for various lengths of time to a red darkroom safelight. The filters in these experiments were made from a film that exhibited OD490 of 0·585 and OD640 of 0·565, as measured in a scanning spectrophotometer. The average of these values, 0·575, was used to calculate intensities of white light produced by the following combinations of filters (all in µm m–2 s–1): 6·75 (one filter), 1·90 (two filters), 0·52 (four filters) and less than 0·01 (six filters). Experimental set‐up and photo‐documentation All data for these experiments were collected by photographing individual developing gametophytes every day or two, over a period of approx. 30 d. Since these transitions in form are complex and do not readily lend themselves to quantification by simple parameters such as numbers of cells or dimensions of structures, much of the Results section consists of photographic sequences that capture the progressive changes arising from cell divisions, cell expansion and cell differentiation. Each of the photomontages, therefore, is the record of a single, representative gametophyte; in every instance I obtained complete photographic records of comparable development in at least six replicate organisms. Working with a 25× dissecting microscope that was placed in a laminar flow hood, individual gametophytes were lifted from the agar surface on the tip of a number 11 scalpel and promptly transferred to the agar surface of a fresh plate. To counteract the desiccating environment of the laminar flow hood, an aluminium foil shield was taped behind the stage of the microscope and, with practice, one could transfer the gametophyte with some adherent agar, which, following transfer, was removed from the new agar surface. To improve visualization of gametophytes on an agar surface, a technique was developed for floating a small (approx. 3 mm × 3 mm) ‘chip’ of clear cellophane (cat. No. SE542; Hoefer Scientific, San Francisco, CA, USA) over each gametophyte. The chips were sterilized in 95 % alcohol for 30 min, followed by two 10 min washes in sterile water. Small, pointed strips of Whatman number 1 paper were used to wick away some of the water between the chip and the gametophyte so that the chip did not drift away but at the same time did not press too closely to the organism. Considerable practice was needed to consistently achieve the desired apposition between the cellophane and the organism. These miniature coverslips were sufficiently porous to allow gas and nutrient exchange; unimpaired growth of gametopytes under these conditions was observed for periods exceeding 60 d. In certain experiments, localized changes in media ion composition were accomplished by adding droplets of media or smaller, ion pre‐loaded chips to the surface of the cellophane overlay. Earlier studies of diffusion through the cellophane were made by preparing two 5 cm × 5 cm layers of hydrated cellophane in the following manner: the lower layer was saturated with a 0·3 g ml–1 solution of bromophenol blue at pH 5, and placed on a glass plate. Bromophenol blue is yellow at this pH. A second layer of cellophane saturated with pH 5 buffer was laid on top of the dye‐containing layer and small KOH pellets were placed on the surface of the second layer. This entire preparation was then covered with a glass Petri dish lid and the edges sealed with water. Bromophenol blue (molecular weight = 670) does not penetrate the cellophane, but KOH penetrated the upper layer instantly, raising the pH and changing the dye colour to blue. The migration of the alkalinizing band could thus be measured over time; the rate of diffusion of ions (D) through cellophane appeared little different to that predicted for diffusion in unstirred water (e.g. for K+, D = 1·9 × 10–4 cm–2 s–1) A knowledge of the approximate rate of lateral diffusion in the cellophane was important in the design of experiments in which a concentrated solution of KCl was added to rhizoid and prothallus ends of the gametophyte grown in cation‐depleted conditions (see Results and Fig. 7). In these experiments, gametophytes that were blocked during prothallial expansion were lifted from the agar surface and placed on a (8 mm × 8 mm) section of plastic, ‘food wrap’ film, resting on the agar. The plastic is hydrophobic and thus impermeable to ions and water. The rhizoid and prothallial ends of the organism were then separately covered with cellophane chips that had been imbibed with sterile water, such that a 50 µm gap remained between them. Given the hydrophobic nature of the underlying plastic film, this increment was sufficient to keep water from establishing continuity, and thus a path for ion diffusion between the chips. The external surfaces of the gametophyte could, of course, serve as a path for diffusion, but this would be expected to be lower than the bulk flows in surrounding water, and, in any case, would not be artificially introduced, since such diffusion would be attributable to inherent features of the organism. Based on studies of KOH migration, it was possible to generate a basal (rhizoid‐end) or acropetal (prothallus‐end) gradient of K+, that was sustained for several days, by placing an additional 1 mm × 1 mm chip, preloaded with 10 mm KCl, on the surface of the cellophane, approx. 2 mm back from the distal ends of the rhizoid or prothallial cells. Observations were made with a Zeiss inverted microscope and photographs were taken with a conventional 35 mm camera, using T max (300 ASA) film. During observations and photography, gametophyte cells were typically exposed to the more intense light from the microscope condenser for less than 5 min; no significant differences in the light‐mediated development of gametophytes were observed that might be attributed to the frequency of examination under the microscope. Photographic negatives were scanned and converted into digital format. Software was used to adjust image brightness, contrast and sharpness to levels which revealed the greatest amount of cellular detail. Microsurgery Portions of filamentous gametophytes were excised with a fresh number 11 scalpel, working under a dissecting microscope at 25× in a laminar flow hood. To prevent the blade from pushing the gametophyte into the agar, the organism was first transferred with the scalpel tip to a 1 cm × 1 cm sheet of sterile, washed cellophane that was placed on a region of the agar surface that was free of gametophytes. The cuts were made with a ‘meat cleaver’ motion and the excised portion was transferred with the scalpel tip back onto the nearby agar. The desired explant was then picked up on a cellophane chip gripped by forceps, transferred to a fresh plate and covered with the chip, as described above. Survival rate for these procedures was quite high; more than 95 % of the surgically removed portions of gametophytes resumed cell division and produced prothalli. RESULTS Light fluence and the conversion of filamentous to planar form Gametophytes grown from germinated spores under a variety of light conditions displayed a consistent relationship between the form attained after 20 d and the amount of white light received in a 24 h period. Organisms grown for 20 d under different total fluences, achieved by adjusting the intensity of light with neutral density filters (Fig. 1A–D), displayed the characteristic range of forms, from filamentous (in complete darkness) to fully expanded prothalli (in unfiltered light), with the filamentous architecture giving way to two‐dimensional forms in organisms grown under less than two layers of neutral density filters (Fig. 1C, D). As described in the Materials and Methods, the light intensity reaching ferns under two layers of filters was 1·9 µm m–2 s–1, giving a 24 h fluence of approx. 165 000 µm m–2. Filaments raised under these light conditions had shortened cells and higher densities of chloroplasts, but rhizoid growth, like that seen in filaments produced under lower light intensities, was minimal. Given the resolution obtainable with layers of identical neutral density filters, there was a clear demarcation between the responses of gametophytes grown under one (Fig. 1C) or two (Fig. 1B) filters. When grown under one filter (6·75 µm m–2 s–1) only prothalli were formed and, except for some extension of the lower (spore end) axis, these were very close in appearance to prothalli produced in unfiltered light. Organisms grown for 20 d in different‐length photoperiods of unfiltered white light were morphologically similar to those obtained under various intensities of continuous white light (Fig. 1E–H). The 24D (continuous darkness) and 24L (continuous light) regimes are, of course, identical to the dark and unfiltered light treatments in the previous experiments. In the photoperiod studies, however, the threshold for maintaining gametophytes in a filamentous state was 2L22D (2 h light and 22 h darkness; Fig. 1F), and these organisms appeared outwardly similar to those grown under two neutral density filters (Fig. 1B). The calculated total fluence for the 2L22D photoperiod of 180 000 µm m–1 compares favourably with that determined above for the two layers of neutral density filters, indicating that these light‐mediated developmental responses are consistent with the predictions for light dose/duration reciprocity. Only planar growth was observed in organisms grown in 4L20D (Fig. 1G), and similar to prothalli produced under one neutral density filter (Fig. 1C), the cordate lamina was somewhat more elongated than that seen in 24L organisms. This extension of the longitudinal axis was apparently due to the presence of several conspicuously elongated cells between the middle and the base of the planar surface. Correlation of prothallial expansion and rhizoid proliferation A consistent feature of every experiment that was undertaken in this study was that a sufficient quantity of white light (>2 µm m–2 s–1) was an absolute requirement for both prothallial and rhizoid development. Figure 2 compares the change in the calculated volume of the prothallus and the aggregate volume of rhizoids in 10 d, dark‐grown organisms as they responded to continuous white light. The rate of prothallial expansion exceeded the rate of rhizoid production and elongation such that there was a consistent six‐ to eight‐fold difference in tissue volumes over a 10 d period. As indicated in the Materials and Methods, the presence of 0·1–1 % glucose or sucrose in the media had no significant effect on any aspect of filamentous or two‐dimensional growth of gametophytes grown in darkness, in filtered light or in unfiltered light (data not shown). Over extended periods of darkness (up to 40 d) gametophytes grew as increasingly longer filaments, but in these organisms there was no extension of the germination rhizoid and no further rhizoids were produced (e.g. see Fig. 5A). At intensities just below the threshold of total fluence which triggered two‐dimensional growth of the apical cell, chloroplast density increased markedly within the filament cells (fx2, Fig. 1B; 2L22D, Fig. 1F), but still no rhizoid growth occurred. On the other hand, total fluences that resulted in the early cell divisions of two‐dimensional growth also promoted rhizoid proliferation (fx1, Fig. 1C; 4L20D, Fig. 1G). These results were always seen, regardless of the age or length of the filament, or the absence of spore end or other filament cells, in surgically severed organisms. Placement of the planar and rhizoid poles exhibits a threshold dependence on axis length In these experiments the range of gametophyte morphologies produced under different light conditions was used to examine how the apparent temporal coordination of two‐dimensional and rhizoid growth might be influenced by the distance between the initial rhizoid pole near the spore, and the apical cell. This distance was generally less than 25 µm (one cell) in full‐light grown organisms and could be extended to nearly 1000 µm in dark‐grown gametophytes (e.g. Fig. 5A). Figure 3 shows the positions and relative growth of 2° rhizoids during the development of selected 2L22D photoperiod‐grown gametophytes with initial filament lengths of 100 µm (A), 200 µm (B) and 250 µm (C). In cases where the distance between the filament apical cell and the spore end of the filament was less than 100 µm, it was consistently observed that all subapical cells of the filament served as sites for 2° rhizoid production, and thus became incorporated into the maturing prothallus over 16 d (Fig. 3A). In organisms that were selected with lengths of approx. 200 µm, the zone of 2° rhizoid growth was extended, over the 16 d observation period, from the spore end towards the cells of the emerging prothallus (Fig. 3B). The series of time‐lapse photographs in Fig. 4 shows that this relocation of 2° rhizoid production occurred as a progressive migration toward the expanding mass of cells at the former filament apex. This phenomenon appeared to depend solely on distance, irrespective of the number of cells that comprised this zone of the filament. For example, the organism in Fig. 4 shows this temporal migration of 2° rhizoids over two cells; the organism in Fig. 7A–C shows this process occurring over the same distance, but involving three to four cells. In situations where the apical cell and spore wall were separated by distances greater than 200 µm (e.g. 250 µm), 2° rhizoids emerged only from cells near the spore wall and those intimately associated with the expanding prothallus (Fig. 3C). Over the 16 d observation period, this relocation of the site of rhizoid emergence established the base for prothallial emergence at the distal end of the filament, and still‐attached cells of the subapical filament and the spore end were not incorporated into the prothallial architecture. In many cases where very long filaments were grown in periods of extended darkness, transfer to full light resulted in the production of two prothalli, one derived from the filament apical cell and the other emanating from the spore end, both attached to the remnant strand of the filament (Fig. 5A–C). Regeneration of gametophytes from excised filament cells Further exploration of the positioning of the rhizoid pole in relation to the emerging photosynthetic surface was undertaken by surgically separating regions of dark‐grown, or 2L22D photoperiod‐grown organisms. When the apical end of the filament was excised from the basal end (including the remnant of the spore), with the explants being exposed to full light, new rounds of cell divisions, starting in the apical cell, gave rise to a prothallus within 30 d (Fig. 5D–F). Prothalli were also formed from the spore end explants in light, with the spore region, as opposed to any of the remaining filament cells, serving as the site for renewed cell divisions (data not shown). Similar filamentous explants, maintained for over 24 h in darkness or under light conditions which did not support photosynthesis, were unresponsive when exposed to full light and were assumed to have died from the interruption of metabolite flow. Regardless of the length of the filament, the number of filament cells (from a single intact apical cell to more than four cells), or the location of excision, only the cells at the tip or base extremities were involved in prothallial regeneration. As was the case with rhizoid emergence in filaments over 200 µm long (described above), new rhizoids were positioned near the base of the expanding cell cluster and their numbers and rate of elongation were positively correlated with the rate of two dimensional expansion. The formation of prothalli at the apex and base of severed filaments was reminiscent of the bipolar formation of prothalli often observed in intact, long filament gametophytes (see Fig. 5A–C), suggesting at first that the intervening, generally elongated cells of the filament did not have regeneration capabilities. It was therefore somewhat surprising to find that excised segments of the mid‐filament were also capable of renewed cell division and ultimately prothallus formation (Fig. 5G–I). In all cases where such regeneration was observed, new rhizoids appeared from one of the filament cells within 2 d, followed by bulging of one of the side walls in the adjacent cell. This bulging was followed by cell divisions which further distorted the original symmetry of the filament, giving rise to a trajectory of pre‐prothallial growth that was oriented at a right angle to the original filament axis. Minimum ion requirements for gametophyte development Regeneration of prothalli from filament cells that were separated from the spore end indicated that the loss of access to organic and mineral nutrients within the spore was readily surmounted if even one intact chlorocyte remained in contact with the basic salt medium and continued to receive sufficient light. To determine how ions in the media might influence regeneration from filament explants, spore end‐excised fragments from 10‐d‐old, dark‐grown or 2L22D‐grown organisms were plated on sucrose‐free, agarose media lacking either K+ or Ca2+ and exposed to full light. Surprisingly, the elimination of added K+ or Ca2+ in the medium had no significant effect on the timing or extent of prothallial regeneration (data not shown; results comparable with those shown in Fig. 5D–F). Since the removal of either K+ or Ca2+ had no effect on gametophyte development, further experiments were undertaken to determine how these organisms were capable of normal development under significantly lower levels of two major nutrient cations. First, by removing both K+ and Ca2+ from the recipe, leaving only NaCl (and the MES buffer) in the agarose‐solidified medium, it was observed that prothalli were still able to form, but the progression was slower and resulted in smaller prothalli (Na+/ose, Fig. 6E–H). Production of prothalli was, however, strongly curtailed by the removal of K+ and Na+ (Ca2+ could be present or absent) from agarose‐based media (Z/ose, Fig. 6I–L). Similar symptoms were obtained in organisms germinated and grown on –K+, –Na+ media and those transferred to this ion‐deficient media 2 d after germination on control media. This inhibition of development could be reversed by adding K+ (or Na+, but not Ca2+) to the organism’s environment; this was done either by transfer to a +K+ plate or by adding a few drops of 0·1 mm KCl to the cellophane chip overlying the organism. Robust rhizoid proliferation and laminar expansion were evident within several days (+K+, d35, Fig. 6L). In the course of these ion‐removal studies, it was also learned that monovalent cation deficiency symptoms could only be elicited on media solidified with low salt agarose; Fig. 6A–D shows the normal development of the prothallus when Bacto‐agar was used as a solidifying agent in a medium that had no added cations. Comparison of the resident cation content in agarose and Bacto‐agar powders indicated that the concentration of each of the major cations is at least five times higher in standard Bacto‐agar. It is not known what proportion of these contaminating ions are unbound and become part of the ionic composition of the media but, judging from these results, it is clear that gametophytes are capable of extracting sufficient ions for growth from unsupplemented Bacto‐agar. Establishing rhizoids as the primary site for ion uptake The monovalent cation deficiency symptoms produced by growing gametophytes on agarose without added K+ or Na+ were used as a starting point for experiments to learn more about the site(s) of osmotic ion uptake in the developing fern gametophyte. As shown in Fig. 7E–H and I–L, deficiency symptoms, as manifested by slower growth, were evident within 10 d. After 23 d, there was no further progression of prothallial development, and indications of general ill health were evidenced by clumping of chloroplasts and a reduction in the number and length of normal‐looking rhizoids. When K+ was added (indicated by +) to alternate sides of a pair of cellophane chips overlying the base (spore end) and the developmentally arrested prothallus, re‐initiation of prothallial expansion was evident only when K+ was able to reach the rhizoids (Fig. 7M–O). Potassium reaching the prothallial portion, but not the rhizoids, did not alleviate the ion deficiency‐induced block on development (Fig. 7P–R). DISCUSSION The mature, two‐dimensional fern gametophyte has the essential structural attributes of many, more familiar, terrestrial autotrophs. The green prothallial surface serves as a photosynthetic solar panel, analogous to a leaf. Rhizoids, with their high surface area to volume ratios, seem the logical domain for ion and perhaps water uptake, and play a similar role to root hairs. The results of this study confirm the predicted functional status of these regions and, furthermore, indicate that there is a strict interdependence between the photosynthate‐producing cells and the zone of osmotic ion uptake. In every experimental circumstance, rhizoid growth required that the laminar surface receive quantities of light sufficient for net carbon fixation, and the enlargement of the photosynthetic portion of the prothallus required a supply of monovalent cations sufficient for the osmotically driven expansion of dividing green cells. The level of interdigitation of these functions was such that following removal of media K+ the organism responded by utilizing sodium ions, taken up by rhizoids, in an effort to meet the osmotic demands of cell expansion in the prothallus. The main contribution to our understanding of plant development to emerge from this study is that the structural alterations which occur during the specification of the dedicated light‐ and ion‐absorbing regions in these fern gametophytes are primarily influenced by cell–cell interactions which, in turn, appear to be inseparable from the organism’s requirements for metabolites and osmotic ions. Specifically, the growth of rhizoids and green cells was never observed to be de‐coupled, and, by repositioning the site of 2° rhizoid emergence, the interstitial zone between these regions was maintained within a span of less than 100 µm in light. Taken together, these observations suggest that ultimate dimensions of the longitudinal axis were subject to control of the developing organism according to its physiological requirements for bidirectional transport of photosynthates and nutrient ions. The driving forces for these transport events were not explored in this study, but given that the distances were in the range of one to four cells and that the translocated species were ions/small molecules, it is not unreasonable to suggest that the nutrient streams are diffusion‐based, perhaps augmented by cytoplasmic streaming. Accordingly, my analysis of the range of morphologies that may be elicited in Onoclea gametophytes proceeds from the standpoint that the controlling parameters are the rates of delivery from photosynthetic and ion‐uptake source regions, and the magnitude of the demands these same regions exert on photosynthate and ion supplies in their corresponding roles as sinks. Each of these is discussed, in turn, below. Analysis of the photosynthetic source–sink relationship The cells of the dark‐grown filament, which develops into the prothallial portion of the gametophyte, quickly increased their numbers of chloroplasts in response to even short exposures of light. These cells of the expanding prothallus were thus assumed to function as the photosynthetic source, and the non‐green rhizoids, which were able to grow only when light struck the prothallus, clearly behaved as a sink for some of the photosynthate produced. As shown graphically in Fig. 2 and photographically in most of the other figures, growth of the photosynthetic and non‐photosynthetic portions of the gametophyte, from the earliest cell expansion to the complete prothallus, appeared to be highly coordinated. Even in surgically excised filament cells, subsequent rhizoid emergence and growth was always correlated with the initial swelling and intense packing of chloroplasts in a green cell. The development of gametophytes in light displayed an all‐or‐nothing type of response with a total fluence threshold of approx. 175 000 µm m–2 per 24 h period, which, administered as continuous irradiation, corresponded to a fluence rate of approx. 2 µm m–2 s–1. This level is below the light compensation point reported for many C3 and C4 terrestrial plants, but ferns are remarkably well adapted to low light conditions and this value compares rather well with other determinations of the photosynthetic compensation point in fern gametophytes (Friend, 1975). As reported in earlier studies, it was noted that expansion of the Onoclea filament tip cell (as well as other filament cells) and dramatic increases in chloroplast density were triggered by total fluences well below the apparent photosynthetic threshold for this organism (e.g. 2L22D photoperiod, Fig. 1F). By convention, therefore, these initial events may be categorized as photomorphogenetic phenomena, and there is a considerable body of evidence which shows that phytochrome and a blue light photoreceptor residing in the plasma membrane and/or membranes associated with the nucleus (reviewed in Furuya, 1983) are responsible for these changes in morphology. In Onoclea, however, the photomorphogenetic effects were clearly separable from the more profound changes in form that required the involvement of photosynthetic, chloroplast‐based machinery. It was observed in this study that filaments with somewhat expanded cells and dense populations of chloroplasts could be held in this state by maintaining the organisms in total fluences below the apparent photosynthetic compensation point. All further processes of cell expansion, cell division and cell differentiation (e.g. 2° rhizoid production) leading to prothallial form, required quantities of light sufficient for net photosynthesis. With regard to earlier studies that reported the presence of media sucrose was necessary to sustain, or enhance, the growth of fern gametophytes in very low light conditions (Miller and Miller, 1961; Kato, 1967), development of Onoclea gametophytes in the present experiments was tied strictly to light conditions that supported photosynthesis, whether or not sucrose was present in the medium. It was also never observed that a supplementary carbon source could induce these ferns to develop as heterotrophs. Growth in darkness was apparently sustained by mobilization of reserves present in the spore; surgical removal of the spore end resulted in the rapid death of organisms in darkness, even with sucrose‐supplemented media, but spore‐less explants developed into prothalli on sucrose‐free media in the light. Furthermore, in considering these earlier studies, is not clear why it was apparently believed that an exogenously applied carbon source might be a viable substitute for the internally produced photosynthates. The green portion of the gametophyte has a waxy cuticle (Wada and Staehelin, 1981), which would be expected to impede uptake, and the rhizoids, by all accounts, show little growth in darkness, irrespective of sugar availability in the media. Reports of species‐specific, or experimentally induced, rhizoid or prothallial growth in low intensity light (Howland and Boyd, 1974) or darkness (Miller and Miller, 1970) must mean that there are circumstances under which the transition to two‐dimensional growth can preferentially utilize prepackaged photosynthates in the spore. Establishing causal links between an organism’s structure and its functions is always desirable; however, pursuing the apparent relationships between the changes in gametophyte structure and the dynamics of photosynthate allocation is encumbered by some deficits in our knowledge which can not be circumvented. First, we do not know the identity of the primary transportable species of photosynthates in ferns. The relevant literature on phloem loading and unloading in higher plants is reasonably unified in the belief that sucrose is the membrane‐transported and vascularly translocated molecule (Ziegler, 1975; Giaquinta, 1983). Secondly, the rate of photosynthesis and photosynthate allocation has never been measured in fern gametophytes. In other systems estimates of net carbon incorporation have been derived from a determination of changes in dry weight, but given the gametophyte’s diminutive size, this would have to be based on a sizeable, developmentally mixed population of organisms. Furthermore, without some independent measure of the rate of carbon uptake, or the rate of consumption from cell divisions and molecular syntheses, these calculations would not bear on the key issue of how domains of the gametophyte detect and then compensate for inequities in supply and demand of fixed carbon. Finally, although the path between source and sink regions is greatly simplified by the absence of vascular tissue, the extent to which symplastic and apoplastic modes of translocation are involved in distribution of photosynthate is not known. Analysis of the osmotic ion source–sink relationship Rhizoids of fern gametophytes are positioned at the base of the expanding prothallus and have the appropriate morphological features of nutrient‐absorbing structures; it is therefore surprising that there is little direct evidence related to their function. Comparing protonema and rhizoid cells on the bases of cell wall composition and the ability to take up vital stains, Smith (1972a) found that rhizoids were much more permeable to water‐soluble dyes, and that the cell behaved as a cation‐exchange medium. Smith (1972b) also showed that marker enzymes associated with phosphate metabolism were present in gametophyte rhizoids, but approaches which established the absorptive capabilities of root hairs, such as ion deficiency experiments or studies with radio‐tracers, have not been applied to fern rhizoids. Because the fern gametophyte grows in direct contact with the medium (in these experiments they were essentially grown in submerged conditions), one cannot safely assume, on the basis of location and appearance, that rhizoids are the sole site for ion uptake in this organism. In fact, the ability of green, thallus‐like surfaces to take up ions directly from media is apparently widespread in the plant kingdom. Examples include ion uptake through the laminar surfaces of the green alga, Ulva (West and Pitman, 1967) and foliar ion uptake exhibited by various terrestrial higher plants (Franke, 1967; Kannan, 1980). Results from the present study suggest that rhizoids appear to be the primary, if not the sole, site of uptake for K+ in Onoclea gametophytes. This is most clearly demonstrated in the monovalent cation ion deficiency experiments in which inhibition of prothallial expansion was alleviated by supplying K+ to the rhizoid end of the organism (Fig. 7M–O). As for most plants, ion uptake systems in fern gametophytes are apparently centred around the accumulation of K+ which, at typical cellular concentrations of 100 mm, is the chief osmotic agent that governs cell turgor and, in turn, turgor‐driven cellular expansion. The internal concentration of K+ in gametophyte cells has not been determined, but measurements of membrane potentials (Racusen and Cooke, 1982), which are highly dependent on outward K+ diffusion, produced values for filament cells that are similar to those seen in a variety of plant cells with known internal concentrations of approx. 100 mm K+ (Findlay and Hope, 1976). The importance of media‐supplied K+ as the osmotic agent that regulates expansion of green cells in the gametophyte is highlighted in experiments where only NaCl is provided in the agarose medium. Although expansion of the prothallus was noticeably restrained, Onoclea gametophytes were still capable of producing the characteristic, heart‐shaped form, apparently because they were able, under K+ deficiency conditions, to substitute Na+ as the osmotic agent. Similar trade‐offs have been documented in other systems, such as the growth of cultured leaf discs (Marschner and Possingham, 1975), and the operation of stomata (Raghavendraet al., 1976). In these cases it appears that the requirement for a monovalent cation species, as part of the regulatory mechanism for cell turgor, has priority over the role that K+ has in stabilizing enzymes (Clarkson and Hanson, 1980). Eliminating both K+ and Na+ from the nutrient solution was the only medium condition which completely (but reversibly) inhibited development of the prothallus. Concentrations of Ca2+ and other nutrient ions were apparently high enough to support growth of a complete prothallus. The effects of monovalent cation deficiency were not, however, immediately obvious; development of the prothallus kept pace with that of control organisms for approx. 10 d. Thereafter, the resulting, roughly triangular structure ceased growth and, over extended time, began to exhibit abnormalities of organelle clumping and physical discontinuity among cells (Fig. 7H, L). The deleterious effects of monovalent cation starvation were ultimately reversed by application of either K+ or Na+; it is possible that the early, unimpeded phase of growth was supported by mobilizing K+ and/or Na+ within the spore. The following calculation suggests that this could indeed be the case. Assuming a combined K+ and Na+ concentration within the spore of 250–300 mm (derived from values in Wayne and Hepler, 1985), and a K+/Na+ concentration inside expanded prothallial cells of 100 mm, it follows that the ion contents of the spore could support a prothallial structure 2·5 to 3 times the volume of the spore. Comparison of the spore and prothallus volumes of gametophytes grown under K+/Na+ deficiency, such as those shown in Figs 6K and 7G, are consistent with this prediction. As is the case with photosynthetic production and dispersal, further analysis of ion source/sink interactions is impeded by a lack of knowledge concerning the rates of ion uptake from solution, and the path of intercellular transport within the developing gametophyte. However, since the ion sink regions accumulate, but cannot metabolically transform the transported entity, it is possible to use changes in volumes of the prothallus and total rhizoids to estimate the demand for K+. For example, assuming that the internal cellular concentration of K+ is approx. 100 mm, then computing the change in total volume of cells (prothallus + rhizoids) between day 4 and day 10 in Fig. 2, and finally coupling these factors with an estimate of the mean surface area of rhizoids (from aggregate length), a net influx for K+ of approx. 5 pm cm–2 s–1 is obtained. This value compares favourably with previous determinations of 1–2·5 pm cm–2 s–1 for K+ influx in freshwater algae (MacRobbie, 1974) and higher plants (Higinbothamet al., 1967). Since the expanding prothallus comprises a volume six to eight times that of the rhizoids, it follows that approx. 85 % of the rhizoid K+ influx is transported to prothallial cells. Determination of polarity and pattern formation coincide with metabolite transport in the gametophyte The light‐induced changes in the position of new cell walls in green apical cells of the fern gametophyte have always held a special prominence for those who utilize this organism as a model system to study plant development, and considerable effort has been directed towards an understanding of the apparently unique physiological (Daviset al., 1974; Racusen and Cooke, 1982), cytoskeletal (Stetler and DeMaggio, 1972; Murata and Wada, 1989), chemical (Raghavan, 1968; Smithet al., 1973) and geometric (Cooke and Paolillo, 1980b) properties that allow it to function as a sort of developmental ‘ground‐cell’ for the organism. In contrast, few previous studies have focused on the ways in which the changes in structure might be related to the overall function of the organism. A prerequisite for development in multicellular organisms is a mechanism for determining the polarity of the main body axis. During germination in ferns, axis polarity is apparently established before the emergence of the rhizoid and protonematal tip cells, indicating that chemical, electrical or structural features of the spore govern the initial placement of the ion‐ and light‐absorbing regions. The regeneration experiments reported here indicate further that remnants of severed filaments are capable, within the span of a few cell divisions, of re‐establishing the poles for a new axis. The determinants of axis polarity in regenerating gametophytes are unknown, but it was consistently observed that the re‐modelled axis comprised two to four green cells that tended not to exceed 200 µm, between the site of 2° rhizoid emergence and the tip cell—a distance comparable with that in other systems in which a diffusible morphogen is involved in the determination of polarity (Wolpert, 1996). As the re‐establishment of polarity in severed gametophyte filaments was dependent on quantities of light sufficient to support photosynthesis, it is conceivable that activation of the cellular processes that increase the supply of available metabolites also establishes, in tandem, the developmental polarity for the gametophyte axis. With increasing transport activity, this initial determination of polarity is reinforced by the biochemical and structural changes attending the production of more nutrient‐absorbing and photosynthetic cells. Non‐vascularized, terrestrial plants with thallus‐type architecture possess inherent structural limitations on their ability to efficiently move photosynthates and ions over long distances. The available escort paths involve either multiple cell‐to‐cell transfers, in this case necessitating coordination of many spatially localized membrane transporters to effect the appropriate direction of transport through the tissue, or movement through extracellular spaces, which, again, would require some mechanism to provide a driving force and to establish direction. In a mature prothallus, the path from cells in the meristematic notch to the rhizoids at the base may consist of as many as eight comparably sized chlorocytes (for example, see Fig. 6D), and, given that the developing prothallus consistently maintains an even shorter distance (two to four cells) between the rhizoids and the expanding two‐dimensional chlorocyte array, this may, in fact, represent the physical limit over which transport is possible in this system. In this context it is noteworthy that in the roots (Drew, 1987), sink tissues (Eschrich, 1989; Woodet al., 1997) and leaves (Philpott, 1953; Giaquinta, 1983) of the much better‐studied vascular plants, it is most common to find cell‐to‐cell transport across fields consisting of fewer than ten parenchymous cells, and it remains an issue of some debate as to the relative involvement of symplastic (plasmodesmatal) and apoplastic (membrane transport) mechanisms (Van Bel, 1993). With this demonstration that the bi‐directional flow of osmotic ions and photosynthates has cause‐and‐effect roles in the development of the prothallus, it seems reasonable to suggest that the orientation of cell divisions, believed to be the main patterning mechanism in the gametophyte, might profitably be examined in light of their potential contribution to the transport of ions or photosynthates. ACKNOWLEDGEMENTS The author gratefully acknowledges Susan Klinedinst for assistance in developing some of the techniques used in this study, Melissa Dooley for help in searching the literature and Dr F. Mark Schiavone for comments during revision. View largeDownload slide Fig. 1. A–D, Development of gametophytes grown under different intensities of continuous white light. Two‐day‐old germinated spores were transferred to Petri dishes with four (f × 4), two (f × 2), one (f × 1) or no (full light; FL) neutral density filters covering the plate lid (see Materials and Methods). E–H, Development of gametophytes grown under different photoperiods of white light. Two‐day‐old germinated spores were exposed to photoperiods of continuous darkness (24D), 2 h light (2L22D), 4 h light (4L20D) and continuous light (24L). All photographs were taken 20 d after the start of the experiment. Bar = 100 µm. View largeDownload slide Fig. 1. A–D, Development of gametophytes grown under different intensities of continuous white light. Two‐day‐old germinated spores were transferred to Petri dishes with four (f × 4), two (f × 2), one (f × 1) or no (full light; FL) neutral density filters covering the plate lid (see Materials and Methods). E–H, Development of gametophytes grown under different photoperiods of white light. Two‐day‐old germinated spores were exposed to photoperiods of continuous darkness (24D), 2 h light (2L22D), 4 h light (4L20D) and continuous light (24L). All photographs were taken 20 d after the start of the experiment. Bar = 100 µm. View largeDownload slide Fig. 2. Changes in aggregate volume of the prothallus (triangles) and rhizoids (circles) during gametophyte development over 10 d. The experiment was started with organisms germinated 2 d earlier in continuous light. Values were calculated by applying simple geometric formulae to measurements of length, width and thickness of representative gametophytes grown in continuous white light. Bars are means ± s.d. for 20 organisms. View largeDownload slide Fig. 2. Changes in aggregate volume of the prothallus (triangles) and rhizoids (circles) during gametophyte development over 10 d. The experiment was started with organisms germinated 2 d earlier in continuous light. Values were calculated by applying simple geometric formulae to measurements of length, width and thickness of representative gametophytes grown in continuous white light. Bars are means ± s.d. for 20 organisms. View largeDownload slide Fig. 3. Position and growth of secondary rhizoids during development of selected gametophytes with 100 µm (A), 200 µm (B) and 250 µm (C) filament lengths. Gametophytes with specific filament lengths were transferred from various‐aged cultures of organisms grown under a 2L22D photoperiod. The x‐axis of each graph represents the length of the filament axes, with the base of the filament at the origin and the apical cell at the highest x‐axis value. For each 25 µm interval along the filament, the lengths of all rhizoids were summed, and this aggregate rhizoid length plotted on the y‐axis. Open bars show rhizoid position and growth after 4 d in full light; hatched bars show rhizoid growth after 20 d in full light. Results are means ± s.d. of eight organisms for each plot. View largeDownload slide Fig. 3. Position and growth of secondary rhizoids during development of selected gametophytes with 100 µm (A), 200 µm (B) and 250 µm (C) filament lengths. Gametophytes with specific filament lengths were transferred from various‐aged cultures of organisms grown under a 2L22D photoperiod. The x‐axis of each graph represents the length of the filament axes, with the base of the filament at the origin and the apical cell at the highest x‐axis value. For each 25 µm interval along the filament, the lengths of all rhizoids were summed, and this aggregate rhizoid length plotted on the y‐axis. Open bars show rhizoid position and growth after 4 d in full light; hatched bars show rhizoid growth after 20 d in full light. Results are means ± s.d. of eight organisms for each plot. View largeDownload slide Fig. 4. Acropetal migration of secondary rhizoids during development of a filamentous gametophyte with an axial length between the apical cell and the spore of approx. 200 µm. The organism at day 0 (d0) had been grown in the dark for 10 d and then transferred to continuous light. Note progression of secondary rhizoid emergence at locations progressively closer to the expanding prothallus. Bar = 100 µm. View largeDownload slide Fig. 4. Acropetal migration of secondary rhizoids during development of a filamentous gametophyte with an axial length between the apical cell and the spore of approx. 200 µm. The organism at day 0 (d0) had been grown in the dark for 10 d and then transferred to continuous light. Note progression of secondary rhizoid emergence at locations progressively closer to the expanding prothallus. Bar = 100 µm. View largeDownload slide Fig. 5. A–C, Prothallial development from a very long (>200 µm) filamentous gametophyte. The organism at d0 had been grown in the dark for 40 d and then transferred to continuous light. D–F, Prothallial regeneration from an excised filament containing the apical cell. Explant at d0 was severed from a 30‐d‐old, dark‐grown gametophyte; the apical cell is to the left and the cut end to the right. From d0 onward, the explant was in continuous light. G–I, Prothallial regeneration from an excised segment of subapical cells; both the apical cell and the spore end have been removed. Explant at d0 was excised from a 30‐d‐old dark‐grown gametophyte; from d0 onward, the explant was in continuous light. Rhizoids were present at d5 onward, but protruded below the structure and are not visible in photographs until d19. Bar = 100 µm. View largeDownload slide Fig. 5. A–C, Prothallial development from a very long (>200 µm) filamentous gametophyte. The organism at d0 had been grown in the dark for 40 d and then transferred to continuous light. D–F, Prothallial regeneration from an excised filament containing the apical cell. Explant at d0 was severed from a 30‐d‐old, dark‐grown gametophyte; the apical cell is to the left and the cut end to the right. From d0 onward, the explant was in continuous light. G–I, Prothallial regeneration from an excised segment of subapical cells; both the apical cell and the spore end have been removed. Explant at d0 was excised from a 30‐d‐old dark‐grown gametophyte; from d0 onward, the explant was in continuous light. Rhizoids were present at d5 onward, but protruded below the structure and are not visible in photographs until d19. Bar = 100 µm. View largeDownload slide Fig. 6. Gametophyte development under conditions where specific media ions were removed. Organisms at d0 had been grown in a 2L22D photoperiod for 4 d, then transferred to continuous light. A–D (Z/agar; ‘zero agar’) shows that normal development occurs when organisms were placed on Bacto‐agar with no added salts. E–H (Na/ose; ‘sodium agarose’) shows that development of the prothallus is slower and smaller when organisms were placed on a low‐salt agarose medium containing only 0·1 mm NaCl. I–L (Z/ose; ‘zero agarose’) shows inhibition of prothallus development at about d7 (J) in organisms placed on low‐salt agarose medium with no added salts. Lower‐diagonal portion of L shows the resumption of prothallus development after a further 8 d (d35), when 0·1 mm K+ was added to the d27, Z/ose organism directly above it. Bar = 100 µm. View largeDownload slide Fig. 6. Gametophyte development under conditions where specific media ions were removed. Organisms at d0 had been grown in a 2L22D photoperiod for 4 d, then transferred to continuous light. A–D (Z/agar; ‘zero agar’) shows that normal development occurs when organisms were placed on Bacto‐agar with no added salts. E–H (Na/ose; ‘sodium agarose’) shows that development of the prothallus is slower and smaller when organisms were placed on a low‐salt agarose medium containing only 0·1 mm NaCl. I–L (Z/ose; ‘zero agarose’) shows inhibition of prothallus development at about d7 (J) in organisms placed on low‐salt agarose medium with no added salts. Lower‐diagonal portion of L shows the resumption of prothallus development after a further 8 d (d35), when 0·1 mm K+ was added to the d27, Z/ose organism directly above it. Bar = 100 µm. View largeDownload slide Fig. 7. Demonstration that the rhizoid end of the gametophyte, and not the prothallial end, is the primary site of ion uptake. A–L shows development of prothalli on control media (agar with all nutrient ions; A–D), on low‐salt agarose with no added salts (Z/ose, E–L). Organisms were grown in a 2L22D photoperiod for 7 d, and then transferred to continuous light at d0. Ion deficiency symptoms are evident in both the Z/ose treatments by d10 (G and K). M–R shows the response of the same two deficient organisms when K+ is added to the cellophane chip overlying the rhizoid end (Z/ose a, M–O), or to the cellophane chip overlying the prothallus end (Z/ose b, P–R). Bar = 100 µm. View largeDownload slide Fig. 7. Demonstration that the rhizoid end of the gametophyte, and not the prothallial end, is the primary site of ion uptake. A–L shows development of prothalli on control media (agar with all nutrient ions; A–D), on low‐salt agarose with no added salts (Z/ose, E–L). Organisms were grown in a 2L22D photoperiod for 7 d, and then transferred to continuous light at d0. Ion deficiency symptoms are evident in both the Z/ose treatments by d10 (G and K). M–R shows the response of the same two deficient organisms when K+ is added to the cellophane chip overlying the rhizoid end (Z/ose a, M–O), or to the cellophane chip overlying the prothallus end (Z/ose b, P–R). Bar = 100 µm. References BanksJA. 1999 Gametophyte development in ferns. Annual Review of Plant Physiology and Plant Molecular Biology  50: 163–186. Google Scholar ClarksonDT, Hanson JB. 1980. The mineral nutrition of higher plants. Annual Review of Plant Physiology  31: 239–298. Google Scholar CookeTJ, Paolillo DJ. 1980a. The control of the orientation of cell divisions in fern gametophytes. American Journal of Botany  67: 1320–1333. Google Scholar CookeTJ, Paolillo DJ. 1980b. Dark growth and the associated phenomenon of age‐dependent photoresponses in fern gametophytes. Annals of Botany  45: 693–702. Google Scholar CookeTJ, Racusen RH. 1982. Cell expansion in the filamentous gametophyte of the fern Onoclea sensibilis L. Planta  155: 449–458. Google Scholar DavisBD, Chen JCW, Philpott M. 1974 The transition from filamentous to two‐dimensional growth in fern gametophytes. IV. Initial events. American Journal of Botany  61: 722–729. Google Scholar DrewMC. 1987 Function of root tissues in nutrient and water transport. In: Gregory PJ, Lake JV, Rose DA, eds. Root development and function. Cambridge: Cambridge University Press, 71–101. Google Scholar DyerAF. 1979. The experimental biology of ferns. New York: Academic Press. Google Scholar DyerAF, King MAL. 1979. Cell division in fern protonemata. In: Dyer AF, ed. The experimental biology of ferns. New York: Academic Press, 307–354. Google Scholar EschrichW. 1989. Phloem unloading of photoassimilates. In: Baker DA, Milburn JA, eds. Transport of photoassimilates. Essex: Longman Scientific and Technical, 206–263. Google Scholar FindlayGP, Hope AB. 1976. Electrical properties of plant cells: methods and findings. In: Lüttge U, Pitman MG, eds. Encyclopedia of plant physiology (new series), Transport in plants, II A. Cells. New York: Springer‐Verlag, 53–92. Google Scholar FrankeW. 1967. Mechanism of foliar penetration of solutions. Annual Review of Plant Physiology  18: 281–300. Google Scholar FriendDJC. 1975. Adaptation and adjustment of photosynthetic characteristics of gametophytes and sporophytes of Hawaiian tree‐fern (Cibotium glaucum) grown at different irradiances. Photosynthetica  9: 157–164. Google Scholar FuruyaM. 1983. Photomorphogenesis in ferns. In: Shropshire W, Mohr H, eds. Encyclopedia of plant physiology (new series), Photomorphogenesis, 16B. New York: Springer‐Verlag, 569–600. Google Scholar GiaquintaRT. 1983 Phloem loading of sucrose. Annual Review of Plant Physiology  34: 347–387. Google Scholar HiginbothamN, Etherton B, Foster RJ. 1967. Mineral ion contents and cell transmembrane electropotentials of pea and oat seedling tissue. Plant Physiology  42: 37–46. Google Scholar HowlandGP, Boyd EL. 1974. Genetic control of photomorphogenesis. Isolation of nonfilamentous mutants after gamma irradiation of Pteridium aquilinum spores. Radiation Botany  14: 281–285. Google Scholar KannanS. 1980. Mechanism of foliar uptake of plant nutrients: Accomplishments and prospects. Journal of Plant Nutrition  2: 717–735. Google Scholar KatoY. 1967. Physiological and morphogenetic studies of fern gametophytes and sporophytes in aseptic culture. VIII. Behavior of Pteris vittata gametophytes on sugar and complex media in red light. Planta  77: 127–134. Google Scholar MacRobbieEAC. 1974. Ion uptake. In: Stewart WDP, ed. Algal physiology. Los Angeles: University of California Press, 676–713. Google Scholar MarschnerH, Possingham JV. 1975. Effect of K+ and Na+ on growth of leaf discs of sugar beet and spinach. Zeitschrift für Pflanzenphysiologie 75: 6–16. Google Scholar MillerJH, Miller PM. 1961 The effect of different light conditions and sucrose on the growth and development of the gametophyte of the fern, Onoclea sensibilis. American Journal of Botany  48: 154–159. Google Scholar MillerJH, Miller PM. 1970. Unusual dark‐growth and antheridial differentiation in some gametophytes of the fern Onoclea sensibilis. American Journal of Botany  57: 1245–1248. Google Scholar MurataT, Wada M. 1989. Organization of cortical microtubules and microfibril deposition in response to blue‐light‐induced apical swelling in a tip‐growing Adiantum protonema cell. Planta  178: 334–341. Google Scholar OotakiT, Furuya M. 1969. Experimentally induced apical dominance in protonemata of Pteris vittata. Embryologia  10: 284–296. Google Scholar PhilpottJ. 1953. A blade tissue study of leaves of forty seven species of Ficus. Botanical Gazette  115: 15–35. Google Scholar RacusenRH, Cooke TJ. 1982. Electrical changes in the apical cell of the fern gametophyte during irradiation with photomorphogenetically active light. Plant Physiology  70: 331–334. Google Scholar RacusenRH, Ketchum KA, Cooke TJ. 1988. Modifications of extracellular electric and ionic gradients preceding the transition from tip growth to isodiametric expansion in the apical cell of the fern gametophyte. Plant Physiology  87: 69–77. Google Scholar RaghavanV. 1968. RNA and protein metabolism in the particulate fractions of the gametophytes of bracken fern during growth in red and blue light. Planta  81: 38–48. Google Scholar RaghavanV. 1989. Developmental biology of fern gametophytes. Cambridge: Cambridge University Press. Google Scholar RaghavendraAS, Rao IM, Das VSR. 1976. Replaceability of potassium by sodium for stomatal opening in epidermal strips of Commelina benghalensis. Zeitschrift für Pflanzenphysiologie  80: 36–42. Google Scholar SmithDL. 1972a. Localization of phosphatases in young gametophytes of Polypodium vulgare L. Protoplasma  74: 133–148. Google Scholar SmithDL. 1972b. Staining and osmotic properties of young gametophytes of Polypodium vulgare L. and their bearing on rhizoid function. Protoplasma  74: 465–479. Google Scholar SmithDL, Robinson PB, Govier RN. 1973. Growth factors produced by gametophytes of Polypodium vulgare L. grown under red and blue light. New Phytologist  72: 1261–1268. Google Scholar StetlerDA, DeMaggio AE. 1972. An ultrastructural study of fern gametophytes during one‐ to two‐dimensional development. American Journal of Botany  59: 1011–1017. Google Scholar TilneyLG, Cooke TJ, Connelly PS, Tilney MS. 1990 The distribution of plasmodesmata and its relationship to morphogenesis in fern gametophytes. Development  110: 1209–1221. Google Scholar Van BelAJE. 1993. Strategies of phloem loading. Annual Review of Plant Physiology and Plant Molecular Biology  44: 253–281. Google Scholar WadaM, Staehelin LA. 1981. Freeze‐fracture observations on the plasma membrane, the cell wall and the cuticle of growing protonemata of Adiantum capillus‐veneris L. Planta  151: 462–468. Google Scholar WayneR, Hepler PK. 1985. The atomic composition of Onoclea sensibilis spores. American Fern Journal  75: 12–18. Google Scholar WestKR, Pitman MG. 1967 Ionic relations and ultrastructure in Ulva lactuca. Australian Journal of Biological Sciences  20: 901–914. Google Scholar WolpertL. 1996. One hundred years of positional information. Trends in Genetics  12: 359–364. Google Scholar WoodRM, Offler CA, Patrick JW. 1997. The cellular pathway of short‐distance transfer of photosynthates and potassium in the elongating stem of Phaseolus vulgaris L. A structural assessment. Annals of Botany  79: 89–100. Google Scholar ZieglerH. 1975. Nature of transported substances. In: Zimmermann MH, Milburn JA, eds. Encyclopedia of plant physiology (new series), Transport in plants, I. Phloem transport. New York: Springer‐Verlag, 59–100. Google Scholar

Journal

Annals of BotanyOxford University Press

Published: Feb 1, 2002

There are no references for this article.