Development of the embryo of Encephalartos natalensis from a rudimentary meristematic structure approximately 700 μm in length extends over 6 months after the seed is shed from the strobilus. Throughout its development, the embryo remains attached to a long suspensor. Differentiation of the shoot meristem flanked by two cotyledonary protuberances occurs over the first 2 months, during which peripheral tannin channels become apparent. Tannins, apparently elaborated by the endoplasmic reticulum, first accumulate in the large central vacuole and ultimately fill the channel. By the fourth month of development, the root meristem is apparent and procambial tissue forming discrete vascular bundles can be discerned in the elongating cotyledons. Between 4 and 6 months, mucilage ducts differentiate; after 6 months, when the seed becomes germinable, the embryo is characterised by cotyledons far longer than the axis. Shoot and root meristem cells remain ultrastructurally similar throughout embryo ontogeny, containing small vacuoles, many well-differentiated mitochondria and endoplasmic reticulum (ER) profiles, abundant polysomes, plastids containing small starch deposits and Golgi bodies. Unusually, however, Golgi bodies are infrequent in other cells including those elaborating mucilage which is accumulated in distended ER and apparently secreted into the duct lumen directly by ER-derived vesicles. The non-meristematic cells accumulate massive starch deposits to the exclusion of any protein bodies and only very sparse lipid, features which are considered in terms of the prolonged period of embryo development and the high atmospheric oxygen content of the Carboniferous Period, when cycads are suggested to have originated.

Very little is known about the pre-and postshedding megagametophyte development and utilisation of accumulated reserves, respectively, in cycads (Zamiaceae). In the present study on developing ovules of the recalcitrant-seeded species, Encephalartos natalensis, cells of the megagametophyte were found to become progressively packed with starch and protein as the two main storage reserves, a limited number of discrete lipid bodies, and occasional mitochondria all of which appeared to be embedded in a homogeneous matrix. ER-derived vesicles (and not Golgi-derived vesicles) appeared to be the principal contributors of the primary cell wall components, pectin and xylan, during megagametophyte cellularisation. This was confirmed by the use of enzyme-gold localisation. High-pressure freezing (HPF) and freeze substitution (FS) of samples the following season showed that while the apparently featureless cytomatrix of the megaspore was an artefact of conventional fixation, there was still an insignificant occurrence of Golgi bodies during primary wall formation. When enzyme-gold localisation was employed on the HPF-FS material, label for pectin and xylan was found only in the regions of ER and vesicles and not in any of the few Golgi bodies or their associated vesicles. Immunocytochemistry revealed that pectin and xylan were restricted to the ER and not to any vesicles or to the occasional Golgi body that was found. This suggests that the ER exclusively, is involved in the deposition of these primary cell wall components during the cellularisation of the E. natalensis megagametophyte. While cellularisation took place over approximately 1-2 weeks, subsequent development of the megagametophyte cells involved the accumulation of storage reserves, this phase lasting approximately 8 months-after which the seeds were shed whether pollination/fertilisation had recently occurred, or not.

No abstract

Plant cell walls (CWs) are dynamic in that they can change conformation during ontogeny and in response to various stresses. Though seeds are the main propagatory units of higher plants, little is known of the conformational responses of zygotic embryo CWs to drying. This study employed cryo-scanning electron microscopy to compare the effects of desiccation on zygotic embryo CW morphology across three gymnosperm species that were shown here to differ in seed desiccation sensitivity: Podocarpus henkelii (highly desiccation-sensitive), Podocarpus falcatus (moderately desiccation-sensitive), and Pinus elliottii (desiccation-tolerant). Fresh/imbibed (i.e. fresh Podocarpus at shedding and imbibed Pi. elliottii) embryos showed polyhedral cells with regular walls, typical of turgid cells with an intact plasmalemma. Upon desiccation to c. 0.05 g g (dry mass basis), CWs assumed an undulating conformation, the severity of which appeared to depend on the amount and type of dry matter accumulated. After desiccation, intercellular spaces between cortical cells in all species were comparably enlarged relative to those of fresh/imbibed embryos. After rehydration, meristematic and cotyledonary CWs of P. henkelii and meristematic CWs of P. falcatus remained slightly undulated, suggestive of plasmalemma and/or CW damage, while those of Pi. elliottii returned to their original conformation. Cell areas in dried-rehydrated P. henkelii root meristem and cotyledon were also significantly lower than those from fresh embryos, suggesting incomplete recovery, even though embryo water contents were comparable between the two states. Electrolyte leakage measurements suggest that the two desiccation-sensitive species incurred significant plasmalemma damage relative to the tolerant species upon desiccation, in agreement with the CW abnormalities observed in these species after rehydration. Immunocytochemistry studies revealed that of the four CW epitopes common to embryos of all three species, an increase in arabinan (LM6) upon desiccation and rehydration in desiccation-tolerant Pi. elliottii was the only difference, although this was not statistically significant. Seed desiccation sensitivity in species like P. henkelii and P. falcatus may therefore be partly based on the inability of the plasmalemma and consequently CWs of dried embryos to regain their original conformation following rehydration.

No abstract