*3.2.1. Variation in tolerance and accumulation characteristics*

A plant showing injury or death due to metal stress is deemed sensitive to its environment. On the other hand, resistant plants can survive and reproduce under metal stress conditions (Ernst *et al.,* 2008). In general, plants can achieve resistance to heavy metals by avoidance or

Avoidance occurs when plants restrict the uptake of metals within root tissue by several strategies. In environments where the soil metal contamination is heterogeneously distributed, plants can prevent metal uptake by exploring less contaminated soil. Another avoidance strategy involves mycorrhizal fungi, where they can extend their hyphae outside the plants rooting zone up to several tens of meters and transfer the necessary elements to the plant (Ernst, 2006; Baker, 1987). Also, these metal tolerant fungi can increase plant metal resistance by changing the metals speciation or by restricting the metal transfer into the plant (Ernst, 2006). Arines *et al.* (1989) found that mycorrhizal *Trifolium pratense* (red clover) plants growing in acid soils had lower levels of Mn in their roots and shoots as compared to the non-mycor‐ rhizal plants. Plants can also restrict contaminant uptake in root tissues by immobilizing metals for example through root exudates in the rhizosphere. A role of root exudates is to chelate metals and stop their entry inside the cell. The cell wall has also been found to be involved in

In the absence of avoidance strategies, some plants can grow and survive in soil contaminated with toxic levels of heavy metals which are otherwise lethal or detrimental for growth and survival of others genotypes of the same or of different species (Maestri and Marmiroli 2012). Plants exhibiting tolerance are internally protected from the stress of metals that have entered the cell's cytoplasm (Baker, 1987). Metallophytes (metal tolerant plant) can function normally even in the presence of higher plant-internal metal levels. Plants adapt to their environments by developing heritable tolerance mechanisms. Tolerance to specific metals has evolved independently several times in different species from local non tolerant ancestral plant populations (Schat *et al.,* 2000). Plants can exhibit tolerance to metals that are present in surplus

According to Bradshaw (1991) most species are in a state of genostasis. It is the restriction of genetic variability which limits the evolution of the population/species. In the absence of avoidance pathways, metal contaminated soil acts as a selection force on a population, where only the plants with tolerant genotype can survive and reproduce. This leads to a bottleneck, where few individuals survive and reproduce. In turn, metal tolerant populations can evolve rapidly following a disturbance such as contamination of soil with heavy metals. Plant adaptation to these sites occurs in populations for which tolerance variability already exists prior to the contamination (Maestri and Marmiroli 2012; Baker, 1987). Genes for the tolerance of metals are pre-existing at a low frequency in non tolerant populations of certain plant species

restricting metal uptake into the cell's cytoplasm (Mganga *et al.,* 2011).

in the soil. Each metal is under control of specific genes.

(Ernst, 2006; Macnair, 1987).

tolerance.

**3.1. Avoidance**

56 Environmental Change and Sustainability

**3.2. Tolerance**

Variation occurs between species, populations and clones for tolerance and accumulation of metals. Assunçãno *et al.* (2003) found differences in the degree of chlorosis and concentration of metal for *Thlaspi caerulescens* (currently named *Noccaea caerulescens*) populations when grown in hydroponic solutions containing various Ni, Cd and Zn concentrations. Visioli *et al.* (2010) found differences in growth, morphology and Ni accumulation capacity when the Ni hyperaccumulator *T. caerulescens* and the non-metal adapted *T. caerluescens* were exposed to different Ni concentrations in hydroponic solutions. Besnard *et al.* (2009) used cleaved amplified polymorphic sites (CAPs) and microsatellites to determine the genetic variation for *T. caerulescens* populations from metalliferous and non-metalliferous sites from Switzerland. They found a correlation between the level of heavy metals in soil and the variation at the target loci for the genes involved in encoding metal transporters. Basic *et al.* (2006) found similar results when they analyzed the genetic variation of different *T. caerulescens* population sampled from different soil types with single nucleotide polymorphism (SNPs) in target and non target genes. These results were also observed in *Populus spp.* Marmiroli *et al.* (2011) compared *Populus* clones and found variation in their capacity to accumulate Cd. This variation in Cd accumulation between clones was correlated with SNPs at some target genes. These results imply that gene flow is limited between individuals found on metal contaminated and those from uncontaminated sites, at least for the loci that are involved in the fitness of the individuals (Visioli *et al.,* 2012). On metal-enriched soil, there is a strong selection of local offspring, which conserves the metal tolerant genotypes (Ernst, 2006).

#### *3.2.2. Genetics of tolerance to metals*

Identifying genes involved in a specific adaptation is challenging. Metal tolerance and accumulation in plants are complex genetic systems. Plants have to modify their physiological processes in order to be able to survive in the environment in which they have germinated. In turn, the survival of a population to the contaminated environment is dependent on the inheritance of favourable traits. Tolerance mechanisms are heritable and variable, resulting from genes and gene products (Maestri and Marmiroli 2012). Variation in the evolution of metal tolerance exists over species, populations and clones (Baker, 1987). Some species do not show variation in tolerance and accumulations. In order to determine genes involved in metal tolerance and accumulation, segregating analyses were used, where parents with contrasting phenotypes were crossed to produce progeny. Studies have determined that in many species, metal tolerance and accumulation are genetically independent (Assunçãno *et al.,* 2006). For example, in *Arabidopsis halleri*, Cd tolerance and accumulation segregated as independent traits while Cd and Zn tolerance and accumulation cosegretated. In this later species, two or more genes were proposed to be involved in Cd and Zn accumulation but only one gene for Cd and Zn tolerance (Bert *et al.,* 2003; Bert *et al.,* 2002). In *T. caerulescens*, no genes involved in Cd, Zn and Ni tolerance and accumulation cosegregated. This suggests that there is a high probability that the genetic and physiological mechanisms for these traits are distinct from each other (Yang *et al.,* 2005a; Maestri *et al.,* 2010; Richau and Schat 2009; Assunçãno *et al.,* 2006; Zha *et al.,* 2004). As a result, it is not possible to conclude that a plant with high levels of metals in aerial biomass is also metal tolerant. The concentration of metals in above ground tissue serves as an indication of the plant's potential metal tolerance (Frérot *et al.,* 2010).

Several techniques have been used to isolate and identify genes involved in heavy metal tolerance in plants, one of which is the quantitative trait loci (QTL) mapping. QTL mapping is a powerful tool in examining complex adaptive traits and in determining the number of genes involved in a trait as well as the genes effects and their interactions (Willems *et al.,* 2007). By mapping QTLs, it can be possible to identify or validate candidate genes involved in a complex trait such as metal tolerance and accumulation (Willems *et al.,* 2007). Other techniques used to identify genes for metal tolerance and accumulation are functional complementation in yeast mutants defective in metal homeostasis. These methods use plant cDNA expression libraries, as well as the identification of hypothesized pathways based on sequence similarities with plant cDNA libraries and genomic sequences (Lal, 2010). Transcriptome analyses have also been used to reveal genes involved in hyperaccumulation by analysing the differences in expression profiles or regulation-level of hyperaccumulator and non hyperaccumulator plants (Colzi *et al.,* 2011).

Few specific major genes have been found for Cd, Cu, Ni and Zn tolerance in *Silene vulgaris* by crossing plants from a metalliferous site with non tolerant plant from a nonmetalliferous site (Schat *et al.,* 1996; Schat *et al.,* 1993). Similar results were reported for Cu tolerance for *Mimulus guttatus* (Macnair, 1993) and Zn tolerance for *Arabidipsis halleri* (Bert *et al.,* 2003). In *S. vulgaris* and *M. guttatus*, modifier genes (minor genes) were involved in Cu tolerance, thus increasing tolerance and enhancing the effect of the major gene(s) (Smith and Macnair 1998). Only two QTL were involved in Ni accumulation and tolerance in *S. vulgaris* (Bratteler, 2005).

Studies aiming at identifying associations between molecular markers and metal tolerance and accumulation trait have been performed using interspecific and intraspecific crosses. When a high Zn accumulating *T. caerulescens* parent was crossed with a low Zn accumulating parent, two major QTLs were found to be involved in the increased of Zn accumulation in root (Assunçãno *et al.,* 2006). Deniau *et al.,* (2006) performed QTL mapping for the hyperaccumu‐ lation of Zn and Cd in *T. caerulescens*. They found two QTLs responsible for Cd and two for Zn accumulation in roots. In addition, one QTL for Cd and three QTLs for Zn accumulation in shoot were characterized. Macnair *et al.* (1999) reported a major gene involved in Zn tolerance from the analysis of F2 progeny derived from a cross between *A. halleri* (tolerant parent) and *Arabidopsis lyrata* (sensitive parent). Willems *et al.* (2007) generated a backcross progeny from the interspecific cross between *A. halleri* (tolerant parent) and *A. lyrata (*sensitive parent) and identified three major additive QTLs involved in Zn tolerance in *A. halleri*. These QTLs were mapped to three different chromosomes (3, 4 and 6) and colocalized with genes that have been known to be involved in metal tolerance and accumulation. *HMA4* (*Heavy Metal ATPase 4)* encodes a P-type ATPase pump localized at the plasmamembrane involved in loading Zn and Cd into the xylem. *MTP1-*A and *MTP1-B* are Metal Tolerance Protein- vacuolar transporters that are involved in Zn tolerance) (Gustin *et al.,* 2009; Krämer, 2005). Three new QTLs were identified and mapped to chromosomes 4, 6 and 7 by Filatov *et al.* 2007, when F2 progenies from a similar interspecific cross were analyzed. Frérot *et al.* (2010) also found Zn accumulation to be polygenic using *A. halleri* X *A. lyrata petraea* progenies. They determined that Zn accumulation is controlled by two QTLs in low Zn concentration and three QTLs in high Zn concentration. Four of the five QTLs mapped for Zn accumulation in their study were also reported in previous studies using *A. halleri* X *A. lyrata petraea* progenies (Frérot *et al.,* 2010; Filatov *et al.,* 2007; Filatov *et al.,* 2006).

aerial biomass is also metal tolerant. The concentration of metals in above ground tissue serves

Several techniques have been used to isolate and identify genes involved in heavy metal tolerance in plants, one of which is the quantitative trait loci (QTL) mapping. QTL mapping is a powerful tool in examining complex adaptive traits and in determining the number of genes involved in a trait as well as the genes effects and their interactions (Willems *et al.,* 2007). By mapping QTLs, it can be possible to identify or validate candidate genes involved in a complex trait such as metal tolerance and accumulation (Willems *et al.,* 2007). Other techniques used to identify genes for metal tolerance and accumulation are functional complementation in yeast mutants defective in metal homeostasis. These methods use plant cDNA expression libraries, as well as the identification of hypothesized pathways based on sequence similarities with plant cDNA libraries and genomic sequences (Lal, 2010). Transcriptome analyses have also been used to reveal genes involved in hyperaccumulation by analysing the differences in expression profiles or regulation-level of hyperaccumulator and non hyperaccumulator plants

Few specific major genes have been found for Cd, Cu, Ni and Zn tolerance in *Silene vulgaris* by crossing plants from a metalliferous site with non tolerant plant from a nonmetalliferous site (Schat *et al.,* 1996; Schat *et al.,* 1993). Similar results were reported for Cu tolerance for *Mimulus guttatus* (Macnair, 1993) and Zn tolerance for *Arabidipsis halleri* (Bert *et al.,* 2003). In *S. vulgaris* and *M. guttatus*, modifier genes (minor genes) were involved in Cu tolerance, thus increasing tolerance and enhancing the effect of the major gene(s) (Smith and Macnair 1998). Only two QTL were involved in Ni accumulation and tolerance in *S. vulgaris* (Bratteler, 2005).

Studies aiming at identifying associations between molecular markers and metal tolerance and accumulation trait have been performed using interspecific and intraspecific crosses. When a high Zn accumulating *T. caerulescens* parent was crossed with a low Zn accumulating parent, two major QTLs were found to be involved in the increased of Zn accumulation in root (Assunçãno *et al.,* 2006). Deniau *et al.,* (2006) performed QTL mapping for the hyperaccumu‐ lation of Zn and Cd in *T. caerulescens*. They found two QTLs responsible for Cd and two for Zn accumulation in roots. In addition, one QTL for Cd and three QTLs for Zn accumulation in shoot were characterized. Macnair *et al.* (1999) reported a major gene involved in Zn tolerance from the analysis of F2 progeny derived from a cross between *A. halleri* (tolerant parent) and *Arabidopsis lyrata* (sensitive parent). Willems *et al.* (2007) generated a backcross progeny from the interspecific cross between *A. halleri* (tolerant parent) and *A. lyrata (*sensitive parent) and identified three major additive QTLs involved in Zn tolerance in *A. halleri*. These QTLs were mapped to three different chromosomes (3, 4 and 6) and colocalized with genes that have been known to be involved in metal tolerance and accumulation. *HMA4* (*Heavy Metal ATPase 4)* encodes a P-type ATPase pump localized at the plasmamembrane involved in loading Zn and Cd into the xylem. *MTP1-*A and *MTP1-B* are Metal Tolerance Protein- vacuolar transporters that are involved in Zn tolerance) (Gustin *et al.,* 2009; Krämer, 2005). Three new QTLs were identified and mapped to chromosomes 4, 6 and 7 by Filatov *et al.* 2007, when F2 progenies from a similar interspecific cross were analyzed. Frérot *et al.* (2010) also found Zn accumulation to be polygenic using *A. halleri* X *A. lyrata petraea* progenies. They determined

as an indication of the plant's potential metal tolerance (Frérot *et al.,* 2010).

(Colzi *et al.,* 2011).

58 Environmental Change and Sustainability

Courbot *et al.* (2007), also using progeny from interspecific cross between *A. halleri* and *A. lyarata* determined three QTLs involved in Cd tolerance. A major QTL region was found to be common to Cd (Courbot et al 2007) and Zn (Willems et al 2007) tolerance and was colocalized with the *HMA4* gene. Hanikenne et al (2008) identified the role of *HMA4* using RNAi-mediated silencing. They reported that when the expression of *HMA4* was down-regulated, less Zn was translocated from the root to the shoot. When this gene was expressed in *A. thaliana*, an increase in Zn translocation to aerial tissue was observed. This increase in Zn translocation in *A. thaliana* plants resulted in signs of Zn hypersensitivity. Therefore, the expression of *AhH‐ MA4* alone was not adequate for Zn detoxification. Additional genes are involved in the *A. halleri* Zn hyperaccumulation (Hanikenne and Nouet 2011; Frérot *et al.,* 2010).

Using segregating progeny resulting from intraspecific crosses between a high Cd accumulating parent and a low Cd accumulating parent for *Glycine max*, Benitez *et al.* (2010) identified a major QTL in seeds that was named cd1. This gene was mapped on chromo‐ some 9. Jegadeesan *et al.* (2010) also identified a major QTL, *cda1*, associated with Cd accumulation in seeds of *G. max*. These two major QTLs were mapped to the same region of chromosome 9 which suggested that cd1 and *cda1* may be identical. Both QTLs were found to be a dominant major gene involved in the control of low Cd uptake*.* By analyz‐ ing the *G. max* genome, Benitez *et al.* (2010) revealed that the cd1 QTL is localized in the vicinity of the P1B-ATPase gene (designated as *GmHMA1)* and proposed that this gene is involved in the transport of Cd. Benitez *et al.* (2012) found a single-base substitution between two cultivars, Harosoy (high Cd content in seed Cd) and Fukuyutaka (low Cd content in seed) in this P1B–ATPase gene. This mutation resulted in an amino acid substitution (glycine in Fukuyutaka and glutamic acid in Harosoy) in GmHMA1a. Since the glycine residue at the amino acid substitution site was conserved in AtHMA3, AtHMA4, AtHMA6 and AtHMA7, it was suggested that the GmHMA1a from Fukuyuta‐ ka was the wild type, responsible for low Cd accumulation in seed (Benitez *et al.,* 2012). A dominant major gene involved in the control of Cd uptake was also observed in wheat (*Triticum aestivum*) (Clarke *et al.,* 1997) and oat (*Avena sativa* L.) (Tanhuanpää *et al.,* 2007). QTL analyses have also been performed in radish (*Raphanus sativus* L.). Xu *et al.* (2012) found a major QTL and three minor QTLs responsible for Cd accumulation in radish roots which were mapped on linkage groups 1, 4, 6 and 9. Induri *et al.* (2012) identified major QTLs for Cd response in *Populus* by performing a pseudo-backcross pedigree of *Populus trichocarpa* Torr. & Gray and *Populus deltoides* Bart. These QTLs were mapped to two different linkage groups. They performed a whole-genome microarray study and they were able to identify nine Cd responsive genes, which included a metal transporter, putative transcription factor and an NHL repeat membrane-spanning protein. Additional candi‐ date genes located in the QTL intervals included a glutathione-S-transferase and putative homolog of a glutamine cysteine ligase.

Several QTL studies on rice (*Oryza sativa* L.) have been conducted to determine the number of genes involved in metal accumulation and tolerance. Three putative QTLs involved in Cd accumulation have been found on chromosomes 3, 6 and 8 (Ishikawa *et al.,* 2010; Ishikawa *et al.,* 2005). Ueno *et al.* (2009) also identified another major QTL for Cd accumulation in *O. sativa* that was mapped on the short arm of chromosome 7. QTLs for the translocation of Cd from roots to sink regions were reported in *O. sativa* (Xu *et al.,* 2012; Tezuka *et al.,* 2010). Tezuka *et al.* (2010) revealed a major QTL (*qCdT7*), mapped to chromosome 7, which controlled the translocation of Cd from roots to shoots. This QTL explained 88% of the phenotypic variation indicating that low Cd accumulation was a dominant trait. Dufey *et al.* (2009), using recombi‐ nant inbred lines, identified in *O. sativa* 24 putative QTLs involved in Fe tolerance which were mapped to chromosomes 1, 2, 3, 4, 7 and 11. In addition, two QTLs, located on chromosomes 2 and 3, were involved in As concentration in shoots and in roots respectively.

In durum wheat (*Triticum durum*, L.), Cd accumulation is controlled by a major gene named *Cdu1* and localized on chromosome 5BL (Knox *et al.,* 2009; Clarke *et al.,* 1997). Further, Ci *et al.* (2012) characterized 26 QTLs involved in Cd tolerance and accumulation in *T. aestivum*, where 16 were involved in Cd stress control, 8 for Cd tolerance and 2 for Cd accumulation in roots. In *A. sativa* L., a single QTL for Cd accumulation in grain has been reported (Tanhuanpää *et al.,* 2007).

In wheat (*T. aestivum* L.), Mayowa and Miller (1991) reported QTLs involved in Cu tolerance and accumulation that were mapped to chromosomes 5A, 4D, 7A, 7B, 7D. Ganeva *et al.* (2003) also characterized QTLs for *T. aestivum* on chromosomes 1A, 1D, 3A, 3B, 4A and 7D. Bálint *et al.* (2003) identified QTLs associated with Cu tolerance located on *T. aestivum* chromosomes 3D, 5A, 5B, 5D, 6B and 7D. In addition, Bálint *et al.* (2007) also determined QTLs for Cu tolerance in *T. aestivum*. They reported one major QTL for Cu tolerance on chromosome 5D and minor QTLs on chromosomes 1A, 2D, 4A, 5B and 7D. A QTL affecting shoot Cu content under Cu stress conditions was mapped on chromosome 1BL and an additional QTL for Cu accumulation was found on chromosome 5AL. The role of these genes located on various chromosomes in these different studies suggests that Cu tolerance is a polygenic character, as well as the possibility of different gene expressions against distinct toxic Cu concentrations in different populations. The accumulation of Cu in the shoots is affected by different QTLs, suggesting a strong metal-specific uptake and/or translocation. Bálint *et al.* (2007) reported a negative correlation between Cu tolerance and accumulation in the shoot indicating that the key tolerance mechanism in wheat could be the restriction of Cu uptake in the roots or the reduced translocation from root to shoot.

#### **3.3. Categories of plants growing on metal contaminated soils**

Baker and Walker (1990) categorized plants into three groups according to their strategy for coping with metal toxicity in soil; metal excluders, indicators and accumulators/ hyperaccumulators.
