Why don't the gas spaces of submerged organs of wetland plants flood extensively when damaged? In addressing this intriguing question, Soukup et al. (pp. 71–75 in this issue) report on the role of rhizome diaphragms as barriers to flooding in Phragmites australis. This should prompt some reappraisal of the ways in which flooding resistance can be realized, even perhaps in undamaged organs.

Most emergent wetland macrophytes have an abundance of interconnected internal gas space, much of it in the form of large voids transversely partitioned at intervals by perforated cellular plates termed diaphragms. Functionally, it provides a low-resistance pathway for internal oxygen transport to support the respiratory needs of submerged and buried organs (Armstrong, 1979; Armstrong et al., 1988; Crawford, 1992) and facilitates carbon dioxide removal. However, it does more than this, since it enables oxygen to be released from the root to where it can support aerobic microbial activity in otherwise anaerobic sediments, and phytotoxin immobilization or destruction (Armstrong et al., 1992; Begg et al., 1994; Gilbert & Frenzel, 1998). This oxygen release is regarded by some as a valuable aid to effluent purification by constructed wetlands. Perhaps a less desirable property of this gas-space provision is its recently discovered role in enhancing the emissions of greenhouse gases such as methane from wetlands (Brix et al., 1992; Chanton & Whiting, 1996; Crutzen, 1991; Dacey & Klug, 1979).

How appropriate that at the turn of a century that has overseen a second industrial revolution in agriculture, there is an increasing appreciation of the central role played by the mycorrhizal symbiosis, first described in the latter part of the nineteenth century (Frank, 1885). Simply stated, nearly all families of plants form root symbiotic organs, termed mycorrhizas, with soil fungi belonging to all the main phyla, namely Zygomycotina, Ascomycotina, Basidiomycotina and the Fungi Imperfecti (Harley & Harley, 1987). The importance of this symbiosis in controlling plant nutrient status and growth is well established (Smith & Read, 1997), but a report in this issue now provides, long awaited, nitrogen and phosphorus budgets for mycorrhizal trees (Perez-Moreno & Read, pp. 301–309).

The view that nutrient acquisition by most plants growing in natural ecosystems is mediated by mycorrhiza-forming symbiotic fungi is now largely accepted. With this perception comes the need to learn more about the identity of the organisms responsible for these key processes. Herein lies a challenge taken up by Vrålstad et al. on pp. 549–563 in this issue.

Monstera acuminata is a plant with a problem – it attains a large body size, yet has no secondary growth. For many years, observers have noted the differences in architecture between monocots and woody dicots, particularly in terms of overall size and comparisons of tree-like forms. However, very little is known about the functional morphology of smaller-bodied and ecologically significant climbing and hemiepiphytic monocots, of which M. acuminata is one. In its wider context, the diversity of plant growth forms contributes to the complex structure of many terrestrial ecosystems, where herbs, trees, shrubs, vines, lianas, root-climbers, epiphytes and hemiepiphytes may be said to comprise the ‘skeleton’ of the ecosystem (Speck & Rowe, 1999). This is seen readily, for example, in the humid tropics, which arguably possess the highest taxonomic and structural diversity that has ever been formed by plants on earth. In this issue, López-Portillo et al. (pp. 289–299) investigate the development of plant architecture in Monstera, and demonstrate a range of novel hydraulic parameters.

SUMMARY 367

I. INTRODUCTION 367

II. NUMBER 368

III. SIZE 379

IV. AIR SPACE IN THE SEEDS 381

V. FLOATATION AND DISPERSAL 383

1. Air 383

(a) Physical considerations 383

(b) Dispersal 387

(c) Birds 415

2. Water 416

(a) Physical considerations 416

(b) Dispersal 416

VI. CONCLUSIONS 417

Acknowledgements 417

References 418

Orchid seeds are very small, extremely light and produced in great numbers. Most range in length from c. 0.05 to 6.0 mm, with the difference between the longest and shortest known seeds in the family being 120-fold. The ‘widest’ seed at 0.9 mm is 90-fold wider than the ‘thinnest’ one, which measures 0.01 mm (because orchid seeds are tubular or balloon-like, ‘wide’ and ‘thin’ actually refer to diameter). Known seed weights extend from 0.31 lg to 24 μg (a 78-fold difference). Recorded numbers of seeds per fruit are as high as 4000000 and as low as 20–50 (80000–200000-fold difference). Testae are usually transparent, with outer cell walls that may be smooth or reticulated. Ultrasonic treatments enhance germination, which suggests that the testae can be restrictive. Embryos are even smaller: their volume is substantially smaller than that of the testa. As a result, orchid seeds have large internal air spaces that render them balloon-like. They can float in the air for long periods, a property that facilitates long-distance dispersal. The difficult-to-wet outer surfaces of the testa and large internal air spaces enable the seeds to float on water for prolonged periods. This facilitates distribution through tree effluates and/or small run-off rivulets that may follow rains. Due to their size and characteristics, orchid seeds may also be transported in and on land animals and birds (in fur, feathers or hair, mud on feet, and perhaps also following ingestion).

Since the beginning of the Bronze Age the acquisition of metals and the mastering of metal processing has been a driving force in human civilisation. It is difficult to imagine, however, how a plant could possibly profit from accumulating excessive amounts of a metal for which there is no metabolic need. Leaving behind an environmentally degraded wasteland, human mining and smeltering activities contributed to providing the niches for such plants, sometimes even accomplishing seed dispersal – through seeds attached to the boots of migrating miners. A stunning example of this is found in the healthy populations of Thlaspi caerulescens in the Cevennes (southern France), the leaves of which have been found to contain the toxic and non-essential metal cadmium (Cd) at concentrations between 1000 and 3000 mg kg−1 dry biomass (R. D. Reeves, unpublished). In this issue, experiments under controlled experimental and glasshouse conditions are reported, confirming an extraordinary level of Cd accumulation and tolerance in these populations (Lombi et al., pp. 11–20).

Summary 167

I. INTRODUCTION 168

II. THE IMPACT OF PROTEIN STRUCTURAL INFORMATION ON THE UNDERSTANDING OF THE PRIMARY REACTIONS IN PHOTOSYNTHESIS 169

1. The structure of the purple bacterial reaction centre 170

2. The structure of PSI 172

3. The structure of PSII 173

4. The structure of antenna complexes 176

(a) Light-harvesting complexes from purple photosynthetic bacteria 177

III. DEVELOPING OVEREXPRESSION SYSTEMS FOR THE STUDY OF PLANT PROTEIN 188

1. Expression of heterologous proteins in Escherichia coli 189

2. Protein expression in the methylotrophic yeast Pichia pastoris 189

3. Expression of recombinant protein in insect cells: the baculovirus system 190

IV. CONCLUSIONS 190

This review sets out the case that now is the time for plant science to establish the technologies required for routinely studying the structure and function of plant proteins. The impact that protein structural information can have is illustrated here with reference to photosynthesis. Our understanding of the precise molecular mechanisms of the light-reactions of photosynthesis has been transformed by the combination of high-resolution protein structural data and detailed functional studies. The past few years have been a particularly exciting time to be engaged in basic plant science research. The application of modern techniques of molecular biology has allowed many key questions to be addressed. The stage is now set for an even bigger revolution as the current plant genome sequencing projects are completed. If these advances are going to be fully exploited, plant science must get to grips with studying proteins, not just genes. Reliable methods for the overexpression of proteins in their native state coupled with routine access to structure determination must become the norm rather than the exception. In 1998 there were about 9000 protein structures deposited in the Brookhaven database. Very few of these are plant proteins. This trend will have to be reversed if research in molecular plant science is to fulfil its potential.

The effects of localized zinc (ZnO) and cadmium (CdS) enrichment on the allocation of root biomass, root length and partitioning of current assimilate within root systems of the Zn hyperaccumulator Thlaspi caerulescens were investigated using a rhizobox system. The rhizoboxes contained either homogeneous soil or juxtaposed control and metal-enriched soil. In the heterogeneous treatments the Zn-enriched soil contained 250, 500 or 1000 mg Zn kg−1. The plants consistently allocated c. 70% of their total root biomass and length into the metal-enriched soil. Moreover, 70% of the current assimilate (14C) allocated to the roots was localized in the Zn-enriched soil of the heterogeneous treatments. The concentration of Zn (250–1000 mg kg−1) in the enriched soil had no effect on these patterns of root allocation, nor were there significant effects of the Zn treatments on the total root or shoot biomass of the plants. The positive responses to the localized metal treatments were therefore characterized by a concomitant reduction in root allocation into the control soil. In contrast to T. caerulescens, when grown with juxtaposed control and Zn-enriched soil the non-accumulator Thlaspi arvense showed reduced root allocation into a patch enriched with 500 mg kg−1 Zn. In a further experiment, two populations of T. caerulescens that differ in their abilities to accumulate Cd were grown with juxtaposed control and Cd-enriched soil. The plants from a population that accumulated Cd also showed increased root biomass and root length allocation into the Cd-enriched soil. Plants from the population that did not accumulate Cd showed no such increase. The possibility that T. caerulescens forages for metals, and the precision of its root allocation with respect to localized metal enrichment is highlighted. The significance of these findings for the selection of hyperaccumulator plants for use in the phytoextraction of Zn and Cd from mine spoils (phytomining) and the phytoremediation of heterogeneous contaminated soils are discussed.

The effects of aluminium (Al) on dividing root-tip cells of Triticum turgidum were investigated with tubulin immunolabelling and electron microscopy. Aluminium affects the mechanisms controlling the organization of microtubule (MT) cytoskeleton, as well as tubulin polymerization, and induces the following aberrations in mitotic cells. (1) It delays the MT disassembly during mitosis, resulting in the persistence of preprophase MT bands in the late prophase cells, the presence of prophase spindles in prometaphase cells, and a disturbance in the shortening of kinetochore MT bundles in anaphase cells. (2) It interferes with the self-organization process of MTs into bipolar systems, inhibiting the formation of prophase and metaphase spindles. (3) Aluminium induces the formation of atypical MT arrays, which in the immunofluorescent specimens appear as ring-like tubulin aggregations in the cortical cytoplasm of the preprophase/prophase cells and as endoplasmic tubulin bundles in prophase and metaphase/anaphase cells; abnormal preprophase MT bands are assembled, consisting of atypical cortical and endoplasmic MT bundles, the latter clearly lining the nuclear envelope on the preprophase MT band plane. (4) It disorders the chromosome movements carried out by the mitotic spindle. In addition, after prolonged Al treatments chromatin condensation is inhibited. The outcome is greatly disturbed organization and function of the mitotic apparatus, as well as inhibition of cells from entering mitosis. This study shows that the MT cytoskeleton is a target site of Al toxicity in mitotic root-tip cells of T. turgidum. The possible mechanisms by which Al exerts its toxicity on MT organization and function are discussed.

The enzyme sucrose: sucrose 1-fructosyltransferase was partially purified from barley leaf growth zones. Four steps (ammonium sulphate precipitation and polyethylene glycol precipitation, followed by chromatography on Concanavalin A-sepharose and hydroxylapatite) yielded a 35-fold purification. The resulting preparation of 1-SST which still contained a number of different activities related to fructan metabolism, was subjected to preparative isoelectric focusing, and sections of the gel were analysed individually for 1-SST and related activities, using sucrose and 1-kestose as substrates. This procedure yielded a 196-fold purification and revealed the presence of two isozymes of 1-SST with pI values of 4.93 and 4.99, as determined by analytical isoelectric focusing of the corresponding fractions. Both isozymes produced glucose and 1-kestose when incubated with sucrose. In addition, small amounts of 6-kestose and tetrasaccharides were formed. In particular, one of the two 1-SST isozymes yielded fructose when incubated with 1-kestose, indicating that it also acts as a fructan exohydrolase. The other isozyme exhibited less fructan exohydrolase activity. Nystose was also degraded by the fructan exohydrolase activity but less than 1-kestose, whereas 6-kestose was not a substrate for the enzyme. Incubation of both 1-SSTs with different concentrations of sucrose showed that the enzyme was not saturated even at 500 mM. As for the barley sucrose: fructan 6-fructosyltransferase, both isozymes of 1-SST yielded two polypeptide bands of molecular weight 50 and 22 kDa upon sodium dodecylsulphate polyacrylamide gel electrophoresis, suggesting their close relationship to invertase (composed of two subunits of similar size), as previously reported for other plants.

The capacity to accumulate cadmium (Cd) and zinc (Zn) was compared in Thlaspi goesingense and four populations of Thlaspi caerulescens. Two populations of T. caerulescens were grown in hydroponics at five concentrations of Cd. In addition, plants were grown in pots containing compost in which three different concentrations of Cd and two concentrations of Zn were added. A field trial was conducted to compare Zn and Cd uptake by three populations of T. caerulescens on nine selected plots of the Woburn Market Garden Experiment (UK) which had been contaminated to different degrees with heavy metals owing to past applications of sewage sludge. Results show that the four populations of T. caerulescens had the same ability to hyperaccumulate Zn but were significantly different in terms of Cd accumulation. Two populations of T. caerulescens from Southern France accumulated much more Cd than the populations from Prayon (Belgium) and Whitesike (UK). Generally, uptake of Cd was not decreased by increased concentrations of Zn in the substrate. These results indicate that the mechanisms of Cd and Zn hyperaccumulation are not identical in this species. This is the first report of hyperaccumulation of Cd by T. goesingense, but the growth of this species was markedly reduced by the large concentrations of Zn in the substrate. Future work should focus on the differences between Cd and Zn uptake in hyperaccumulator plants at the species and population level.