Pod Dehiscence

Introduction

Seed production is an invaluable process in all higher plants as it provides the means for propagation. During development, oilseed rape undergoes the

Prior to dehiscence a number of anatomical, molecular and biochemical changes occur within the pod, culminating in shatter due to a combination of components, but primarily from the generation of tensile and torsional forces within the pod between the lignified valve edge cells (VEC) of the endocarp and the unlignified cells of the dehiscence zone (DZ). This is coupled with the weakening and loss of cohesion of the DZ cell walls due to stimulated hydrolytic enzyme activity, thus rendering the pod susceptible to shatter under conditions of elevated physical stress (eg. wind or harvesting machinery).

Fruit dehiscence and the associated seed loss occurs in a number of crop plants (McGraw & Beuselinck, 1983; Erskine, 1985). Pod shatter in B. napus is known to have economic significance in terms of seed loss even though the loss of seed dispersal mechanisms are normally considered one of the first steps towards the domestication of a crop plant. For example, Bilsborrow (1985) reported field losses of 5-10% of seed yield due to shatter and a potential loss of 50% in adverse conditions prior to harvest (Macleod, 1981). Lutman (1993) estimated that seed loss can accumulate to reach up to 10,000seeds/m². Subsequently both timing and harvesting methods are factors in limiting field losses. In conjunction with yield loss, dehiscence also presents the problem of volunteer, or run-away plants that contaminate the following crops, with about 23% of winter cereal crops being contaminated with oilseed rape in the UK (Whitehead & Wright, 1989). These volunteer plants may act competitively as weeds (require herbicide application) or cause genetic contamination as a result of cross-pollination which would prove significant if the cultivars were of different varieties (eg. low erucic acid volunteers cross with high erucic acid field crops).

As a result of these serious seed losses, much research into pod development has been conducted over recent years to establish the factors involved, with an aim to modifying them to form shatter resistant varieties, such as the 'conventionally' bred varieties of B. rapa (turnip rape) and B. juncea.

For pod shatter to be limited, harvesting tends to occur prior to complete seed maturation (asynchronous ripening) which has the adverse effect of collecting immature seeds. These seeds don't directly affect seed yield, although they do effect the oil quality because they contain more moisture and chlorophyll which are undesired traits.

The regulated control of hydrolytic enzyme activity plays a crucial role in the degradation of the cell walls in the DZ. The enzymes identified are comparable to those in abscission and include the elevated activity of both cellulase (Beta-1,4-glucanase) (Meakin & Roberts, 1990) and polygalacturonase (PG, pectinase) (Jenkins et al, 1996).

So far it had been shown that the cells of the dehiscence layer separate, but there was no evidence to suggest any loss of cellular cohesion. Josefsson (1968) was first to identify the cellular basis of dehiscence, as he recorded a decline in the pectic material of dehiscence zone cell walls. Pectins constitute a major part of the cell wall middle lamellae, suggesting a role for enzymic mediation. The dissolution of the middle lamella and its solubilisation of pectins are closely related to the abscission process (Sexton & Roberts, 1982). Josefsson (1968) also noticed that dehiscence (valve shedding) doesn't necessarily succeed zone weakening, therefore stresses created by pod desiccation and external physical factors (eg. wind) are also required. This was explained by the identification of sclerenchymatous bridges connecting the valves to the replum, so as to require the additional separational force.

Picart (1983) supported these observations by reporting the occurrence of cell separation due to complete cellular autolysis of the DZ, becoming progressively (ie. regulated) more apparent from mid pod development onwards. He too was unable to offer any insight into the potential enzymes involved.

The following study by Picart & Morgan (1986) discovered that the application of synthetic auxin (4-chlorophenoxyacetic acid, CPA) causes delayed maturation and dehiscence in pods. CPA acts to delay the programmed autolysis of the thin-walled parenchyma cells of the DZ and reduce the rate of pod wall senescence, thereby reducing the rate of desiccation of the pod walls and seeds. The equivalent antagonistic effect of auxin on abscission was demonstrated by Sexton & Roberts (1982), further highlighting their homology. Although crudely similar, many cytoplasmic changes differ, including the activation of the endomembrane system reported during abscission (Sexton & Roberts, 1982) which is not apparent during DZ cell maturation.

The anatomical features of dehiscence had now been characterised, although there was little knowledge associated with the biochemical activity within the DZ that could cause the observed loss of cellular cohesion. Meakin & Roberts (1990a) summarised the anatomical changes, that pod shatter is mediated by a senescence-related autolysis of cells within the DZ. This is supported by DZ tissue-specific cytoplasmic changes (eg. reduction in both volume and organelle content) that proceed replum lignification (40 days prior to separation) and are comparative to events during senescence as DZ cell wall degradation occurs following rupture of the plasma membrane. This would therefore require the release of protoplasm containing compounds capable of cell wall degradation (eg. cellulase, polygalacturonase & pectin methyl esterase). Such mechanisms of cell autolysis have been described in many other systems (O'Brien, 1970).

The paper by Meakin & Roberts (1990b) provided the first evidence as to the presence and role of isolated cell wall degrading enzymes. The known loss of cell cohesion as described by Josefsson (1968) was shown to be associated with a DZ cell-specific increase in activity of the hydrolytic enzyme, cellulase (Beta-1,4-glucanase). However an increase in the activity of the enzyme, polygalacturonase (PG, pectinase) would have explained the loss of pectic material from the DZ cell walls, but was found to be unchanged. Pod ethylene production was shown to peak at 40 DAA, in accordance with the tissue-specific cellulase activity, thus illustrating the antagonistic effect of auxin and ethylene with respect to the delayed pod maturation after CPA application (Picart & Morgan, 1986). They concluded that the loss of cellular cohesion during dehiscence was due to the disruption of the middle lamella as a result of enzymic degradation (ie. cellulase).

The observed rise in cellulase activity was significant from 40 DAA even though visible cell wall degradation proceeds it by 15-20 d, a lag phase greater than that for abscission (Sexton et al, 1985). This maintained integrity of the cell walls during elevated enzyme activity has also been observed during citrus leaf abscission (Huberman et al, 1983). The walls of the VEC are protected by heavy lignification, but the cells of the DZ confer no obvious resistance to cellulase, so therefore between 40-60 DAA the cellulase must be confined, away from its apoplastic substrates. This was supported by the recordings that increased numbers of DZ cells undergo protoplasmic autolysis from 60 DAA, thereby providing a mechanism for releasing their contents (ie. cellulase) into the apoplastic region in order to facilitate DZ cell wall dissolution.

Meakin (1988) discovered that pod midge (Dasineura brassicae) infestation was found to induce premature shatter. It was expected that the mechanisms used by the pod midge may further elucidate the fundamental changes that occur during shatter. Unfortunately their infestation induced a process which showed closer homology to abscission events as there was no distinct desiccation of cells in the DZ, instead they merely lose cellular cohesion so as to render them susceptible to shatter under physical environmental stress. As with abscission, the tissue-specific increases in cellulase and PG activity occurred in association with the pod midge damage (Meakin & Roberts, 1991).

Coupe (1993) conducted an investigation into the changes in gene expression during pod development of Brassica napus by the technique of comparative genome hybridisation (CGH). Unlike the results from Meakin & Roberts (1990), he isolated and identified three mRNA's that accumulate immediately prior to dehiscence, one encoding PG (pectolytic enzyme). This is in accordance with the findings presented by Josefsson (1968) all those years ago, with respect to the loss of pectic material from the middle lamella prior to dehiscence! Jenkins et al (1996) further characterised this mRNA and termed the equivalent full length cDNA, SAC66. The other two mRNA's encoded a proline-rich protein (SAC51) (Coupe et al, 1993) and a protein which is homologous with oxidoreductases (SAC25) (Coupe et al, 1994).

Dehiscence and abscission are regarded as comparable systems as both involve the controlled degradation of cell wall material and the cell separation of a specific layer of cells. The effects of the plant hormones, ethylene and auxin (indole-3-acetic acid, IAA) as regulators are important factors in the timing of abscission (Gonzalez-Carranza et al, 1998), but contrastingly their role in dehiscence is unclear. Meakin & Roberts (1990) showed that ethylene levels peak at 40 DAA that correlates with an increase in cellulase activity, although Meakin (1988) was unable to induce dehiscence by ethylene exposure. Chauvaux et al (1997) investigated the role of auxin (IAA) during cell separation in the DZ to reveal that IAA peaks at 40 DAA, then progressively declines prior to dehiscence in accordance with the increase in cellulase activity in the DZ. Also, pod treatment with synthetic auxin (CPA) resulted in a delay in both cellulase activity and cell separation in the DZ, thus indicating that auxin plays a role in regulating the timing of pod dehiscence.

Currently, as a result of these enzyme identifications much work is aimed at manipulating their gene expression to study whether it affects the timing or capacity of B. napus to undergo pod shatter.

So far the process of dehiscence is basically understood as changes in the respective enzyme activities which are involved in the middle lamella dissolution have been characterised. However, the underlying mechanism of gene expression regulation for both spatial and temporal control must require an intricate signal transduction pathway. Previous work in other plant species (eg. Arabidopsis thaliana) has revealed proteins analogous to the bacterial two-component systems and are considered to be involved in the reception and transduction of extracellular signals (Chang et al, 1993; Urao et al, 1998). Subsequently Whitelaw et al (1999) investigated using differential display, the genes specifically expressed during pod development in Brassica napus. They isolated an up-regulated mRNA (respective cDNA termed SAC29), encoding a putative individual response regulator protein and attempted to ascertain its role as a signalling molecule during the events leading to shatter. After comparisons with various bacterial and the related Arabidopsis thaliana systems, there were a number of potential roles for SAC29 in dehiscence. It is up-regulated specifically in the DZ prior to dehiscence and itself may be regulated by phytochromes, such as ethylene (Meakin & Roberts, 1990b). They believe that once induced (autophosphorylated in response to an external elicitor), pSAC29 interacts with an upstream histidine kinase to transfer a phosphoryl group to pSAC29. Such phosphorylation is associated with a conformational change that enables it to interact either directly with DNA as a transcription factor, or via a MAP kinase system so as to initiate a cascade that culminates with the alteration of dehiscence related gene expression (up- and/or down-regulation). Further work is required in this field to determine the mechanisms of gene regulation before methods of dehiscence manipulation can be established for commercial crop production.

Although there is a lot of interest in oilseed physiology, much of the research occurs in the related Arabidopsis thaliana as it provides a model experimental system which can later be applied to engineer cruciferous crop plants such as B. napus for shatter resistance.