7. Adverse Soil Mineral Availability

7.2.1 Elements in the Soil

The soil represents an immensely complex physical, chemical and biological substrate (Chap. 11, Sect. 11.​1.​1 for a more detailed description). Nutrient availability strongly varies in space and time. Soil types differ tremendously in mineral content. Large and element-specific fluctuations occur within a soil—for instance, depending on changes in pH (Fig. 7.3), water status or microbial activity. Gradients develop horizontally and vertically. Elements can be found in patches because of uneven distribution of factors influencing availability—for example, litter fall and decomposition. Mobility within the soil is strongly element-specific.

Soil consists of solid, liquid and gaseous phases. The mineral nutrient supply is influenced by all three phases. The solid phase contains most of the nutrients. Inorganic soil matter is a reservoir for nutrients such as K and Fe, and organic soil matter is the main N reservoir. Minerals originate mostly from the weathering of bedrock—the sediments that give rise to soil. The exception is nitrogen, which predominantly stems from nitrogen fixation (Sect. 7.4.2). Weathering rates are strongly influenced by age and environmental factors. For instance, many tropical soils are highly weathered and the soils are therefore depleted in plant-available phosphorus.

The liquid phase contains ions available for uptake by roots. Only a very small fraction of total soil minerals are in soil solution. The bulk is bound to soil particles (e.g. clay, humic acids) that carry mostly negative charges, thus providing binding sites for cations. Minerals are in a dynamic equilibrium between the two phases. Ion exchange processes result in slow release into the soil solution. The binding capacity of the particles is an important soil fertility factor, and the average particle diameter is therefore a parameter used for soil classification. The high binding capacity of smaller particles reduces leaching and thus increases reserves.

While roots tap nutrient resources by growth (Sect. 7.3.5), nutrient ions ultimately have to reach the root surface via mass flow and diffusion. Elements differ strongly in their mobility within the soil solution. Phosphate is about three orders of magnitude less mobile than nitrate or sulphate (Lambers et al. 2015) because of stronger interaction with soil particles. This has major consequences for the biology of nutrient acquisition. About 90% of all plant species live in a mycorrhizal symbiosis with fungi. The fungal hyphae greatly enhance the capacity to unlock immobile phosphate in the soil (Sect. 7.4.1).

Soil particles form pores that differ in size between macro- and micropores. They are partly filled with water and partly with air, depending on the type of the soil and the amount of precipitation (Fig. 5.1). Air-filled pores are important for the gas exchange of the respiring roots (autotrophic respiration) and of other soil organisms (heterotrophic respiration). Gas exchange of roots and soils influences nutrient availability. For example, CO₂ released from roots as the product of respiration dissolves in H₂O and forms hydrogen carbonate (HCO₃ ⁻) and H⁺. These ions can desorb nutrient ions from soil particles through ion exchange and thereby enhance bioavailability.

Box 7.1: Redox Potential of the Soil

The redox potential of soil results from the ratios between the oxidised forms and the reduced forms of metals:

(7.1)

where n denotes the number of electrons exchanged between the oxidised and the reduced form, a _Ox is the activity of the oxidised form and a _red is the activity of the reduced form.

Well-aerated soils have redox potentials of up to +0.8 V, and poorly aerated soils at the level of groundwater or peat soils have redox potentials of up to −0.35 V.

The reductive potential of the soil is characterised by the pe value. The p_e value (analogous to pH) is the negative log of the “concentration of electrons (n)” in the soil:

(7.2)

The conversion factor between E and p_e is: p_e = E (V) × 16.9.

The p_e values of a paddy rice field are between +4 (surface) and −3 (middle layer).

Strong variation exists between species and also within species (i.e. between ecotypes, cultivars, varieties) in their ability to acquire nutrients from soil. Thus, soil mineral availability has a strong influence on the distribution and composition of natural vegetation (Marschner 2012) (compare global soil map Fig. 11.​2). This is illustrated by widely used classifications such as calcicoles versus calcifuges—that is, plants thriving on alkaline lime-rich soil versus plants with a preference for acidic soil. Some plant species have evolved specific adaptations to particularly nutrient-impoverished soils—for example, highly weathered ancient soils in Australia and South Africa, or soils in cold climates with very slow mineralisation of organic matter.

7.2.2 Element Toxicity

For macroelements the adequate range is usually very broad. This is different for some micronutrients. Their reactivity is the reason why they can easily become toxic (Fig. 7.2). Thus, the concentration range between deficiency and toxicity is comparatively narrow. Mn availability can become very high in acidic soils, and the resulting Mn toxicity is a secondary problem of low pH besides the (far more important) Al toxicity (Sect. 7.5.4). A negative redox potential of flooded soil is another reason for high Mn availability and toxicity (Box 7.1). Plants in many arid regions such as southern Australia suffer from boron toxicity (Sect. 7.5.1).

Similarly, beneficial elements are quite often available in the soil at a concentration that exceeds the toxicity threshold. Unlike silicon, aluminium and sodium are required in only very small quantities for growth stimulation. Many soils and soil conditions exist, however, that are associated with high availability of these elements and their respective toxicities. Survival of plants in salt marshes entails the expression of particular salt tolerance mechanisms (Sect. 7.5.3). Growth in acidic soils where Al³⁺ becomes available for uptake by plant roots requires specific adaptations to cope with the associated toxicity (Sect. 7.5.4).

Another relevant aspect to consider is the potential availability of elements that have no biological function and are potentially highly toxic (Clemens 2006). Some are taken up into cells because of their chemical similarity to essential elements. The most important examples are arsenic (As) and Cd. The arsenate anion [(AsO₄)^3-] closely resembles the phosphate anion and cannot be discriminated against accurately enough by phosphate uptake systems. Cd is below Zn in the periodic table and therefore shares many chemical characteristics with this essential element. Again, transporters for the essential Zn²⁺ ion are not perfectly selective and transport Cd²⁺ ions as well.

Some plants have evolved the capacity to survive in soils with very high levels of toxic elements that far exceed the concentrations tolerable for most terrestrial plants. In the case of Na they are referred to as halophytes (Sect. 7.5.3). Vegetation on metal-rich sites is dominated by metallophytes (Sect. 7.5.2). The adaptations to such extreme habitats represent very instructive models for rapid evolution in action and will therefore be described from this angle.

7.3.1 Modulation of Nutrient Availability

Plant roots actively influence the rhizosphere—that is, the immediate vicinity of the roots—in order to change the availability of nutrients. Because of the important role of soil pH, acidification by proton pumping is one prevalent mechanism that enhances availability of Fe, Zn, B and Mn. Organic acids such as malate and citrate are among the major components of root exudates. Deposition of carbohydrates in the rhizosphere accounts for a substantial fraction of the 20–60% of photosynthetically fixed carbon that is transferred below-ground by plants (Kuzyakov and Domanski 2000). Release of organic acids can occur passively through damaged root cells and via controlled secretion through anion channels in the plasma membrane. Stimulation of the latter has been observed especially in response to phosphate deficiency. Citrate and other carboxylates can mobilise sparingly soluble phosphate adsorbed to Fe or Al oxides. Their exudation is particularly pronounced in cluster roots of certain plants adapted to phosphate-poor soils (Fig. 7.19). Phosphate can in addition be mobilised by the secretion of enzymes (e.g. phosphatases) that can hydrolyse organic P esters (Fig. 7.5).

Because of the complexity of the processes and the difficulties in experimentally accessing the rhizosphere, detailed molecular understanding of root exudates and their contribution to nutrient availability is still limited. In contrast, a well-understood and very important example not only of nutrient mobilisation is Fe acquisition (Kobayashi and Nishizawa 2012), which will therefore be a recurring theme throughout this chapter. Iron is a particularly problematic element with respect to bioavailability. The usage of Fe as a redox-active element in biological systems evolved at a time when conditions in the Earth’s atmosphere were reducing, making Fe readily available because Fe sulphides are highly soluble (Frausto da Silva and Williams 2001). With the advent of oxygenic photosynthesis, the bioavailability of Fe gradually and dramatically dropped by about eight orders of magnitude. In an oxidising atmosphere, Fe is mostly present as insoluble Fe oxides. Massive Fe precipitation resulted in the formation of red bands in sedimentary rock about 2.5 billion years ago. Thus, Fe is one of the most abundant elements in the Earth’s crust, yet it is scarcely available in many habitats because it is mostly present in the oxidised form Fe(III), which is barely soluble, especially at pH values above neutral (Fig. 7.3).

Terrestrial plants evolved two distinct strategies to mobilise and to take up Fe (Römheld and Marschner 1986). Strategy I, expressed by dicots and non-graminaceous monocots, consists of subsequent acidification, reduction and uptake steps. Protons are secreted into the rhizosphere by proton pumps (P-type H ⁺-ATPases) to enhance solubility of Fe(III). Plasma membrane–localised ferric reductases reduce Fe(III) chelate complexes to Fe(II), which is then taken up into root epidermal and cortex cells. This strategy is supplemented by the secretion of phenolic compounds such as coumarins, which may act as chelators and/or reductants of Fe(III) (Clemens and Weber 2016). Strategy II is characteristic for grasses. Fe(III)-chelating phytosiderophores such as mugineic acids are secreted by root cells. Phytosiderophores form complexes with Fe(III), which are substrates for specialised transporters that mediate the uptake of these complexes. Many of the proteins mediating these processes have been molecularly identified in Arabidopsis thaliana and maize. Loss-of-function mutants have demonstrated the essentiality of the different steps for growth under Fe-limited conditions. Maize mutants lacking the phytosiderophore uptake transporter Yellow stripe 1 (ys1) show characteristic Fe deficiency–caused chlorosis phenotypes of the leaves (Curie et al. 2001) (Fig. 7.6).

The synthesis and secretion of Fe-chelating molecules is strongly activated under Fe limitation. In rice, barley and maize, as representative grass species, several genes encoding enzymes of the mugineic acid synthesis pathway are up-regulated. Examples include nicotianamine synthase (NAS), nicotianamine aminotransferase (NAAT) and deoxymugineic acid synthase (DMAS) (Fig. 7.7). The same applies to the transporters for mugineic acid or Fe(III)–mugineic acid complexes. Variations within and between grass species in the ability to thrive on alkaline soil have been explained by differences in phytosiderophore secretion rates. Barley shows much stronger mugineic acid release than rice. When rice was engineered to produce more phytosiderophores through transfer of a more strongly expressed nicotinamine aminotransferase gene from barley, growth and yield on alkaline soil was significantly improved (Takahashi et al. 2001)—an example for the engineering of stress tolerance.

It is important to note that P and Fe are merely the two best-understood examples of active modulation of nutrient availability. The aforementioned rhizodeposition of organic compounds strongly influences the density and the activities of microbial communities around the root. These activities in turn have pronounced effects on nutrient chemistry (Sect. 7.3.2.1, the conversion of N sources). The questions as to whether and how plant roots actively recruit certain microbial populations (the rhizosphere microbiome) can only now begin to be addressed, owing largely to the revolution in DNA sequencing technologies (Bulgarelli et al. 2013).

7.3.2 Cellular Ion Transport Mechanisms

Mineral nutrients in the soil solution are practically always present as ions and carry an electric charge. Thus, they cannot cross biological membranes at sufficiently high rates without the involvement of proteins that form specialised pores allowing passage through the membranes. Ions can move into the cytosol and out of the cytosol. Movement from the exterior into the cytosol is called uptake; transport into the extracellular space is called efflux. All other transport processes occur across organellar membranes. Most relevant for mineral nutrition are uptake, efflux and transport into and out of the vacuole (Fig. 7.8). A large number of transport proteins mediating these processes are encoded in plant genomes, as we know from the model species A. thaliana and rice. For example, around 50 genes encode K⁺ channels and K⁺ transporters in A. thaliana alone (Ward et al. 2009).

Depending on the driving force, three categories of transporters are distinguished (Fig. 7.9). Facilitated diffusion refers to transport that is energetically favourable because it occurs along an electrochemical potential gradient. Typical proteins enabling facilitated diffusion are channels such as K⁺ channels (Figs. 7.11 and 7.12). Active transport moves an ion (or a metabolite) against an electrochemical potential gradient—for example, from the soil solution with a low concentration to the cytosol of a root cell with a high concentration (enrichment)—or an anion against the negative potential of the plasma membrane (by definition, the membrane potential of a cell is negative when there is a surplus of negative charges on the cytosolic side). Primary active transport is directly energised by the hydrolysis of adenosine triphosphate (ATP). Secondary active transport uses the energy supplied by a gradient in electrochemical potential through coupling of the energetically favoured movement of one molecule to the unfavourable movement of another (Fig. 7.9).

The dominant primary active transport in plant cells is the establishment of a proton gradient across the plasma membrane and the tonoplast by the activity of H ⁺-ATPases (and additionally of pyrophosphatases in the case of the tonoplast). This proton gradient, the proton motive force, provides the driving force for myriad transport processes in plants (Fig. 7.10). It can be used by carrier proteins (Fig. 7.9). Many of them co-transport protons with ions or metabolites. This co-transport can be a symport (i.e. both molecules move in the same direction) or an antiport (i.e. the molecules move in opposite directions). While most carriers involved in nutrient acquisition couple transport to the movement of protons and are therefore secondary active, there are many others that couple the transport of two metabolites or of a metabolite and an ion and are passive. Examples are the phosphate translocator in the inner plastid membrane, which exchanges phosphate with 3-phosphoglycerate, or the malate/oxaloacetate shuttle. A third type of carrier—and passive too—is the uniporter that mediates transport of one molecule along a potential gradient.

The activities of cation and anion channels depend on the cell’s external and internal ion concentrations, which establish a specific membrane potential (Fig. 7.10). The so-called rectifying channels allow charges (ions) to pass more easily either into the cells (inward rectifier) or in the outward direction (outward rectifier), thus also affecting the membrane potential (Fig. 7.11). The same holds for channels in the tonoplast membrane. At a membrane potential negative to the equilibrium potential of, for example, K⁺, inward potassium rectifiers support the flow of the potassium cation into the cell, rendering the membrane potential more positive until the equilibrium potential (also termed the Nernst potential) of K⁺ is reached. Anion-rectifying channels operate in a similar way with anions. Channels whose activities change the membrane potential can, on the other hand, be controlled by that potential. They are termed voltage-gated channels, in contrast to channels that are controlled by specific ligands (ligand-gated channels). Gating of these channels can control K⁺ mineral nutrition (Sect. 7.3.2.2), signalling, and abiotic as well as biotic stress responses. The regulation of inward- and outward-rectifying K⁺ channels plays a key role in controlling the turgor of guard cells and the apertures of stomata (Sect 6.​3.​3).

Depending on the nutrient ion in question, uptake into the root symplast—that is, the cytosol—has to be either energised or not. The plasma membrane potential of plant cells is negative (around −150 mV), owing largely to the proton pumping activity of H⁺-ATPases. Thus, cations such as K⁺ or Fe²⁺ can in principle move passively into the cytosol along an electric potential gradient through channels or uniporters. In contrast, anions such as phosphate, nitrate and sulphate enter a root cell against a potential gradient. This is enabled by H⁺-coupled symport (Figs. 7.8 and 7.10). Conversely, efflux out of the symplast into the apoplast for xylem loading requires energisation for the cations and is energetically favourable for the anions.

One of the hallmarks of plant mineral uptake is the existence of multiphasic uptake systems with varying affinities, as classically shown for K⁺ (Epstein et al. 1963). Depending on the concentration in the soil solution, low-affinity or high-affinity transport systems with affinities in the millimolar or micromolar range, respectively, are in operation (Fig. 7.12). As for enzymes, the affinity for the substrate is expressed as a K_M (Michaelis–Menten) value—that is, the substrate concentration at which the transport rate is half maximal. Typically, a plant possesses several isoforms of both low-affinity and high-affinity transporters, which vary in localisation and timing of expression (Fig. 7.13). This applies to both monocot and dicot plants.

7.3.2.4 Uptake of Other Nutrient Elements

Sulphate acquisition shows many similarities to nitrate acquisition. High-affinity uptake by proton-coupled symporters (in the case of sulphate these are referred to as sulphate permeases) mediate entry into the cytosol. All steps of sulphate assimilation occur in plastids. The metabolite that is analogous to glutamate in N assimilation is cysteine for S assimilation.

Uptake of Fe(II) by strategy I plants such as A. thaliana is mainly dependent on IRT1. The absence of this uniporter causes severe growth inhibition when Fe availability in the soil is low (Fig. 7.14).

Boron and the beneficial element Si are taken up as boric acid and silicic acid, respectively, by aquaglyceroporins of the nodulin 26–like intrinsic protein (NIP) family. These proteins form pores in the membrane for facilitated diffusion. The protein responsible for boron uptake in A. thaliana is NIP5;1. The very pronounced uptake of Si in rice is mediated by Lsi1. This transporter represents an example of the polar localisation of at least some nutrient transporters in root cells. Lsi1 is exclusively localised on the distal side facing the soil (Ma et al. 2006).

7.3.3 Modulation of Nutrient Uptake in Response to Deficiency

Plants need to be able to respond and acclimate to strongly fluctuating external nutrient availability. The major targets of acclimation are nutrient uptake capacities and the root architecture (described in Sect. 7.3.5). Generally relevant factors are the external supply of a particular nutrient and the internal status. Both can in principle be sensed and translated into a response. A second distinction is that between a local response (i.e. a cell or a tissue perceives a problem and triggers countermeasures) and a systemic response (i.e. leaves or another organ distant from the site of nutrient acquisition monitor the nutritional status and send a signal down to the root, where uptake is modulated appropriately).

A basal feature of plant nutrient acquisition is the switching from low-affinity uptake to high-affinity uptake when external concentrations drop below certain thresholds (Fig. 7.12). The activities of the different systems often correlate quite strongly with the expression levels of the genes encoding the transporters. Thus, transcriptional regulation is one key to the adjustment of uptake capacities. While low-affinity systems tend to be expressed constitutively, many high-affinity transporter genes (e.g. NRT2, AMT1, Pht1, KUP/HAK) have been shown to be up-regulated upon nutrient deficiency (Miller et al. 2009).

Another aspect of transcriptional control is activation in the presence of a substrate and repression by an end product. Nitrate serves as an important signal for nutrient availability and elicits the transcription of nitrate uptake transporter genes. Conversely, glutamine—as the product of ammonium transfer onto glutamate during nitrate assimilation—represents a negative feedback signal for nitrate uptake.

The P deficiency response is a relatively well-understood example of systemic signalling. Studies on A. thaliana and rice mutants with deregulated P_i acquisition (i.e. with low shoot P levels even when external supply is high) or with excessive P accumulation regardless of P status led to the identification of a systemic regulation of P uptake which, in its principal form, is present in many plants. PHO2 is a regulatory protein crucial for the down-regulation of Pht1-dependent high-affinity uptake. pho2 mutants over-accumulate P in their shoots. The PHO2 messenger RNA (mRNA) carries target sites for microRNA (miRNA) 399. This miRNA represents a phloem-mobile signal of P status in the leaves. Low P levels stimulate miRNA399 synthesis. In roots, miRNA399 then suppresses PHO2 expression, which in turn results in higher expression of Pht1 transporters and thus higher transport capacity (Fig. 7.15). The abundance of miRNA399 is under the control of a transcription factor, Phr1, which controls many P deficiency responses.

Locally controlled transcription and systemically controlled transcription do not fully explain the plasticity of nutrient uptake. The responsible transporters are also regulated post-translationally. The attribution of low- and high-affinity uptake to distinct classes of transporters is in fact not without exceptions. CHL1/NRT1.1, the first nitrate transporter identified in plants, can switch between a low-affinity state and a high-affinity state. In response to low external nitrate concentrations, a threonine residue is phosphorylated by a kinase (Fig. 7.16). Similar affinity switching is known to occur in high-affinity K⁺ uptake transporters in the HAK/KUP family. The K⁺ channel AKT1 is activated by phosphorylation when external K⁺ is low.

Ammonium is a nutrient that can be toxic to cells and therefore should not accumulate. Thus, uptake of ammonium needs to be controlled by a mechanism that allows rapid shut-off if a critical ammonium concentration is exceeded. AMT proteins are oligomeric proteins. Their C-terminal domain functions as an allosteric regulator of activity, which is controlled by phosphorylation.

Also, transporter activity can cause problems when the substrate specificity is not very high. This is illustrated by the Fe(II) uptake system IRT1. IRT1 expression is transcriptionally up-regulated when the plant’s Fe status is low. Because IRT1 mediates entry also of Zn, Ni and Cd ions, among others, Fe deficiency can lead to supraoptimal uptake of these metal ions, resulting in toxicity. In the case of IRT1, the tight control through the Fe status and rapid shutdown when enough Fe is present intracellularly is brought about by cycling of the transporter within the endomembrane system (see aquaporin cycling in Chap. 6, Sect. 6.​2 for a similar phenomenon). The residence time in the plasma membrane and protein stability respond immediately to Fe via post-translational mechanisms (Brumbarova et al. 2015), again enabling rapid inactivation of uptake capacity.

7.3.4 Intracellular Transport and Cellular Aspects of Long-Distance Transport

Nutrient supply to organelles and vacuolar storage require transporters too. Principally the same transporter types responsible for uptake also mediate transport into organelles (Miller et al. 2009). Examples are phosphate transporters in the Pht family. While Pht1 resides in the plasma membrane and takes up P_i, Pht2 is localised to the plastid inner membrane, Pht3 to the mitochondrial inner membrane and Pht4 to the Golgi compartment to mediate P_i transport into these organelles (López-Arredondo et al. 2014). The vacuole is an important storage site for some nutrient ions, including nitrate and phosphate. Vacuolar storage of nitrate is dependent on NRT2 nitrate transporters and nitrate/H⁺ antiporters of the CLC family (see also stomatal regulation in Chap. 6, Sect. 6.​3).

Not all nutrients are present in the cytosol as hydrated ions that can readily be accepted as substrates for transporters in organellar membranes. Many micronutrients, with Cu as the most extreme example, are too reactive to be available in this form for interaction with proteins and other molecules (Clemens 2001). According to the Irving–Williams series—that is, the observation that the stability of transition metal complexes with organic ligands generally follows the order Mn(II) < Fe(II) < Co(II) < Ni(II) < Cu(II) > Zn(II) (Irving and Williams 1948)—Cu(II) has a higher affinity for organic ligands than any other divalent cation in biology. This applies to N-, O- and S-ligands—for example, amino acids, organic acids or thiols such as GSH. Therefore, Cu(II) (and Cu(I) after reduction) have to be chelated by designated metallochaperones that deliver Cu ions to various pumps (proteins in the Heavy Metal ATPase (HMA) family) residing in the membranes of cellular compartments (or the plasma membrane for efflux). It has been estimated that in fact not a single hydrated Cu ion is present in the cytosol of a prokaryotic or eukaryotic cell. For Zn(II) the concentration of the so-called labile pool of ions that can readily be transported or bound is in the picomolar range—that is, several orders of magnitude below the total cellular concentration, which is in the micromolar range. Thus, most Zn(II) ions are chelated either by proteins or by low molecular weight ligands (Sect. 7.5.1). Estimates for the other micronutrients including Fe are less precise (Fig. 7.17).

Following symplastic passage from sites of uptake in the root cortex (Fig. 7.13), nutrients are loaded for long-distance transport into the xylem. The xylem is apoplastic. Thus, efflux across the plasma membrane of the xylem parenchyma or pericycle cells is necessary. Energetically the situation is reversed in comparison with uptake. Anions can be loaded via facilitated diffusion; cations may require active transport. Many transporters involved in this loading process are known in the model systems A. thaliana and rice. Loss of function leads to reduced xylem sap concentrations of the nutrient in question. This results in lower shoot concentrations.

In A. thaliana, nitrate, which can (in contrast to ammonium) be transported over long distances, is loaded into the xylem by the low-affinity transporter NRT1.5. The cytosolic nitrate concentration is high enough for this type of transport. NRT1.5 shows the typical localisation of expression in pericycle cells adjacent to the protoxylem to fulfil its loading function (Fig. 7.13). Similarly, borate is exported into the xylem by transporters known from A. thaliana and rice (BOR1). K⁺ ions reach the xylem through outward-rectifying Shaker-type channels—SKOR in A. thaliana. Its opening probability decreases with rising external K⁺ concentrations. This ensures that only efflux is mediated (Chérel et al. 2014). Zn is an example of a nutrient that is actively pumped into the xylem. The proteins responsible are HMAs. In A. thaliana, HMA2 and HMA4 load Zn into the xylem (Sect. 7.5.2 for the role of HMAs in metal hyperaccumulation).

7.3.5 Plasticity of Root Architecture and Responses to Nutrient Deficiency

New cells arise through cell division in the root meristem near the root tip. The root tip itself is protected by a root cap. In many species the root cap releases cells (border cells) that influence the rhizosphere—for example, by enhancing nutrient availability. Cells formed in the meristem differentiate into the epidermis, the cortex, the endodermis and the stele with its pericycle, xylem and phloem. Lateral roots are initiated in the pericycle; adventitious roots are initiated in the stem. Root hairs, which are particularly important for increasing the root surface, arise through the tip growth of designated epidermal cells, the trichoblasts.

Root architecture plasticity is governed by a variety of processes that occur in response to either the nutrient availability in the soil or the nutritional status of the plant (Fig. 7.18). Responses to nutrient availability are predominantly those triggered by the presence of nutrients. Responses to the nutritional status are triggered by nutrient deficiency. Detailed investigations have revealed clear nutrient specificity of the root architecture modulation. For instance, the strong inhibition of primary root growth and the stimulation of root hair development and elongation in phosphate-deficient plants are very well documented. Conversely, phosphate availability stimulates growth of lateral roots. Phosphate resources tend to be concentrated in the topsoil. The root architecture responses enhance exploitation of this soil layer. Perhaps the most striking response to phosphate deficiency is the formation of cluster roots—for instance, in a range of plants native to nutrient-poor soils in Australia (Lambers et al. 2015) (Fig. 7.19).

Stimulation of lateral root growth by a lack of adequate nitrogen or Ca supply represents a foraging behaviour of roots. This strategy is physiologically limited inasmuch as a minimum nutrient supply has to be available to sustain the foraging (Giehl et al. 2014). When a nutrient-rich patch of soil is found, the growth of lateral roots is further stimulated. This has been shown for nitrate, phosphate and Fe (Fig. 7.18). In many experiments with various plant species, localised supply of nitrate in the substrate triggered growth of lateral roots located near or in the nutrient patch. In contrast, the systemic response of the lateral roots as a whole to conditions of ample nitrogen supply is a reduction in the growth rate. Less investment in root biomass is needed, thus more resources can be channelled into shoot growth. The local response enables rapid exploitation of a nutrient resource and thereby gives a plant a distinct competitive advantage over its neighbours. It is dependent on the perception of nitrate as a signal and not a mere consequence of nutrient supply. Generally, phosphate and nitrogen supply elicit the strongest responses of the root system development (Osmont et al. 2007). For many root architecture responses, however, it remains to be determined whether they represent strategies to mitigate the deficiency or symptoms of either the nutrient deficiency stress or the availability of nutrients to one part of the root but not another.

Like all plant developmental processes, the formation of the root system is controlled by the delicate, fine-tuned interaction of hormones (for detailed descriptions, see plant physiology and plant molecular biology textbooks). Key events are cell cycle activity, cell differentiation and cell elongation. The interplay between auxin and cytokinin in controlling root architecture is well documented. One important aspect is the balance between cell division and cell differentiation in the root meristem (Petricka et al. 2012). An auxin gradient establishes in the root and controls stem cell maintenance, cell proliferation and differentiation. Cytokinin, on the other hand, antagonistically inhibits cell division and reduces the size of the root meristem.

Emergence of lateral roots originates from pericycle cells adjacent to the xylem poles. Auxin triggers the formation of lateral roots. The sensing of phosphate deficiency, for instance, may sensitise pericycle cells to auxin, possibly through the up-regulation of auxin receptor expression. Overall, the exact mechanisms underlying the nutrient status–dependent modulation of root architecture by the hormonal network are not well understood yet.

7.3.6 Sensing of Nutrient Availability and Nutrient Status

Regulation of nutrient acquisition and modulation of root system architecture require accurate sensing of external nutrient availability and internal nutritional status. The latter is monitored both locally and systemically (Chap. 2, Sect. 2.​2.​4). The respective sensing then has to be translated into transcriptional changes (for instance, the up-regulation of genes encoding high-affinity transporters), post-translational control of transporters or regulatory proteins, and changes in the concentrations and distribution of hormones.

As alluded to in previous chapters, the molecular understanding of the primary sensing of environmental parameters generally lags behind the insight into the downstream signal transduction events. This is no different for the sensing of nutrient availability in the rhizosphere or the monitoring of the nutritional status. Local and systemic sensing of phosphate status, for instance, is not understood, while the regulation via miRNAs is well established (Fig. 7.15). The knowledge on nitrate perception is most advanced and can therefore serve as an example. The nitrate transporter NRT1/CHL1, already introduced as a protein able to switch affinity in a phosphorylation-dependent manner (Fig. 7.16), functions, in addition, as a nitrate sensor (Ho et al. 2009). Both the primary response to nitrate availability (the up-regulation of nitrate transporter, nitrate reductase and nitrite reductase genes) and the stimulation of lateral root growth in nitrate-rich soil patches (Fig. 7.18) are dependent on NRT1/CHL1 in A. thaliana. This association is not explained by transporter activity of NRT1/CHL1. Instead, external nitrate regulates the phosphorylation status of the protein and its ability to activate responses to high external nitrate availability. NRT/CHL1 is the first plant example of a so-called transceptor, a transporter protein that in fact functions like a receptor. The nutrient is not a substrate but a ligand with a signalling function. Many similar proteins may be involved in other nutrient-sensing processes.

7.4.1 Mycorrhizae

About 90% of all land plants—most monocots and dicots, and nearly 100% of gymnosperms—engage in a mycorrhizal association with fungi. On the fungal side, tens of thousands of different taxa are involved. Mycorrhizae are present in most terrestrial habitats colonised by plants. Rare exceptions include very dry soils, waterlogged soils or extremely phosphorus-impoverished areas where plants with cluster roots dominate (Fig. 7.19). Also, most members of a few plant families (e.g. Brassicaceae, Chenopodiaceae) do not form mycorrhizae. Mycorrhizae are evolutionary ancient and most likely were already instrumental when plants first colonised land about 450 million years ago (van der Heijden et al. 2015). This is suggested by fossils of the earliest land plants, which show signs of mycorrhizae.

Two major types are distinguished, which vary in morphology, type of partner and physiology. An endomycorrhiza is characterised by growth of fungal hyphae within the root cortex of the plant host and the formation of structures within cortex cells, while in an ectomycorrhiza the fungal hyphae do not grow into root cortical cells. Ectomycorrhizal fungi sheath the root with a mantle of tightly woven hyphae and grow between cortical cells (Hartig’s net) (Fig. 7.20). Within the endomycorrhizal symbioses, three types are differentiated further: the arbuscular mycorrhiza, ericoid mycorrhiza and orchid mycorrhiza. By far the most common form (occurring in around 75% of all plant species) and evolutionarily the oldest is the arbuscular mycorrhiza. It is also the one that has been studied intensively at the molecular level and will thus be the focus of this section.

A large fraction (up to 80%) of the main limiting macronutrient, phosphorus, is provided to plants by mycorrhizal fungi. Phosphorus is much less mobile in the soil solution than nitrogen or other macronutrients. Therefore, the ability of fungal hyphae to grow into soil micropores that are too small for the diameter of plant roots makes more phosphorus available for uptake. Like plant roots (Fig. 7.5), mycelia actively mobilise phosphorus bound to soil particles—for instance, by the secretion of phosphatases. Zn is another comparatively immobile nutrient that is supplied to the host by the fungal partner. However, this is not well understood. The contribution of the arbuscular mycorrhiza to nitrogen supply for the host is much less pronounced. This is different for ericoid mycorrhizae and ectomycorrhizae. In addition to nutrients, the fungal symbiont supplies water to the host.

Both the molecular physiology of mycorrhizal transport processes and the events that enable symbiosis formation are being investigated predominantly in the model systems Medicago truncatula and Lotus japonicus. These species are studied in place of A. thaliana which, as a member of the Brassicaceae, does not engage in mycorrhizal symbioses. The added benefit of M. truncatula and L. japonicus as model species is that they are legumes; therefore, biological N₂ fixation can be studied as well.

Fungal arbuscules are structures with a large surface within root cortical cells (Fig. 7.21). Thus, the exchange of nutrients and sugars between the fungus and the plant occurs through two membranes that are in close vicinity: the invaginated plant-derived periarbuscular membrane, which is continuous with the plasma membrane of the host cell; and the plasma membrane of the arbuscule. Nutrient transfer from the soil into host cells requires uptake by the fungal mycelium, efflux into the symbiotic interface (the periarbuscular space) and uptake by the plant cell. Transfer of sugars arriving via the phloem from the plant to the fungal partner entails two membrane passages: efflux across the root cell plasma membrane and uptake by the arbuscular membrane.

Experimentally it is extremely challenging to detect and identify the transporters mediating these different steps. Consequently, a molecular dissection of the exchanges between the symbiotic partners is far from achieved. Sequencing of the genomes of plants and mycorrhizal fungi, as well as transcriptomes of mycorrhized and non-mycorrhized plants, is establishing inventories of transporters that are potentially involved (Garcia et al. 2016). The understanding of phosphate transport is most advanced. In host plants, genes encoding specific subgroups of phosphate transporters (the Pht1 family; Sect. 7.3.4) are up-regulated upon fungal colonisation of the root. In turn, other transporters involved in direct phosphate uptake from the soil solution are down-regulated. This indicates that phosphate supply through the symbiosis is prioritised. How this regulation occurs molecularly is not known. The up-regulated phosphate transporters in colonised root cells take up phosphate from the interface. Energy is provided, as in the case of anion uptake from the soil solution, by the pumping of protons. The proteins involved in the release of phosphate from the arbuscules are unidentified. Uptake of phosphate from the soil solution into the mycelium is mediated by high-affinity transporters similar to the ones expressed by plants (Harrison and van Buuren 1995). An equally fragmentary aspect is the assignment of essential sugar transport steps—that is, efflux of glucose out of the colonised root cell and uptake into the fungal mycelium—to particular proteins.

Even less is known at the molecular level about the possible exchange of reduced carbon, macronutrients and micronutrients between neighbouring plants via mycorrhizal connections. The host specificity—especially of the fungi involved in arbuscular mycorrhizae, the Glomeromycota—is low. Thus, they can be simultaneously associated with individuals of several different plant species. Also, many plant roots are colonised by more than one fungal species. The resulting underground hyphal network may well enable, for example, the support of shaded tree saplings with assimilates supplied by connected adult trees (van der Heijden et al. 2015). Also, the first direct evidence of carbon trading between trees has recently been found through carbon isotope labelling. A substantial fraction of the carbon isotope label introduced via photosynthesis into one tree was later found in neighbouring trees, suggesting an exchange mediated by ectomycorrhizal connections (Klein et al. 2016). The exact fluxes in such underground networks, however, have not really been determined yet, and the relevant molecular knowledge (e.g. about responsible transporters) is not yet available.

7.4.2 Nitrogen Fixation

Biological nitrogen fixation represents the major route of nitrogen supply to plants in natural terrestrial environments (an estimated 80–90%), while input into agricultural ecosystems originates largely from industrial production of NH₃ through the Haber–Bosch process. The conversion of inert atmospheric N₂ into organic nitrogen is of utmost importance for the global N cycle. The enrichment of soils with nitrogen has a massive impact on ecosystems (Chap. 23, Sect. 23.​2).

Molecularly the Fabaceae–rhizobia symbiosis is the best-understood symbiosis and will therefore be the focus here.

The symbiotic bacteria live in specialised structures (nodules in the case of the Fabaceae–rhizobia symbiosis) within the plant host. These provide suitable conditions for the supply of reduced carbon to the bacteria and the transfer of fixed nitrogen to the host. Another essential requirement is protection of the feature enzyme of all N₂-fixing bacteria—nitrogenase—from oxygen. Nitrogenase catalyses the reaction:

(7.3)

It is immediately clear how energy demanding the nitrogenase reaction is. This probably explains why symbiotic N₂ fixation has such a dominant role in nature. For free-living bacteria it is extremely difficult to provide the ATP needed for NH₃ production. In contrast, symbiotic bacteria are supplied with assimilates by the host. The oxidation through respiration produces enough ATP to ensure efficient nitrogen fixation. A major challenge is the co-occurrence of respiration and oxygen-sensitive nitrogenase reaction. The mechanisms involved in generating a low-oxygen atmosphere in the vicinity of the nitrogenase are a bacterial cytochrome oxidase with an unusually high O₂ affinity and the synthesis of O₂-binding leghemoglobins by the host cells.

Within root nodule cells, rhizobia live as bacteroids in a specialised structure, the symbiosome (Fig. 7.22). The symbiosome is separated from the cytosol of the nodule cell by a membrane. Carbohydrates arriving at nodules via the phloem are converted into organic acids. One fraction of these is taken up by the bacteroids, where they are oxidised in the citric acid cycle (alternatively called the TCA cycle). ATP and reducing equivalents are then used for the nitrogenase reaction. The resulting NH₄ ⁺ is assimilated into glutamine in the cytosol of the nodule cell. The other fraction of the organic acids provides the carbon skeletons for the nitrogen export compounds. Depending on the host plant, either glutamine itself can be exported or it can serve as an amide donor for other metabolites (asparagine, allantoin) supplying organic nitrogen to tissues outside the nodules.

7.4.3 The Common Sym Pathway

Establishment of both a mycorrhiza and nitrogen-fixing symbiosis has to occur for every single plant individually. It requires mutual recognition by the partners through chemical communication. Plant-derived signals perceived by the microorganisms stimulate processes that facilitate formation of the symbiosis. Conversely, signals released by the microbial partner trigger far-reaching developmental changes in the host plant that make entry of the fungal hyphae or the bacteria possible. The command over the process of symbiosis establishment lies entirely on the plant side, as the host has to actively allow infection by the symbiotic partner (Oldroyd 2013). Plants growing in an environment with an ample supply of phosphate do not engage in mycorrhizae, and legumes provided with enough nitrate and/or ammonia do not form nodules. This illustrates why the establishment of these symbioses can be regarded as an acclimative response to nutrient scarcity.

Flavonoids released by plant roots stimulate rhizobia to synthesise nod factors. These are lipo-chito-oligosaccharides, consisting of a chitin oligomer backbone with several modifications. Nod factors are perceived by plant roots and trigger many changes in the host that enable colonisation by the bacteria. First, root hairs curl and thereby build a protected environment for the rhizobia. Following a few bacterial cell division cycles, an infection thread forms inside the root hair. Rhizobia enter the plant root through these threads. Concomitantly, nodule organogenesis occurs in the root cortex. The cell cycle is stimulated, and newly dividing cells develop into the nodule (Fig. 7.23).

The plant signals that stimulate spore germination and hyphal branching of endomycorrhizal fungi have been identified as strigolactones (Akiyama et al. 2005)—that is, root-synthesised phytohormones involved in controlling shoot branching. Myc factors are the signals on the fungal side. They are chemically similar to nod factors in that they are lipo-chito-oligosaccharides as well (Maillet et al. 2011). Upon recognition of myc factors, plant roots undergo massive changes to allow penetration of hyphae towards the root cortex. A pre-penetration apparatus is formed in rhizodermal cells. It represents a transcellular “tunnel”, which is then used by the hyphae to grow into the root cortex (Parniske 2008). In the cortex the hyphae grow intercellularly before they form arbuscules within cells in the inner cortex (Fig. 7.23).

In spite of the fundamental differences between the symbiotic microorganisms and the plant developmental changes allowing colonisation of the host, the recognition and early signal transduction events in host cells proceed through a pathway shared by arbuscular mycorrhizae and the Fabaceae–rhizobia symbiosis: the common sym pathway. Molecular analysis of symbiotic interactions in the legume model systems has revealed that the symbiosis pathway enabling the formation of mycorrhizal associations in most plant species was, much later in evolution (around 150 million years ago), recruited by legumes for the establishment of biological nitrogen fixation. Research using mutants of M. truncatula and L. japonicus defined several genes that are essential for both mycorrhizae and root nodule symbiosis. A lack of any of these genes (e.g. the SYMRK gene) abolishes the early responses of plant roots to the microbial signals (Stracke et al. 2002). The chemically very similar nod and myc factors are perceived by closely related receptor-like kinases residing in the plasma membrane of root cells (Fig. 7.24). The receptors and co-receptors, whose exact functions with regard to the recognition of specific ligands are not resolved yet, share LysM domains typical for proteins binding chitin fragments. Early signal transduction event in root cells stimulated by symbiosis signals are Ca oscillations (i.e. rapid fluctuations in Ca concentrations, Chap. 2, Sect. 2.​2) in the nucleus. Components that have been identified are ion channels of the nuclear envelope (CASTOR and POLLUX) required for the induction of the calcium oscillations, and components of the nuclear pore (Oldroyd 2013). Perception of the Ca signal is mediated by Ca- and calmodulin-dependent kinases (CCaMK), which interact with another common sym pathway component of unknown molecular function, CYCLOPS (Fig. 7.24). Even among the transcription factors downstream from the Ca signalling, overlaps in functional roles between mycorrhizal establishment and nodulation exist.

The common sym pathway exists in mosses and lycophytes, illustrating once more how ancient mycorrhizal associations are in terrestrial plants. Apparently, rather subtle changes in the genomes of legumes enabled them to establish a second symbiosis much later in evolution, this time with rhizobia, by principally employing the same basic mechanisms. It has been shown that many nodulation genes can be replaced by orthologous genes from plant species that are not able to form nodules, indicating that no fundamentally new innovations were required for the legume–rhizobia symbiosis. Nonetheless, a legume root initiates nodule formation in response to nod factors and prepares for penetration of fungal hyphae in response to myc factors. The specificity of the recognition and signalling processes is presumably conferred by specific functions of particular receptor-like kinases perceiving the signals and by specific transcriptional regulators.

7.5.1 Essential Metal Toxicity and Tolerance

The essential element boron has a particular narrow optimal range (Fig. 7.2a). Like boron deficiency, boron toxicity limits agricultural productivity in many regions around the world. B toxicity occurs frequently in arid environments on soils rich in B, such as alkaline soils of marine origin. It was first described in southern Australia. Typical symptoms are leaf burns—necrotic spots at the tips of older leaves where higher concentrations of B accumulate because of the transpiration stream that moves mineral nutrients with the water (Nable et al. 1997). Tolerance is conferred by transporters mediating the efflux of boron into the root apoplast (Fig. 7.26). This limits the translocation of boron to the shoot. At high external boron concentrations the efflux transporters can lower cytosolic concentrations (Miwa et al. 2007).

There is a documented adaptation to boron toxicity within several crop species, including wheat and barley. Wheat cultivars bred for southern Australia have a yield advantage in their region of origin and perform less well than cultivars from northern Australia when grown there. Genetic loci controlling boron tolerance are known from wheat and barley. Intraspecific differences have been traced to variations in boron efflux transporters. In a highly boron-tolerant barley cultivar this trait is associated with much stronger root expression (>100-fold higher mRNA abundance) of a boron efflux transporter, which is at least partly attributable to an increase in the gene copy number (Sutton et al. 2007).

After Fe, Mn is the second most abundant transition metal in the Earth’s crust. Mn availability is positively correlated with proton concentrations in the soil solution; thus, in acidic soils Mn can become toxic to plants. A second factor causing increasing Mn availability up to potentially growth-inhibitory concentrations is the reducing condition of waterlogged soils, which promotes the more bioavailable Mn(II) over the much less available Mn(IV) (Foy et al. 1978). Consequently, plant species adapted to growth under such conditions can tolerate much higher Mn content in their tissues. Rice, for example, can cope with a Mn content of 5000 μg/g d.w. in leaves, while many other plants show toxicity symptoms when the Mn content in the leaves exceeds 150 μg/g d.w. (Marschner 2012).

The dominant metal tolerance mechanism is compartmentation. An excess of metal is removed from the cytosol to prevent deleterious effects. In the case of Mn, transporters in the MTP family (metal tolerance proteins; the original name is CDF, for cation diffusion facilitator) mediate sequestration of Mn in the vacuole as the major storage site or in the endoplasmic reticulum (ER) and Golgi (Fig. 7.26). The transporters are present in plants with basal tolerance, as well as those with specific adaptations. Rice is highly Mn tolerant and MTP proteins are involved in Mn tolerance. The same applies to Stylosanthes species (e.g. S. hamata)—tropical legumes thriving in acid soils. However, to date it is not clear whether differences in expression or specific properties of these transporters in the Mn-tolerant species explain adaptation to high-manganese soils. Alternative compartmentations away from the cytosol are accumulation in the apoplast and exclusion from plant tissues. A possible mechanism, similar to the mode of Al tolerance (Sect. 7.5.4), is the secretion of organic acids that form 1:1 complexes with Mn.

7.5.2 Metal Hyperaccumulators as Models for Adaptation to Extreme Environments

Other essential micronutrients such as Zn or Ni can reach toxic concentrations in some soils too. While toxic Zn concentrations are largely restricted to mining-impacted soils, potential Ni toxicity of serpentine soils is more common. A number of plant species have evolved the ability to colonise metal-rich sites. They are called metallophytes.

As described in this chapter, the concentrations of metals in plant tissues are under tight physiological control. Many elements are toxic when present in excess. However, approximately 500 taxa (i.e. about 0.2% of all flowering plant species) have evolved the ability to hyperaccumulate metals or metalloids (arsenic, selenium) in their leaves. Hyperaccumulation is defined as a metal concentration that is above an element-specific threshold in above-ground tissues of plants grown in the field. This threshold should be a concentration that is two to three orders of magnitude higher than what is normally found in plants growing in soils that are not enriched with particular metals, and one to two orders of magnitude higher than what non-hyperaccumulating species show at a site where the hyperaccumulator grows (Krämer 2010). For example, the hyperaccumulation thresholds are 100 parts per million (ppm) for Cd, 1000 ppm for Ni and As, and 3000 ppm for Zn, while plants normally contain, for example, about 50–100 ppm Zn in their organs. Important aspects of this definition are (a) that extreme accumulation is found in plants grown in the natural habitat (not only in plants cultivated under laboratory conditions) and (b) that the accumulation in above-ground tissues is due to active translocation via the roots and not caused by aerial deposition onto the leaves.

Hyperaccumulators represent a subgroup of the metallophytes—that is, plants that are part of a special type of vegetation found on metal-rich sites. Such sites can be rich in metals either naturally (geogenic) or because of human activities such as metal mining or processing (anthropogenic). Typical geogenically metal-rich habitats are serpentine (“ultramafic”) soils with high concentrations of Ni; calamine soils rich in Zn; soils in the African copper belts characterised by high concentrations of Co, Cu, Cr, Ni and Zn; and seleniferous soils enriched in Se (Baker 1989). An example of typical metallophyte vegetation is the “Galmei flora” found in calamine soils in Belgium and the region around Aachen in Germany. A characteristic species is the endemic Viola calaminaria.

All metallophytes have evolved mechanisms to survive and reproduce on metal-rich soil. They share the ability to tolerate metal or metalloid concentrations that would strongly inhibit or even kill most other plants. With respect to the accumulation of metals, three strategies are distinguished. Many metallophytes are able to grow in the presence of metal/metalloid concentrations that are intolerable for most plants, because they can efficiently exclude the metal from their cells. Even in the presence of high concentrations of bioavailable metal in the soil, the concentration within the plant is maintained at a very low level (Fig. 7.27). Examples are the aforementioned V. calaminaria and other more common metallophytes such as Armeria maritima and Silene vulgaris. Some metallophytes are bioindicators. They can tolerate a wider range of concentrations than normal plants, and the accumulation in leaves is linearly correlated with concentrations of available metals in the soil. The third strategy is hyperaccumulation. The plants very efficiently translocate metals from the soil solution via root uptake and long-distance transport into the shoots. Some hyperaccumulator species even hyperaccumulate when growing on sites not enriched in metals at all. They are pseudo-metallophytes, meaning that they occur not only at metal-rich sites but also at sites with normal metal availability in the soil. Examples of such species are the model hyperaccumulators Arabidopsis halleri and Noccaea caerulescens (in the Brassicaceae; see below).

A given metallophyte species does not tolerate toxic concentrations of any metal. Instead, naturally evolved hypertolerance—that is, a degree of metal tolerance exceeding the tolerance found in most plant species—is specific to certain metals. Plants adapted to serpentine soils can thrive in the presence of high Ni concentrations but not necessarily in the presence of high Zn concentrations. There is pronounced variation even within species. Accessions of N. caerulescens adapted to calamine soils tolerate Zn and Cd very well but are sensitive to Ni. Conversely, accessions from ultramafic sites are extremely Ni tolerant yet are as sensitive to Zn and Cd as non-metallophytes (Halimaa et al. 2014).

The large majority of hyperaccumulator species—that is, about 450 out of 500—hyperaccumulate nickel and typically occur in serpentine (ultramafic) soil. Hotspots for Ni hyperaccumulators are Cuba and New Caledonia. They are rich in serpentine sites, and more than 150 Ni-hyperaccumulating species grow on these islands. Around 15 taxa hyperaccumulate either Zn, As or Se. A few species accumulate Cd (Krämer 2010). Hyperaccumulating species are strongly overrepresented among the Brassicaceae, indicating a propensity of species in this family to evolve hyperaccumulation. The adaptation of metallophytes to metal-rich soils has attracted attention from evolutionary biologists and ecologists because the toxicity of metals exerts extreme selective pressure. On serpentine soil a large number of endemic species are found, indicating the need for specific adaptations to cope with such edaphic conditions (which include a high Mg to Ca ratio and scarcity of N, P and K, in addition to high metal concentrations). Colonisation of sites that have been metal contaminated by human activities represents an example of rapid evolution—of “evolution in action”, as stated by Antonovics et al. (1971). Within a range of a few metres the environment can differ dramatically, allowing many plants to grow in a meadow and allowing only very few adapted plants to establish themselves on a neighbouring heap of mining waste.

Since the 1990s, research into the ecophysiology and genetics of metallophytes has, in addition, been fuelled by several applications envisioned for metal-hypertolerant plants. They can be used for phytoremediation of metal-contaminated areas—for example, mining sites—to enable gradual revegetation. Metallophytes can facilitate the growth of other plants by reducing the bioavailability of toxic pollutants (phytostabilisation). Metal-hyperaccumulating plants could even be used for phytoextraction to remove metal contaminants (Salt et al. 1998). A related application is phytomining to exploit substrates that are too poor for conventional mining.

Mechanistic insights into the evolution of hyperaccumulation can illustrate the path leading to adaptation to extreme environments or, more generally speaking, the emergence of new traits in nature. Metal hyperaccumulators show exceptional mobility of metals and metalloids, which enables the transfer to above-ground tissues and thereby a fundamentally different partitioning of metals between below-ground and above-ground tissues. The relatedness of hyperaccumulator species such as N. caerulescens and A. halleri to the molecular genetics model A. thaliana has enabled molecular approaches to the unravelling of metal hyperaccumulation. Intra- and interspecific crosses—for example, between contrasting N. caerulescens accessions or between A. halleri and its non-hyperaccumulating relative Arabidopsis lyrata—have revealed the genetic architecture of metal hypertolerance and metal hyperaccumulation. The two traits are partially independent. Some of the mapped quantitative trait loci are important for both characteristics, others only for one of the two. Several studies compared the transcriptomes of hyperaccumulators with that of A. thaliana (Becher et al. 2004; Weber et al. 2004). The fundamental realisation was that many genes involved in metal homeostasis in normal plants show altered regulation in hyperaccumulators. Genes encoding metal transporters and enzymes catalysing the synthesis of metal ligands are constitutively more strongly expressed (Fig. 7.28). The best-documented example is the metal-pumping ATPase HMA4, a protein that effluxes Zn ions across the plasma membrane (Sect. 7.3.4). In A. thaliana, HMA4 loads Zn into the xylem for translocation to the leaves. Much stronger expression of HMA4 in A. halleri results in very efficient root-to-shoot translocation of both Zn and Cd. In addition the efflux activity protects root cells from Zn and Cd toxicity (Hanikenne et al. 2008). This difference explains a substantial fraction of the Cd/Zn hypertolerance and hyperaccumulation of A. halleri. Higher expression evolved through an expansion of the gene copy number and variation in promoter sequences. The proteins apparently do not differ in affinities, activities or any other functional features. Thus, an important conclusion of general importance is that adaptation has evolved through transcriptional changes via copy number variation and changes in cis-regulatory sequences—that is, rather subtle changes that can easily occur and then be selected. Indeed, for the triplicated HMA4 locus in the A. halleri genome, strong signs of a selective sweep were detected (Hanikenne et al. 2013). Among individuals from different populations, nucleotide diversity was strongly reduced around the HMA4 locus. This is indicative of strong selection causing rapid spread of this particular allele in A. halleri populations.

7.5.3 Sodium Toxicity

The soils of more than 6% of the world’s terrestrial surface contain high concentrations of salt, mostly NaCl. To a lesser extent, Na-carbonate and Ca-sulphate are also found, especially in the vicinity of salt lakes. Needless to say, two thirds of the Earth’s surface—namely, the oceans—represent saline habitats. Halogenides of alkali and earth alkaline ions are easily soluble in water and thus these ions are washed out from suboceanic or terrestrial minerals of the Earth’s mantle and finally accumulate in the sea. From there, saline aerosols are transported landwards by the wind, leading to continuous salt deposition not only in the coastal regions but also further inland. In arid and semi-arid areas, upward movement of the soil water results in deposition of dissolved salt upon evaporation of the solvent, frequently giving rise to salt crusts. This process also takes place in irrigated arid areas. It is estimated that by the middle of the current century, increased salinisation will result in up to 50% arable land loss (Wang et al. 2003). The impact of soil salinity on agriculture is enormous, as it affects plants during their entire life cycle and results in huge losses in biomass production and yields.

Soils are considered saline when the electrical conductance exceeds 0.4 Siemens per metre. This value corresponds to approximately 40 mM NaCl and an osmotic potential of −0.2 MPa. The threshold value is derived from agriculture. Many crops react with yield reduction when grown in soils of higher salinity. Salt tolerance is usually determined as the percentage biomass production or crop yield in saline versus control conditions over an extended period of time, or in terms of survival, which is especially useful in experiments with seedlings.

The conductivity of seawater (3% salt: 480 mM Na⁺, 50 mM Mg²⁺ and 560 mM Cl⁻) is 4.4 S/m—more than ten times the threshold for soil salinity—with an osmotic potential of −2.7 MPa. The conductivity of water used for irrigation must be less than 0.2 S/m, notwithstanding the fact that some plants—for example, glassworts (Salicornia species) or even special cultivars of barley (cv. California Mariout)—require or at least tolerate irrigation with seawater. For as yet unknown reasons, salinity in soils is often accompanied by toxic concentrations of boric acid (Tester and Davenport 2003) (Sect. 7.5.1).

Na⁺ leaks into plant cells via Ca²⁺-permeable non-selective cation channels (NSCCs) or K⁺/Na⁺ transporters. The molecular identities of these proteins are still uncertain. Many K⁺ transport systems such as HAK/KUP family members (Sect. 7.3.2.2) have some affinity for Na⁺. High-affinity Na⁺ uptake has also been observed but is unlikely to play a role under conditions of salt stress with high external Na⁺ concentrations (Munns and Tester 2008). The Na⁺ electrochemical potential gradient across the plasma membrane suggests that facilitated diffusion is the principal mode of Na⁺ influx, while Cl⁻ is transported against the electrochemical potential. Uptake therefore requires proton coupling, while efflux can be passive. Anion influx can be passive too, provided that a permeable anion channel is present and the concentration gradient of the ion across the membrane is high enough, as can be the case in saline environments. Furthermore, depolarisation will result from the uptake of cations such as Na⁺. This lowers the electrochemical potential gradient and facilitates Cl⁻ uptake. Thus, exposure of a cell or tissue to high salt concentrations results primarily in passive influx of Cl⁻, followed by active uptake after the membrane potential has returned to more negative values.

Salt stresses plants in several ways (Fig. 7.29): by dehydration (Chap. 6), Na⁺ (and Cl⁻) toxicity, nutrient imbalances and reactive oxygen species production (ROS production). Saline soil solutions have a very negative osmotic potential, which has to be overcome by the water-absorbing surface of the plant, typically the root. Shortage of water affects growth of cells and plant organs. Thus, for water uptake, the root cells have to produce and maintain an even more negative osmotic potential. Na⁺ in particular, but also Cl⁻, can be regarded as biologically aggressive solutes on account of their small ionic diameters and the corresponding high surface charge. These properties result in high attraction for water molecules and thus the binding of much water in water shells, which enhances intracellular water scarcity. Also, accumulation of small cations such as Na⁺ can strongly interfere with the intracellular balance of ion pools and charges, affecting the membrane potential, which may result in inactivation of voltage-dependent membrane functions. This is especially the case when K⁺ is displaced by Na⁺ with its higher charge density. Other inorganic plant nutrient relations—for example, of Ca²⁺, Mg²⁺ and anions such as nitrate and malate—may be affected as well.

7.5.3.1 The Osmotic Component of Salt Stress

When a plant, be it salt tolerant or sensitive, is subjected to a sudden increase in the NaCl concentration in the medium or soil, a fast and a slow reaction can be differentiated (Fig. 7.30 and 7.31). Leaves of various cereals and dicots show an instantaneous halt of expansion. Because of its rapidity and the subsequent partial recovery, this fast response must be due to changes in cell water relations. Since the same phenomena can be triggered with KCl, mannitol or polyethylene glycol instead of NaCl, this response is not salt specific (Munns 2002). For most plants, the threshold for that response is around 40 mM NaCl, but it may be lower for particularly salt-sensitive species. Several minutes after the initial decline in leaf growth, a partial recovery takes place and a new steady growth rate is attained after approximately 30 min. The time required for recovery depends on the concentration of the salt solution. Removal of the salt results in an equally sudden overshoot followed by a fast return to the original growth rate.

Under long-term salinity stress, inhibition of shoot growth encompasses reduced leaf expansion, delayed formation of new leaves and delayed or even suspended bud break, resulting in fewer branches or lateral shoots. Contrary to the expectation, root growth in saline media is less affected than shoot growth (Chap. 6). This might be due to the fact that a reduction in shoot growth also reduces the water consumption by the plant and mitigates the transpiration-induced increase in salinity in the rooted soil volume.

7.5.3.2 The Ionic Component of Salt Stress

Homeostatic ion concentrations in the cytosol of a non-stressed cell of a glycophyte—that is, a plant without adaptation to saline habitats—are 100–200 mM K⁺, 1–10 mM Na⁺ and Cl⁻, and 0.1–0.2 mM Ca²⁺. These concentrations are maintained in the plant cell mainly through ion transport energised by H⁺-ATPases and H⁺-pyrophosphatases in the plasma membrane and the tonoplast, respectively. In addition, channels and pumps are involved (Sect. 7.3.2). The combined activities result in a membrane potential of around −150 mV at the plasma membrane and a tonoplast potential between +20 and +50 mV (the cytosolic side is always more negative). These membrane potentials are associated with differences in pH between the cytosol (pH 7.5) on the one hand and the apoplast and the vacuole on the other, whose pH values are around 2 units more acidic (Fig. 7.32).

High salinity subjects this homeostasis to considerable strain (Fig. 7.31). In the majority of plant species grown under salinity, Na⁺ reaches a toxic concentration before Cl⁻ does, and so plant chloride relations have attracted much less attention than the plant’s response to sodium (Munns and Tester 2008). Chloride is an essential micronutrient (e.g. for photosynthesis), acts as counter anion to stabilise membrane potentials and is involved in turgor and pH regulation. In the presence of high external salt concentrations, large intracellular Na⁺ pools build up, partly at the expense of the K⁺ pools. More than 50 enzymes are controlled or activated by K⁺, and Na⁺ cannot substitute for K⁺ in this role. High concentrations of Na⁺ or high Na⁺ to K⁺ ratios therefore disturb or even disrupt various enzymatic processes. Photosynthesis and respiration are among the processes that are most sensitive to salt stress. Protein synthesis requires high concentrations of K⁺—for example, for the binding of transfer RNA (tRNA) to the ribosomes—and is thus also highly affected by high intracellular Na⁺. Finally, ion imbalances (e.g. in photosynthesis and respiration) and water shortage result in oxidative stress which, in combination with the aforementioned impairments, can easily result in cell and organ death.

7.5.3.3 Variations in Salt Tolerance: Plant Functional Types with Respect to Salinity

Plants vary greatly in their tolerance of salinity (Fig. 7.33). For example, after some time in a 200 mM NaCl solution, a salt-tolerant species such as sugar beet may have a reduction of only 20% in dry weight, a moderately tolerant species such as cotton might have a 60% reduction and a sensitive species such as soybean might be dead. A plant from a salt marsh (e.g. Suaeda maritima), however, may be growing at its optimum rate (Flowers and Colmer 2008). Salt-sensitive species such as the monocot rice and the dicot A. thaliana are termed glycophytes, whereas species with a high salt tolerance or even salt requirement for growth are referred to as halophytes. While glycophytes are clearly more salt sensitive than halophytes, they still possess a basal tolerance mechanism, and most molecular knowledge on salt tolerance mechanisms originates from work with A. thaliana.

Not all developmental stages of a plant are equally sensitive to salinity. With respect to crops, this provides an opportunity to minimise salt injury at the sensitive stages by using irrigation water of differing salinity during the season. Notwithstanding their relatively high salt tolerance, sugar beet, barley and cotton are relatively sensitive during germination or early seedling growth. In contrast, corn, pea and beans are more sensitive during later stages of development. In tomato, salt tolerance is low in young plants, becomes much higher during vegetative growth and decreases again during flowering.

7.5.3.4 Salt Tolerance Mechanisms

At the cellular level, plants have several principal mechanisms to cope with salinity (Fig. 7.29). They can:

• Minimise initial entry into the root

• Maximise efflux from the root into the soil

• Minimise loading into the xylem or maximise retrieval from the xylem fluid before Na⁺ reaches the shoot

• Maximise recirculation out of the shoot into the phloem

• Maximise intracellular compartmentation or allocation to particular parts of the shoot (e.g. pith cells or old leaves)

• Secrete salt via glands to the surface of the leaf or into specific bladder hairs

The extent to which these mechanisms operate in plants varies from species to species, and even within species, and depends on the severity of the stress. Because the control—especially of Na⁺ influx into the root and efflux from the vascular parenchyma into the xylem apoplast—is weak owing to the limited specificity of many nutrient transporters, the most important cellular aspects of salt stress tolerance are efficient removal of Na⁺ from the cytosol especially in young tissues and regulation of Na⁺ distribution within the plant (Tester and Davenport 2003).

Maintenance of low cytosolic Na⁺ is achieved by secretion into the apoplast or sequestration in the vacuole. This applies to cells in the outer part of the root—that is, the rhizodermis and the cortex—as well as the metabolically active cells in the shoot. Fig. 7.34 shows the basic mechanisms by which the cell can manage its cytosolic Na⁺ concentration at the expense of ATP. The dominant systems in the plasma membrane, as well as in the tonoplast, are Na ⁺/H ⁺ antiporters, which use the proton gradients produced by H⁺-ATPases or the vacuolar pyrophosphate-driven proton pump to extrude Na⁺ from the cytosol. The transporters were first identified in A. thaliana and named SOS1 (salt overly sensitive, the protein in the plasma membrane) and NHX (for Na⁺/H⁺ exchanger). Their activity is an important component of Na ⁺ tolerance in both glycophytes and halophytes. Knockout or knock-down of the antiporter expression results in a dramatic increase in salt sensitivity in A. thaliana, tomato and also the halophyte Eutrema salsugineum (formerly known as Thellungiella salsuginea) (Fig. 7.35).

NHX antiporters belong to the large cation/proton antiporter 1 (CPA1) family which, by sequence similarity and intracellular localisation, is further subdivided into vacuolar (class I) and endosomal (class II) NHX transporters. Most of the plant species sequenced to date contain both types of NHX (Bassil et al. 2012). The cation selectivity of AtNHX1 represents an instructive example of post-translational transporter modification (Fig. 7.16). It appears to be controlled by the C-terminal domain reaching into the lumen of the vacuole. Depending on the vacuolar Ca²⁺ concentration and the pH, it binds to a calmodulin-like protein, AtCaM15. Interaction with AtCaM15 decreases the Na⁺ transport activity of AtNHX1 while maintaining the K⁺ transport activity almost unchanged (Yamaguchi et al. 2013). Under normal physiological conditions—that is, a high vacuolar Ca²⁺ concentration and an acidic pH—binding of AtCaM15 favours the K⁺/H⁺ antiport mode. However, as salinity stress causes vacuolar alkalinisation, AtCaM15 dissociates from AtNHX1, which then exhibits higher Na⁺/H⁺ antiport activity and promotes sequestration of Na⁺ into the vacuole. Overexpression of NHX genes improves the salt tolerance of a range of plant species, with a concomitant increase in tissue Na⁺ (Apse and Blumwald 2002).

Vacuolar sequestration and efflux into the apoplast are also the cellular mechanisms underlying salt secretion in specialised leaf structures. In salt bladders (specialised trichomes), salt accumulates in the large central vacuole of the bladder cells. Bursting of the cells then deposits salt on the leaf surface. In salt glands (Fig. 7.36), vacuolar vesicles filled with salt fuse with the plasma membrane for exocytotic release of salt, or Na⁺ ions are directly transported out of the cell across the plasma membrane. Thus, it can be postulated that NHX- and SOS1-like proteins play a key role also in these specific adaptations.

Another major mechanism of salt tolerance is the control of Na ⁺ translocation to the shoot. This is largely a function of re-uptake of Na⁺ from the xylem into xylem parenchyma cells. The transporters responsible for this retrieval are those in the HKT (high-affinity K ⁺ transporter) family (Deinlein et al. 2014). HKTs belong to a superfamily of potassium transporters, which have been found in microorganisms, yeasts, plant cells and parasites such as trypanosomes (Yamaguchi et al. 2013). Two classes can be differentiated on the basis of functional and structural traits: class I, which is more selective for Na⁺; and class II, encompassing K⁺/Na⁺ co-transporters. A. thaliana AtHKT1;1 loss-of-function mutants are Na⁺ hypersensitive and accumulate more Na⁺ in the leaves. Conversely, overexpression under the control of a stele-specific promoter has been shown to reduce Na⁺ transport to the leaves and result in higher salt tolerance (Møller et al. 2009). AtHKT1;1 resides in the plasma membrane of xylem parenchymal cells and phloem tissues. The latter explains recirculation of Na⁺ from the shoot to the root via AtHKT1;1, which may contribute to salt tolerance (Box 7.2).

A common acclimative response of a plant to salinity is lowering of the water potential of its cells. This is achieved by accumulation of low molecular weight compounds (the so-called compatible solutes or osmolytes) in the cell—for example, quaternary ammonium compounds such as glycine betaine, polyamines, open-chain sugar alcohols (polyols) such as mannitol and glycerol, oligosaccharides such as trehalose, and proline (Chap. 6). In addition to their colligative effects, osmolytes can partially replace water, thereby stabilising proteins and cellular substructures. Some osmolytes are rather salinity specific—that is, they are less commonly produced under other stresses such as drought. Very common osmolytes in halophytes—for example, mangroves or ice plants—are cyclic sugar alcohols or cyclitols. In contrast to the open-chain polyols, they show slow metabolism, which prevents their consumption in situations of throttled carbohydrate availability (e.g. when stomata are closed), thus safeguarding the osmolyte function. Their biosynthesis starts from glucose-6-phosphate, which is cyclised to inositol-3-P, the basic compound for a variety of cyclitols. They are varied by relatively simple biochemical reactions such as epimerisation (e.g. L-quebrachitol) or transfer of methyl groups (e.g. D-ononitol), which renders them metabolically rather inactive. Because halophytes can tolerate comparatively higher cytosolic salt concentrations owing to compatible solute accumulation and efficient vacuolar sequestration, they can also use Na⁺ and Cl^– as osmolytes to lower the osmotic potential.

Like cells affected by low water availability, salt-stressed cells synthesise proteins that are assumed to protect cellular structures such as membranes and protein complexes by associating with them. Typical representatives are the dehydrins or LEA proteins, whose structural features and functions in cell biology are discussed in Chap. 6.

Box 7.2: Generation of Plants with Increased Salt Tolerance

The genes encoding SOS1, NHX and HKT transporters have been successfully utilised as genetic tools for enhancing the salt tolerance of model and crop plants. The genes Nax1 and Nax2, identified as the sodium transporters TmHKT1;4-A2 and TmHKT1;5-A (Fig. 7.34) in Triticum monococcum, were transferred to Triticum aestivum (bread wheat), using marker-assisted selection for hexaploid plants containing one or both genes. Expression of Nax1 reduced the Na⁺ content in the leaf blades by 50%, expression of Nax2 reduced it by 30% and expression of both genes together decreased it by 60%. This decrease was at the expense of sodium accumulation in the leaf sheath; nevertheless, the salt tolerance of the bread wheat was substantially improved (Munns et al. 2012) (Fig. 7.37). However, evidence has been provided that these transporters are also involved in other cellular processes. This limits the possibility of further increasing salt tolerance by stronger overexpression of any of them.

7.5.3.5 Sensing of Salinity Stress and Intracellular Signalling

The overall response of a plant to salt stress is highly complex. It is not only removal of Na⁺ from the cytosol and the synthesis of osmolytes that are activated. The transcript abundance of up to several thousand genes changes within hours in the roots of plants exposed to toxic concentrations of NaCl. In A. thaliana the majority of these changes occur in the root cortex, where most of the Na⁺ accumulates (Deinlein et al. 2014).

Plant cells are able to separately sense the two components of salinity. Osmotic stress elicits responses distinct from those to ionic toxicity. The molecular nature of the sensors is still elusive (Fig. 7.34). A change in osmolality generates a stretch force on the plasma membrane, which may activate osmosensors. Mechanosensitive channels are known from yeast and other eukaryotes but not from plants. The plant sensors are expected to be closely associated with Ca ²⁺ channels. Cytosolic Ca²⁺ increases within seconds upon osmotic stress and represents the earliest documented response (also to other stresses such as cold; Chaps. 2 and 4). A recently identified hyperosmolality-gated calcium-permeable channel (OSCA1) is a candidate for an osmosensor (Yuan et al. 2014). Sensing of ionic stress caused by NaCl is less well understood.

The best-characterised salinity-specific signalling pathway is the one leading to the activation of the H ⁺/Na ⁺ antiporter SOS1 (Fig. 7.34). The elevated Ca²⁺ signal is sensed by the calcineurin B–like protein CBL4 (termed SOS3), which responds with dimerisation. The dimer can interact with the protein kinase CIPK24 (CBL-interacting protein kinase), known as SOS2. The SOS3/SOS2 complex is targeted at the plasma membrane, where it phosphorylates the SOS1 protein. Activation of this antiporter requires phosphorylation of its auto-inhibitory domain. Calcium-dependent or calcium-controlled proteins (such as CBL4 or calmodulin) or protein kinases can transduce the salinity signal further downstream and trigger or attenuate gene expression (Fig. 7.34). Some of the calcium-dependent protein kinases regulate the response to abscisic acid (ABA), which accumulates under drought, as well as under salinity stress (Hirayama and Shinozaki 2010) (Chap. 2, Sect. 2.​2).

When a plant is subjected to a salt shock, a dramatic but transient increase in ABA takes place, which is similar to the reaction under drought (Fig. 7.38). Under prolonged salt stress the level of ABA might remain elevated or might return to close to the original concentration, depending on the plant species and plant organ, as well as on the experimental conditions. Part of this reaction might indeed be due to the osmotic stress imposed by high salinity. ABA signalling leads via transcription factors to the enhanced formation and accumulation of compatible solutes and protective proteins such as dehydrins. This process is presented in detail in Chap. 6.

7.5.4 Aluminium Toxicity and Tolerance

Aluminium is the most abundant metallic element in the Earth’s crust, where it appears in many (usually insoluble) compounds and complexes. Below a soil pH of around 5, Al becomes available for plant uptake as the highly phytotoxic Al³⁺ ion. This process is the major reason for the limitation of plant productivity in acidic soils, which are especially prevalent in tropical and subtropical regions. Around 30% of the global land area is affected by soil acidity.

The primary effect of exposure to the highly reactive Al³⁺ is rapid inhibition of root growth (Kochian et al. 2004). Within 1–2 h, cell elongation is halted. Then, with a delay of several hours, cell division is inhibited too. The apex is the most sensitive part of the root. Only a small fraction of the available Al enters the symplast, via routes that are molecularly not well understood. The larger Al fraction (usually >80%) is bound by the cell wall. Several extra- and intracellular structures and processes are affected by Al³⁺. In the apoplast, Al³⁺ interacts with the negative charges of pectins. This compromises the cell extensibility and thereby cell expansion. Inhibition of crucial cell wall enzymes such as expansins is another possible cause of extracellular toxicity. Transport of Ca²⁺, as well as cytosolic Ca signalling, is inhibited. Al³⁺ interferes with Mg²⁺-dependent processes. For instance, Al-ATP interacts far more strongly with enzymes such as hexokinase than Mg-ATP does. Al³⁺ disrupts the dynamics of the cytoskeleton by interacting with microtubules and actin filaments. It can damage DNA and elicits generation of ROS.

Two major mechanisms allow plants to withstand Al toxicity. The first one can be categorised as an avoidance strategy and results in effective exclusion of Al³⁺ from the symplast, thereby reducing the actual exposure of cellular sites to the toxicity of Al³⁺. The second mechanism confers tolerance. It comprises processes that detoxify and sequester Al³⁺ ions after they have entered cells.

Considerable variation exists in both exclusion and intracellular detoxification. Because the genetics of Al exclusion are comparatively simple in some crop species—including wheat, barley and sorghum—it has been possible to molecularly characterise the responsible loci. Differences in the ability to exclude Al³⁺ from root cells are strongly correlated with differences in the secretion of organic acids, mostly malate and citrate, into the rhizosphere (Fig. 7.39).

The organic acids form complexes with Al³⁺ and thereby prevent uptake of Al³⁺ into the root symplast (Fig. 7.40). The transporters responsible for the efflux of organic acids belong to two different classes. The first Al tolerance locus was cloned from wheat and was found to encode an aluminium-activated malate-secreting anion channel (TaALMT1) (Sasaki et al. 2004). The activation is specific to Al³⁺, as it is not generally triggered by trivalent cations (e.g. La³⁺, Sc³⁺). Expression of this channel in other plant species enhances malate secretion and Al³⁺ tolerance. ALMT-like transporters fulfil comparable functions in rye or A. thaliana. Proteins of the MATE (multidrug and toxic compound extrusion) family account for citrate efflux in sorghum, barley and presumably many other plant species (Magalhaes et al. 2007). MATE proteins represent a large group of secondary active transporters in eukaryotes and prokaryotes. They export a wide variety of substrates. Sorghum bicolor SbMATE, the barley protein HvAACT1 and several orthologues in various species mediate Al-activated citrate secretion. The mechanism of Al³⁺ activation is not understood for any of the ALMT1 and MATE proteins. In addition to this post-translational activation, there is up-regulation of expression in some species. Tight regulation of organic acid secretion helps to limit the carbon costs of this mechanism. Secretion is contingent on the presence of potentially dangerous Al³⁺ concentrations and is apparently restricted to particular zones in the root tip where most of the Al³⁺ damage occurs (Kochian et al. 2015).

Under aluminium stress, Al³⁺-sensitive cultivars excrete malate and citrate too, but much less so than tolerant cultivars (Fig. 7.39). The molecular explanation for natural variation in Al exclusion provides an instructive example of how intraspecific differences in stress tolerance can arise. Variations in ALMT1 and MATE expression at the root tip correlate well with the genotypic variation in Al³⁺ tolerance. Regardless of whether gene expression is responsive to Al³⁺ exposure, tolerant genotypes always show higher transcript levels of the organic acid transporter genes than sensitive genotypes (Delhaize et al. 2012). Different types of promoter polymorphisms have been selected that result in higher expression levels of the version present in the tolerant cultivars. In wheat the ALMT1 promoters differ in the number of repeats. More repeats are correlated with stronger promoter activity (Fig. 7.41). High expression of HvAACT1 at the root apex has been traced to a transposon insertion in the promoter in more tolerant cultivars. Sorghum promoters also differ in repeat numbers.

Resistance to high intracellular Al concentrations is observed in several species adapted to acidic soils. Among them are a few exceptional plants, which can even be referred to as Al accumulators, with shoot Al concentrations >10 times higher than those in other plants—that is, up to 3 g/kg of dry biomass. Old leaves of tea accumulate Al to this level and sometimes even higher levels (30 g/kg of dry biomass has been reported). Buckwheat is another example. The accumulation of Al in the vacuoles of Hydrangea is remarkable. The sepals turn from red to blue with the addition of Al³⁺ to the irrigation water. The colour change is due to the formation of a complex of Al with delphinidin-3-glucoside and 3-caffeoylquinic acid (Ma et al. 1997) (Fig. 7.42). Corresponding to the high content of soluble Al³⁺ in acidic soils in the tropics, substantial concentrations of Al (up to 1 g/kg) occur in wild plants (e.g. Melastoma and Vaccinium) in rainforests.

Al tolerance results from effective sequestration of Al by formation of cytosolic complexes and transport into vacuoles. Most of this sequestration occurs in root cells and, in Al-accumulating plants, additionally in the leaf symplast. For example, Al accumulates in buckwheat as an Al–oxalate complex. In Hydrangea the counter anion for Al³⁺ is citrate.

Mechanistic insight into the intracellular sequestration of Al is predominantly available for rice, the most tolerant cereal species. Al tolerance is genetically much more complex in rice than in the other cereals, which indicates a multitude of factors contributing to tolerance. As in barley and sorghum, rice roots secrete organic acids via a transporter in the MATE family (OsFRDL4). However, the contribution of this transporter to overall Al tolerance is much less pronounced. Instead, a major determinant of natural variation in Al tolerance is allelic variation in the Al transporter NRAT1 (Li et al. 2014). This transporter allows entry of apoplastic Al³⁺ into root cells. The beneficial effect of Al³⁺ uptake in rice demonstrates that at least some of the toxicity of Al³⁺ is explained by interaction with extracellular targets. At the same time it indicates efficient intracellular sequestration of Al³⁺. Cytosolic Al³⁺ (whose concentration around neutral pH is very low) and/or Al–organic acid complexes are the substrates for tonoplast-localised ABC transporters (e.g. OsALS1 in rice) or aquaporins.

Accumulation of Al in the leaves of a small subset of Al-tolerant species requires Al mobility. Al is translocated to the shoot complexed with organic acids, taken up into leaf cells and transported into the vacuoles by aquaporins, as shown for Hydrangea macrophylla (Kochian et al. 2015). Transport forms in the xylem of other Al accumulators such as tea, buckwheat and Melastoma are of the same chemical nature—that is, Al complexes with citrate, malate or oxalate.

7.5.5 Non-Essential Toxic Metals

Several non-essential and potentially highly toxic metals or metalloids are present in the environment. They can be taken up and accumulated by plants when available in the soil. The most important elements in this category are Cd, As, Pb and Hg. The causes of their release into the environment can be either natural or anthropogenic. Volcanic emissions are important sources of Hg and As. Cd and Pb are present in Zn minerals and become available through weathering. Potentially toxic concentrations, however, are mostly due to human activities. Metal pollution is a consequence of mining, metal processing and agricultural practices such as the use of fertilisers with toxic metal impurities (Clemens and Ma 2016). Perhaps surprisingly, all plants apparently express genes conferring non-essential metal tolerance, even those that in their natural habitats are very unlikely to encounter threatening concentrations.

As referred to in Sect. 7.2.2, entry of non-essential metal ions into plant cells is due to hitchhiking on transporters of essential macro- and micronutrients or beneficial elements (Clemens 2006). This is well documented for the two main inorganic As species present in soil, arsenate (AsV) and arsenite (AsIII) (Fig. 7.43). The former is taken up by phosphate transporters, the latter by aquaglyceroporins. In rice plants the aquaglyceroporin Lsi1 functions as a silicon transporter. In plants with naturally selected As hypertolerance (e.g. the perennial grass Holcus lanatus) the suppression of high-affinity phosphate uptake represents an important part of the adaptation.

Cd, As, Pb and Hg are thiophilic elements—that is, they have a high affinity for sulphur groups in biological molecules. Uncontrolled binding renders such molecules inactive and eventually causes damage. Thus, the ability to cope with exposure to toxic non-essential elements is conferred by mechanisms that suppress such unwanted interactions. Like the tolerance of essential metal excess, the detoxification of non-essential metals is predominantly achieved by sequestration and efflux to protect the cytosol. The main sequestration route is the phytochelatin (PC) pathway (Fig. 7.43). PCs are derivatives of GSH with the general formula (γ-Glu-Cys)_n-Gly (n is usually between 2 and 5). PCs are synthesised non-ribosomally by the enzyme PC synthase and bind several metals with high affinity. An excess of metals activates the enzyme in the cytosol. Upon formation of PC–metal complexes, these are transported into the vacuole by ABC-type transporters. A second mechanism to maintain low cytosolic concentrations depends on metal-effluxing HMAs—P_1B-type ATPases localised in the plasma membrane or the tonoplast. They play an important role in Cd hypertolerance in Cd-hyperaccumulating plants (Sect. 7.5.2).