In search of the Holy Grail – a further step in understanding metal hyperaccumulation?

Abstract

Hyperaccumulation of heavy metal ions is a striking phenomenon exhibited by < 0.2% of angiosperms. These hyperaccumulator plants are considered endemic to metalliferous substrates, and are characterized by their tolerance and sequestration of exceptional quantities of metals like Zn, Ni and Cd in their shoots at concentrations that would be toxic to ‘normal’ plants. The last decade has seen an exponentially increasing interest in the phenomenon of metal hyperaccumulation since the property has been proposed as harnessable in the emerging technology of phytoextraction, both in terms of clean-up of contaminated land and phytomining otherwise noneconomic mineral deposits. Most recently, there has been a plethora of papers attempting to dissect the mechanisms of metal uptake, transport and accumulation, both at the physiological and molecular level, largely in the ‘model’ plant Thlaspi caerulescens and its close relatives. In this relentless race to identify the genes responsible, certain important issues have largely been cast to one side – notably, plant performance under natural field conditions and the need for more conventional genetic analysis of the hyperaccumulator trait for phytoextraction to become a commercial reality rather than the justification for even more sophisticated experimentation. Two papers in this issue (Bert et al. (2002) pp. 47–57; Macnair (2002), pp. 59–66) go some way to redress this balance and extend our knowledge of Zn and Cd hyperaccumulation in Arabidopsis halleri, a plant which is increasingly exciting the ‘gene tweakers’ since it was strategically moved from the genus Cardaminopsis in 1997 on the basis of rDNA analysis. ‘This is most comprehensive screening of any single hyperaccumulator species to date’ Like T. caerulescens, A. halleri has long been known as a metallophyte but only recently investigated for its heavy metal tolerance and hyperaccumulation of Zn (Bert et al., 2000; Zhao et al., 2000) and Cd (Dahmani-Muller et al., 2000). It is a notable member of similar metallophyte communities throughout Europe and appears to readily colonise newly contaminated sites. Unlike Thlaspi, however, it seems to be largely allogamous, seeds profusely over a longer growing season and can also spread vegetatively by stolons. The two species occur in both contaminated and uncontaminated areas and so provide ideal opportunities for comparative techniques in testing the evolutionary basis of metal tolerance and hyperaccumulation. Samples from all populations of metal-hyperaccumulating plants must be analyzed to identify polymorphisms in the accumulation trait. Consequently, the ‘model’ hyperaccumulator T. caerulescens has been much screened in the search for new and more extreme ecotypes, uncovering some remarkable Cd-hyperaccumulating populations (McGrath et al., 2001a; Lombi et al., 2002). The papers of Bert et al. and Macnair present perhaps the most comprehensive screening of any single hyperaccumulator species to date, examining metal concentrations in 33 and 17 populations of A. halleri, respectively. Bert et al. explore the interpopulation variation in Zn- and Cd-accumulation by A. halleri growing in central Europe. They sampled leaves and soils from established populations of A. halleri with > 4 km between nearest neighbours. This study was particularly novel because 13 of the sites were not contaminated with metals, and is certainly the most extensive investigation of any hyperaccumulator plant growing on ‘clean’ soils. Significantly, A. halleri growing on the clean soils bioaccumulated metals to the same order of magnitude as those growing on the contaminated soils, suggesting that Zn- and Cd-accumulation are probably constitutive traits in this species and not in any way a function of the soil metal content. There was, however, some variability in the absolute concentrations of metals accumulated. Some field samples had lower metal concentrations than the nominal thresholds of 10 000 and 100 mg kg−1 for Zn and Cd hyperaccumulation (Baker & Brooks, 1989). Given these lesser levels of metal accumulation, though several-fold higher than ‘normal’ plants on metalliferous soils, it poses the question: can we now define all populations of A. halleri as Zn- and Cd-hyperaccumulating sensu stricto? Macnair took the next logical step in dissecting the Zn-hyperaccumulating ability of A. halleri by growing plants from 17 populations under standardized conditions. This erudite study employed nondestructive determination of Zn concentration in the shoots to look at the effects of different external Zn concentrations supplied sequentially to the same individuals. It confirmed Zn hyperaccumulation as a constitutive trait in A. halleri from the sites sampled. Macnair then demonstrated that the heritability of Zn accumulation was between 25 and 50%, the highest yet recorded for any hyperaccumulator, probably because of the outbreeding nature of A. halleri. Intriguingly, the genetic variability of Zn accumulation in A. halleri was manifest more when grown in low-Zn media than in high. He suggests that in high Zn media ‘all plants accumulate substantial amounts of zinc, and the variation between plants with different phenotypes is more likely due to chance, local environmental effects, or ontological variation than genes.’ This significant finding highlights the importance of evaluating the methodology, especially the concentration of metal used, when assessing the results of experiments testing metal hyperaccumulation in hydroponic experiments. Moreover, experiments such as these, which confirm the heritability of hyperaccumulation and identify the optimal conditions under which selection will be most successful, should offer potential for breeding improved hyperaccumulators for commercial application. One notable issue of general importance highlighted by these two studies of A. halleri is the definition of hyperaccumulation and how the nominal thresholds should be interpreted. Various authors have chosen to employ the operational definitions given by Baker & Brooks (1989) very literally or liberally. The criteria suggested are merely guidelines for recognizing extreme behaviour, not absolute. Thus for some metals, notably Zn, individual plants are regularly found to have intermediate (but very high) shoot Zn concentrations, frequently in the 1000 s mg kg−1 range but not as high as 10 000 mg kg−1. This is clearly evident in the field data presented in these two papers. Is this then hyperaccumulation or not? The answer of course is ‘yes’ when the spread of concentration data is considered and the working definition of Reeves (1992) applied. The basis for this is elegantly argued in Macnair’s paper where the relative impacts of environmental (soil Zn concentrations) and genetic effects are explored. The ultimate confirmation of hyperaccumulator status can only come from hydroponic experimentation where it is emerges that: the plant is tolerant to extremely high concentrations of bioavailable metal; it accumulates metal in its shoot (leaf) tissues to exceptional concentrations relative to other genotypes/species; and the shoot/root or leaf/root quotient for metal concentration is > 1 indicating preferential partitioning of metals to the shoot (Baker, 1981). For example, Shen et al. (1997) demonstrated that all the Zn accumulated in the roots of T. caerulescens was rapidly translocated to the shoot when the Zn was abruptly withheld from solution-cultured plants. This argument provides a physiological definition. Here again, however, in their enthusiasm to report new hyperaccumulator plants, many authors have regularly assumed that if a normally nonaccumulating plant can take up > 10 000 mg kg−1 Zn from hydroponic culture it can be designated hyperaccumulator status. Not so: almost any plant can be made to do this, albeit with lethal consequences. Hence, caution is required. Zinc- and Cd-hyperaccumulation appear to be constitutive traits in A. halleri and T. caerulescens, even if variable between populations. The fact that these plants are able to bioaccumulate when growing in environments with low Zn status raises a suite of questions as to the mechanisms conferring the response, the evolutionary history of hyperaccumulation and its adaptive significance. Our understanding of the internal processes that confer the hyperaccumulation phenotype is advancing in leaps and bounds, and, for species, in the genera Thlaspi and Alyssum at least, the mechanisms of transport, tolerance and sequestration are partially elucidated (Lasat, 2002). Ironically, the constitutive expression of hyperaccumulation is a hindrance to progress in dissecting its physiological and genetic basis, because there are no nonaccumulating genotypes for comparative studies. There are still many conundrums that must be unraveled to fully understand constitutive hyperaccumulation and its evolutionary significance. The most immediate enigma is just how does A. halleri manage to scavenge Zn so effectively from soils of low Zn status? The same high accumula-tion of Zn is found in T. caerulescens growing on low-Zn soils (Meerts & Van Isacker, 1997; Escarréet al., 2000). Identifying and manipulating the processes occurring in the rhizosphere of these species is crucial if phytoextraction is to be successful (McGrath et al., 2001b). Perhaps more fundamental to our scientific understanding of hyperaccumulation is the uncertainty relating to the evolutionary history of the hyperaccumulation trait. Why does Zn hyperaccumulation in A. halleri occur within the entire species and not just those populations on metalliferous soils? Hyperaccumulation is clearly heritable (Macnair et al., 1999; Macnair, 2002), so where and under what selective pressures did the trait evolve? There is also little or no information on when and how such disjunct populations have arisen and persist. The absence of metal-enrichment in the soil at ‘clean’ sites suggests that selection for the development of the trait in situ at these sites is improbable. Indeed, Macnair suggests that the populations on these low-metal sites may be adventive rather than relict. These are but a few of the unknowns in the ecological and causal factors leading to hyperaccumulation and the variability within this trait. The answers to these questions, however, are likely to emerge all too slowly due to the small number of new phylogenetic and ecological studies of hyperaccumulators. Our comprehension of the processes by which metals are accumulated by plants is progressing rapidly. The momentum of discovery is further likely to increase, driven by the ‘hype’ (Ernst, 2000) of exploiting the phenomenon in cost-effective methods of extracting metals from soil. Two cautions must be exercised however. First, there appears to be a widening gap between science and practicality. Successful applications of metal-accumulating plants for remediation or metal recovery purposes, even at the pilot-scale, are still alarmingly few and far between. It has become almost expedient in concluding any paper or justifying a research grant application involving work on hyperaccumulators to state that ‘this study has potential applications to the development of phytoremediation technology’. In 1995 Ernst predicted it would be a minimum of 20 yr before any molecular biology involving hyperaccumulator plants could find practical application – if at all (Ernst, 1995). His prediction will be true unless there are more efforts to develop implementation. Second, considerable expectations are placed on genetic modification (GM) to produce idealized plants for use in commercial phytoextraction. While we are confident that GM can produce optimized and enhanced plants to bring this technology to fruition, the ever-increasing public concern over the release of GM organisms could force regulators to veto their use. This is certainly the case in much of the developed world at this time. It would therefore be prudent to complement GM approaches with searches for ‘natural’ accumulators and further explore conventional plant breeding practices (Chaney et al., 2000), which has the additional benefit of cataloging and, we hope, conserving the unique global biodiversity in mineralized environments (Whiting et al., 2002).