Investigations into the responses of axes of recalcitrant seeds to dehydration and cryopreservation.

Abstract

Achieving long-term storage of germplasm is critical for the conservation of plant biodiversity. Seed storage practices require that degradative reactions causing ageing be limited. By reducing the water content, cytoplasmic viscosity is increased to levels that minimise deteriorative reactions. Reducing the storage temperature additionally increases the storage lifespan by further reducing the rate at which such deleterious processes occur. Two broad categories of seeds can be distinguished based on their storage behaviour. Orthodox seeds are desiccation-tolerant; generally shed in the dry state and are metabolically quiescent. Such seeds are usually stored at low water contents (e.g. 5%), and their high cytoplasmic viscosity prevents freezing damage during cooling to subzero temperatures. On the other hand, desiccation-sensitive (recalcitrant) seeds do not undergo a maturation-drying phase, they are metabolically active at shedding, and sensitive to extreme or prolonged drying. Accordingly, recalcitrant seeds cannot be stored under conventional conditions because they do not survive drying to low water contents and are damaged by sub-zero temperatures, even when dried to the lowest water content tolerated. Therefore, procedures that facilitate harmless drying and cooling to low temperatures are required to achieve long-term storage of recalcitrant germplasm. Recalcitrant seeds that are dried rapidly can attain relatively lower water contents without injury. However, these seeds are usually large and this limits the drying rates that can be achieved even under favourable conditions. Isolating embryonic axes from the rest of the seed facilitates faster drying, and a consequent reduction in the water content at which damage occurs. In axes of many species, the level of drying attained before lethal desiccation damage occurs is sufficient to limit freeZing damage during cryogenic exposure and facilitate survival in vitro. However, many others are damaged when dried to water contents that preclude freezing, and also are killed if cooled to sub-zero temperatures at higher water contents. In such instances, the window of permissible water contents leading to survival may be small or nonexistent. A basic premise explored in this thesis is that by restricting the growth of intracellular ice crystals using increasingly rapid cooling rates, the range of permissible water contents can be widened, facilitating survival of axes at higher water contents. An overview of the problems associated with the long-term storage of recalcitrant germplasm, and the rationale behind such rapid cooling approach are presented in Chapter 1 of the present thesis. Subsequent chapters report investigations on the effects of variables required to dry and cryopreserve embryonic axes with minimum damage, in keeping with this approach. Collectively, those studies aimed at establishing a robust cryopreservation procedure for the conservation of recalcitrant germplasm with broad applicability across species. The approach presently adopted entailed manipulating the water content of excised axes using rapid drying to discrete water content ranges, and also using different methods to cool axes to cryogenic temperatures at various rates. The calorimetric properties of water in axes were investigated for Camellia sinensis (L.) O. Kuntze using differential scanning calorimetry (DSC). For all species, the effect of any drying or cooling treatment tested was determined by assessing the survival of axes in vitro, which provided the most reliable indicator of cellular damage. Additionally, the effects of different treatments upon the structural and functional integrity of axes were assessed using light and electron microscopy as well as measurement of electrolyte leakage. The studies undertaken are presented in a similar sequence to that in which they took place during the course of the experimental phase of this work. These are summarised below. Partial drying plays a pivotal role in the approach developed, and microscopy has contributed towards increasing present understanding of desiccation damage. Microscopy was used to determine the effects of drying rate upon the ultrastructure of recalcitrant axes. It was necessary to find reliable protocols to prepare specimens for light and electron microscopy that did not alter the architecture of the cells in the dry state. Freeze-substitution and conventional aqueous fixation were compared in Chapter 2 using variously dried material from three species. The results obtained revealed that an unacceptably high extent of artefactual rehydration occurs during aqueous fixation, and highlight the need for anhydrous processing of dehydrated samples. Significantly, that study also revealed that many cellular events commonly associated with desiccation damage (e.g. withdrawal, tearing and/or vesiculation of the plasmalemma) are not seen in freeze-substituted preparations, and are likely artefacts of aqueous fixation. Freeze-substitution was used subsequently (Chapter 3) to assess the effects of slow drying (2 - 3 days) or rapid drying (min) upon the survival of embryonic axes of jackfruit (Artocarpus heterophyllus Lamk.) Results confirmed the beneficial effects of rapid drying, and also provided insights into ultrastructural changes and probable causes underlying cellular damage that occur during a drying/rehydration cycle. Efforts subsequently focused on determining the effect of cooling rate upon survival of recalcitrant axes at various water contents. The study on embryonic axes of recalcitrant camellia sinensis (tea; Chapter 4) tested the hypothesis that rapid cooling facilitates survival of axes at higher water content by restricting the growth of ice crystals to within harmless dimensions. The presence of sharp peaks in DSC melting thermograms was indicative of decreased survival in vitro. These peaks were attributed to the melting of ice crystals sufficiently large to be detected by DSC as well as to cause lethal damage to axes. Increasing the cooling rate from 10°C min-1 to that attained by forcibly plunging naked axes into sub-cooled nitrogen increased the upper limit of water content facilitating survival in vitro from c. 0.4 to 1.1 - 1.6 g H20 g-1 (dry mass [dmb]). Subsequent studies tested whether a similar trend occurred in other recalcitrant species cooled under similar conditions. In order to investigate further the relationship between water content, cooling rate and survival it was necessary to achieve cooling rates reproducibly, and to quantify these reliably. The plunging device required to achieve rapid cooling, and instruments required to measure the cooling rates attained, are described in Chapter 5. That study investigated the effects of cryogen type, depth of plunge and plunging velocity on the cooling rates measured by thermocouples either bare or within tissues of similar size and water content as encountered in cryopreservation experiments. This plunger was used in subsequent studies to achieve the fastest cooling conditions tested. Favourable cooling conditions were selected, and the efficacy of this procedure to cryopreserve relatively large axes was tested (Chapter 6) using embryonic axes of horse chestnut (Aesculus hippocastanum L.) Axes at water contents above c. 0.75 g g-1 could not be cooled faster than c. 60°C S-1, but cooling rates of axes below this water content increased markedly with isopentane, and to a lesser extent with subcooled nitrogen. Axes were killed when cooled at water contents above 1.0 g g-1 but survived fully (albeit abnormally) when cooled in isopentane between 1.0 and 0.75 g g-1. Complete survival and increasingly normal development was attained at water contents below 0.75 g g-1, especially if isopentane was used. The study on horse chestnut axes emphasised that water content and cooling rate are co-dependent during non-equilibrium cooling. Accordingly, that study could not determine whether survival at lower water contents increased because of the corresponding increase in cooling rates measured, or because of the higher cytoplasmic viscosity that resulted from drying. That uncertainty was addressed by the study discussed in Chapter 7, using axes of the trifoliate orange (Poncirus trifoliata [L.] RAF.) That study investigated the effect of cytoplasmic viscosity upon survival of axes cooled and warmed at different rates. Survival and normal development was high at lower water contents, and seemingly independent of cooling rate at about 0.26 g g-1. At higher water contents the range of cooling rates facilitating survival became narrower and displaced towards higher cooling rates. This study revealed two conspicuous inconsistencies that questioned the beneficial effect of rapid cooling. Firstly, the fastest cooling rates did not necessarily facilitate the highest survival. Secondly, survival of fully hydrated axes was higher when cooled under conditions that encouraged - rather than restricted - the growth of intracellular ice crystals. These inconsistencies were explored further using embryonic axes of silver maple (Acer saccharinum L.). Freeze-fracture replicas and freeze-substitution techniques provided valuable insights into the way in which ice crystals were distributed in cells cooled using different methods at rates ranging between 3.3 and 97°C S-1. Extensive intracellular freezing was common to all treatments. Unexpectedly, fully hydrated axes not only survived cryogenic exposure, but many axes developed normally when cooled using the relatively slower methods (77 and 3.3°C S-1) if warming was rapid. The most conspicuous ultrastructural difference between plunge cooling and the relatively slower methods was the exclusion of ice from many intracellular compartments in the latter. It is possible that even the fastest warming cannot prevent serious cellular damage if ice crystals form within such 'critical' compartments. It is proposed that the intracellular location of ice is a stronger determinant of survival that the size attained by ice crystals. The study of A. saccharinum also investigated the recovery of axes cooled fully hydrated either rapidly (97°C S-1) or slowly (3.3°C S-1). This facet of the study showed that cell lysis became apparent immediately after warming only where damage was most extensive. In other cells damage became apparent only after 2.5 to 6 h had elapsed, thus cautioning against inferring survival from the ultrastructural appearance of cells immediately after warming. Microscopy enabled cell repair as well as the pattern of growth of cryopreserved tissues to be appraised at the cellular, tissue and organ levels. Similar studies are required to understand further the nature of freezing damage, and how those events affect cell function. The salient trends observed in previous chapters are brought together in Chapter 9.