Abstract

Understanding genetic variation for the response of Populus to heavy metals like cadmium (Cd) is an important step in elucidating the underlying mechanisms of tolerance. In this study, a pseudo-backcross pedigree of Populus trichocarpa Torr. & Gray and Populus deltoides Bart. was characterized for growth and performance traits after Cd exposure. A total of 16 quantitative trait loci (QTL) at logarithm of odds (LOD) ratio ≥ 2.5 were detected for total dry weight, its components and root volume. Major QTL for Cd responses were mapped to two different linkage groups and the relative allelic effects were in opposing directions on the two chromosomes, suggesting differential mechanisms at these two loci. The phenotypic variance explained by Cd QTL ranged from 5.9 to 11.6% and averaged 8.2% across all QTL. A whole-genome microarray study led to the identification of nine Cd-responsive genes from these QTL. Promising candidates for Cd tolerance include an NHL repeat membrane-spanning protein, a metal transporter and a putative transcription factor. Additional candidates in the QTL intervals include a putative homolog of a glutamate cysteine ligase, and a glutathione-S-transferase. Functional characterization of these candidate genes should enhance our understanding of Cd metabolism and transport and phytoremediation capabilities of Populus.

Introduction

One of the challenges posed by phytoremediation of cadmium (Cd) contamination is the development of plants that are tolerant of the deleterious effects of this toxin, which has no known nutritional value in plants. Symptoms of Cd toxicity include leaf roll, chlorosis and reductions in root and shoot growth, even at minute concentrations (Milone et al. 2003, Almeida et al. 2007). Cadmium compromises the structure and function of many proteins by acting on sulphydryl groups (Vanassche and Clijsters 1990) and affects the permeability of membranes by altering lipid composition (Ouariti et al. 1997).

In addition to disrupting the ion and water balance of plants, Cd affects various physiological processes such as photosynthesis, respiration, nitrate assimilation, metabolite accumulation and enzyme activity (Smeets et al. 2005, Solti et al. 2008). Cadmium decreases the photosynthetic activity of plants by changing chloroplast ultrastructure (Sandalio et al. 2001) and altering stomatal closure. Cadmium also interferes with the uptake and distribution of other nutrients (Shao et al. 2008, Rodriguez-Serrano et al. 2009), most notably iron (Fe) (Siedlecka and Krupa 1999, Fodor et al. 2005, Solti et al. 2008). Cadmium has also been shown to affect the electron transport system, interact with antioxidative defense systems and induce lipid peroxidation (Smeets et al. 2005, Rodriguez-Serrano et al. 2009).

Most phytoremediation research so far has been carried out on herbaceous plants that accumulate high levels of metals in their cells naturally. However, their slow growth rates, annual habit, small stature, low biomass and/or narrow geographic adaptability make them unsuitable for phytoremediation on an operational scale (Eapen and D'Souza 2005). Populus, with its rapid growth, makes an ideal plant species for phytoremediation of soils polluted by heavy metals (Cunningham and Ow 1996). Furthermore, Populus is readily propagated from vegetative cuttings, and has a wide range of adaptability, high bulk root volume, high transpiration rate and large stature, all of which should enhance the efficacy of phytoremediation (Rockwood et al. 2004). Previous studies have assessed the potential of using Populus to remediate soils polluted by atrazine (Burken and Schnoor 1997), trichloroethylene (Newman et al. 1997), Cd (Robinson et al. 2000), selenium (Pilon-Smits et al. 1998) and zinc (Di Baccio et al. 2009). In addition, some hybrid Populus clones have shown elevated tolerance of Cd in field plantings relative to other tree species (Migeon et al. 2009), and there is considerable natural variation within the genus that could be exploited to enhance Cd tolerance further using biotechnological approaches.

To further characterize Cd tolerance in Populus, we (i) quantified phenotypic and genetic variation for Cd tolerance and Cd accumulation in a pseudo-backcross multiple generation Populus pedigree that was grown under greenhouse conditions; (ii) detected quantitative trait loci (QTL) for traits related to Cd tolerance including dry weight components and root volume; and (3) identified candidate genes in the QTL intervals based on transcriptional data from a microarray study.

Material and methods

Plant materials

An interspecific mapping population was developed at the University of Minnesota Natural Resources Research Institute by crossing Populus trichocarpa Torr. & Gray clone 93-968 from western Washington to Populus deltoides Bart. clone ILL-101 from southern Illinois. The resulting F1 hybrid, 52-225, was in turn backcrossed to P. deltoides clone D-124 from Minnesota. Two-hundred and fifty-two full-sib progeny from the ‘pseudobackcross’ pedigree (Family 52-124) were used for the current study. Dormant cuttings were collected from a stoolbed in early spring prior to bud break and were stored at 4 °C for 5 weeks to meet chilling requirements. This facilitated uniform rooting of the cuttings, and enabled selection of plants with approximately uniform sizes at the initiation of our experiment. Two ramets per clone were used for both control and Cd treatments. A total of 1008 vegetative cuttings (252 genotypes × 2 treatments 2 treatments × 2 cuttings per genotype) were examined in this experiment.

Cuttings with approximate lengths of 15 cm and diameters of 1 cm were dipped in 1 : 10 dilution of a commercial liquid rooting hormone solution (Dip N Grow 1% indole-3-butyric acid, 0.5% naphthalene acetic acid) for 15 s, surrounded by foam stoppers and plugged into 3 cm holes at 8 cm spacing in the lids of polyethylene containers. Approximately 5 cm of the base of the cuttings was submerged beneath a dilute, continuously aerated nutrient solution, 0.1 × Johnson's solution (Siddique et al. 1990). The nutrient solution was composed of 400 µM NH4NO3, 400 µM KNO3, 200 µM Ca(NO3)2·4H2O, 100 µM MgSO4·7H2O, 50 µM K2HPO4, 20 µM KCl, 25 µM H3BO3, 0.5 µM Na2MoO4·2H2O, 2 µM MnSO4·4H2O, 0.5 µM CuSO4·5H2O, 2 µM ZnSO4·7H2O, 0.5 µM CoCl2 and 20 µM Fe-Na EDTA. The pH was ∼5.4. One primary shoot was maintained per cutting.

Once the cuttings developed roots and shoots, they were transplanted into 126 polyethylene containers (24 cm × 12 cm × 38 cm) following a partially balanced incomplete block design, with two treatments (Cd and control), 126 containers as incomplete blocks and two replicates of each genotype in each treatment. Each container had one ramet from each of eight different genotypes at 9 cm spacing and contained 7 l of the 0.1× Johnson's solution described above. The solutions were changed every 3 days throughout the experiment (from 19 May 2007 to 10 July 2007) to maintain nutrient and oxygen levels and to avoid algal growth.

The experiment was carried out with a target day temperature of 25 °C and a night temperature of 21 °C and a photoperiod of 18 h. Artificial lighting was provided by high-density mixed halide lights and photosynthetically active radiation (PAR) was recorded for every container used in the study, at plant height (average PAR = 224.05 µmol/m2/s and standard deviation = 55.08). Temperature was maintained by an evaporative cooling system.

Cuttings were grown for 40 days and one-half of the containers were randomly selected to receive a 25 µM CdCl2 treatment for an additional 2 weeks. The remaining half of the containers were maintained as described above for 2 weeks and used as controls. The Cd concentration was selected based on pilot experiments that revealed substantial phenotypic variation in this family without excessive mortality. Fresh CdCl2 was added each time solutions were changed. Before and after 20 days of treatment, measurements were taken on shoot length, cutting diameter, shoot diameter, PAR, root collar diameter (the diameter recorded right above the origin point of the uppermost root on a cutting) and rooted length for each cutting (portion of the cutting occupied by roots). Furthermore, root volume was estimated using water displacement in graduated cylinders. After 2 weeks, roots and shoots were harvested, separated into paper bags and dried in an oven at 60 °C for 6 days. Roots were washed with deionized water and blotted on paper towels before being stored in the bags. Dry weights were recorded for leaves, roots and shoots separately, and total dry weights were calculated by summing all components.

Statistical analysis of phenotypic data

Phenotypic data were tested for normality using the Shapiro–Wilk test (Shapiro and Wilk 1965) and logarithmic transformations were performed to mitigate heteroscedasticity of variance versus predicted values. Data were analyzed with mixed linear models using SAS JMP, version 8.0 (SAS Institute Inc., Cary, NC, USA). The general form of the statistical model was

Yijk = μ + Ti + Bj + Gk + εijk

where Yijk was the measurement of response variable Y (total dry weight, root dry weight, leaf dry weight, stem dry weight or root volume) for genotype k in container j under treatment i, μ was the population mean, Ti was the effect of treatment i, Bj was the effect of incomplete block (container) j, Gk was the effect of genotype k and εijk was the experimental error. When significant (P ≤ 0.05), the rooted length and dry weight of each cutting and PAR for each container were also included as covariates. Genotype and all interactions with genotype were treated as random effects in the model, whereas all remaining variables were treated as fixed effects.

Broad-sense heritabilities (H2) were calculated as H2 = VG/(VG + Vε) based on the variance components for genotype (VG) and error (Vε) estimated using restricted maximum likelihood (REML) in mixed linear models that were fitted separately for control and Cd treatments. This was done to assess differences in variances and heritabilities under these two conditions.

The effect of Cd on each genotype (hereafter referred to as ‘Cd effects’) was estimated by subtracting its best linear unbiased predictor (BLUP) under Cd conditions from that under control conditions (Coles et al. 2010) for all biomass traits described above. The genotypes with the largest BLUP differences were identified as Cd sensitive and those with the smallest BLUP differences were identified as Cd tolerant. Genetic correlations among traits were calculated using BLUPs and the REML method in SAS JMP, version 8.0.

Linkage map construction

To construct a genetic map for Family 52-124, 188 progeny were genotyped using 590 amplified fragment length polymorphism markers. In addition, 418 progeny were genotyped using 287 simple sequence repeat (SSR) markers (Yin et al. 2009) that were chosen based on their physical locations along chromosomes to achieve an even distribution of markers on each linkage group. Four aneuploid trees were identified by genotyping with fully informative SSR markers and these individuals were later removed from the mapping data set. Marker generation, genotyping and nomenclature were performed as described previously (Tuskan et al. 2004, Yin et al. 2004). Map construction was conducted using JoinMap 3.0 under the CP cross type (Van Ooijen 2001).

Quantitative trait loci analysis

Quantitative trait loci were identified using MapQTL version 5.0 (Van Ooijen 2004). Interval mapping was performed initially to identify markers associated with putative QTL. Markers with significant effects in this analysis were then used as cofactors in restricted multiple QTL model mapping (Van Ooijen 2004). Logarithm of odds (LOD) scores were calculated at 1.0 cM intervals. Only QTL with LOD score > 2.5 are reported here.

Microarray study

A whole-genome microarray study was conducted using two genotypes that demonstrated differential Cd effects in the above QTL hydroponic study: a Cd-tolerant genotype (1–183) and a Cd-susceptible genotype (182). The experiment consisted of six polyethylene containers as described above but with only four plants per container. Each genotype was represented by two ramets per container with randomly assigned locations. Three containers were treated with 25 µM CdCl2 and the remaining three containers were under control conditions. Three ramets from each genotype × treatment combination were harvested at 24 h, and three at 72 h after Cd treatment. Roots from each plant were thoroughly washed in deionized water and blotted dry and all material was quick frozen in liquid nitrogen and stored at − 80 °C.

Plant tissue was ground in liquid nitrogen, RNA was extracted using the Qiagen RNA Mini kit (Qiagen, Germantown, MD, USA), and double-stranded DNA was synthesized using the Invitrogen SuperScript™ double-stranded cDNA Synthesis Kit (Invitrogen, Carlsbad, CA, USA). Labeling, hybridization and scanning were performed by NimbleGen (Roche NimbleGen, Madison, WI, USA). The raw data were normalized using the robust multichip averaging procedure, which performs a convolution background correction, quantile normalization and summarization based on a multi-array model fit robustly using the median polish algorithm (Bolstad et al. 2005). Differentially expressed genes (Cd treated versus control) were determined using rank product analysis (Breitling et al. 2004) using 1000 permutations and a percentage of false positives cutoff of 0.05.

Results

Cadmium effects

The effect of Cd was evident for all measured traits (Table 1). Cadmium symptoms included necrotic spots on leaves and browning of roots (Figure 1). The genotypes under Cd conditions showed poor growth compared with those under control conditions (Table 1). The effects of genotype, treatment and their interaction were highly significant (P < 0.001) for total dry weight, root dry weight, leaf dry weight, stem dry weight and root volume. The covariate effects of cutting rooted length, cutting dry weight and PAR were also significant (P < 0.001) for all traits except for PAR for root volume (Table 2).

Figure 1.

Morphological effects of Cd treatment in a hydroponic experiment: (a) control roots; (b) Cd-treated roots; (c) control leaf (d) and Cd-treated leaf.

Figure 1.

Morphological effects of Cd treatment in a hydroponic experiment: (a) control roots; (b) Cd-treated roots; (c) control leaf (d) and Cd-treated leaf.

Table 1.

Summary statistics and heritability estimates for total dry weight and other measured traits along with their variance components.

 Mean (g) SD1 Range SE2 VG3 Vp4 H2BS5 
Control        
Total dry weight C 6.35 4.11 24.98 0.19 0.170 0.471 0.361 
Leaf dry weight C 3.57 2.28 13.05 0.10 0.160 0.499 0.321 
Root dry weight C 1.14 0.88 7.34 0.04 0.201 0.581 0.346 
Stem dry weight C 1.63 1.16 7.39 0.05 0.233 0.597 0.390 
Root volume C 1.62 0.94 6.00 0.04 0.045 0.235 0.191 
Cadmium        
Total dry weight Cd 2.52 1.63 8.24 0.07 0.153 0.450 0.340 
Leaf dry weight Cd 1.46 1.01 5.23 0.04 0.147 0.533 0.276 
Root dry weight Cd 0.45 0.33 2.51 0.01 0.179 0.558 0.322 
Stem dry weight Cd 0.60 0.43 2.70 0.02 0.153 0.529 0.288 
Root volume Cd 0.97 0.64 9.00 0.02 0.024 0.218 0.110 
Cd mg/g 0.34 0.27 1.51 0.01 0.108 0.859 0.126 
 Mean (g) SD1 Range SE2 VG3 Vp4 H2BS5 
Control        
Total dry weight C 6.35 4.11 24.98 0.19 0.170 0.471 0.361 
Leaf dry weight C 3.57 2.28 13.05 0.10 0.160 0.499 0.321 
Root dry weight C 1.14 0.88 7.34 0.04 0.201 0.581 0.346 
Stem dry weight C 1.63 1.16 7.39 0.05 0.233 0.597 0.390 
Root volume C 1.62 0.94 6.00 0.04 0.045 0.235 0.191 
Cadmium        
Total dry weight Cd 2.52 1.63 8.24 0.07 0.153 0.450 0.340 
Leaf dry weight Cd 1.46 1.01 5.23 0.04 0.147 0.533 0.276 
Root dry weight Cd 0.45 0.33 2.51 0.01 0.179 0.558 0.322 
Stem dry weight Cd 0.60 0.43 2.70 0.02 0.153 0.529 0.288 
Root volume Cd 0.97 0.64 9.00 0.02 0.024 0.218 0.110 
Cd mg/g 0.34 0.27 1.51 0.01 0.108 0.859 0.126 

1Standard deviation.

2Standard error.

3Genetic variance.

4Phenotypic variance (VG + Vε).

5Broad-sense heritability.

Table 2.

Results of ANOVA for traits measured in the greenhouse hydroponic study that showed significant differences between Cd and control conditions.

 Total dry weight
 
Leaf dry weight
 
Root dry weight
 
Stem dry weight
 
Root volume
 
Trait F P F P F P F P F P 
Genotype 1.97 <0.0001 1.86 <0.0001 1.76 <0.0001 1.97 <0.0001 1.61 <0.0001 
Treatment 589.94 <0.0001 525.75 <0.0001 335.03 <0.0001 555.28 <0.0001 215.37 <0.0001 
Genotype × treatment 1.49 0.0002 1.41 0.0009 1.28 0.0124 1.53 <0.0001 1.34 0.0044 
Rooted length of cutting 10.08 0.0016 5.99 0.0147 16.76 <0.0001 7.71 0.0057 13.92 0.0002 
Cutting dry weight 139.78 <0.0001 130.92 <0.0001 56.23 <0.0001 146.18 <0.0001 51.94 <0.0001 
PAR 6.87 0.0091 4.85 0.0281 4.30 0.0388 9.18 0.0026 0.70 0.4027 
 Total dry weight
 
Leaf dry weight
 
Root dry weight
 
Stem dry weight
 
Root volume
 
Trait F P F P F P F P F P 
Genotype 1.97 <0.0001 1.86 <0.0001 1.76 <0.0001 1.97 <0.0001 1.61 <0.0001 
Treatment 589.94 <0.0001 525.75 <0.0001 335.03 <0.0001 555.28 <0.0001 215.37 <0.0001 
Genotype × treatment 1.49 0.0002 1.41 0.0009 1.28 0.0124 1.53 <0.0001 1.34 0.0044 
Rooted length of cutting 10.08 0.0016 5.99 0.0147 16.76 <0.0001 7.71 0.0057 13.92 0.0002 
Cutting dry weight 139.78 <0.0001 130.92 <0.0001 56.23 <0.0001 146.18 <0.0001 51.94 <0.0001 
PAR 6.87 0.0091 4.85 0.0281 4.30 0.0388 9.18 0.0026 0.70 0.4027 

There was extensive variation in Cd responses among genotypes (Figures 2 and 3). Broad-sense heritabilities of total dry weight and its components were relatively low and comparable between control (range H2 = 0.19–0.39) and Cd conditions (range H2 = 0.11–0.34, Table 1). Root volume showed the lowest heritabilities under either control (H2 = 0.19) or Cd (H2 = 0.11) conditions. In contrast, total dry weight had heritabilities of 0.36 and 0.34 in control and Cd conditions, respectively (Table 1).

Figure 2.

Genotypic variation in visible effects of Cd treatment. (a) Differences between control (right two plants) and Cd (left two plants) treatments. (b) Variation among genotypes under Cd conditions. Blue bar is 30.5 cm.

Figure 2.

Genotypic variation in visible effects of Cd treatment. (a) Differences between control (right two plants) and Cd (left two plants) treatments. (b) Variation among genotypes under Cd conditions. Blue bar is 30.5 cm.

Figure 3.

Distributions of the difference in BLUPs of total dry weight between control (C) and cadmium (Cd) treatments. Genotypes from the left portion of the curve are relatively tolerant of Cd treatment.

Figure 3.

Distributions of the difference in BLUPs of total dry weight between control (C) and cadmium (Cd) treatments. Genotypes from the left portion of the curve are relatively tolerant of Cd treatment.

Quantitative trait loci identification

Quantitative trait loci were mapped for 16 phenotypes with a minimum LOD score of 2.5. Four of these QTL were identified in control conditions and 12 under Cd conditions. Quantitative trait loci were located on LG III and XVI (Figure 4; Table 3). The phenotypic variation explained by QTL ranged from 5.9 to 11.6% and averaged 8.2% across all traits. Excluding co-localizing QTL (±1 LOD) and accounting for correlations among phenotypic traits (Table 4; see Figure S1 available as Supplementary Data at Tree Physiology Online), these 16 QTL correspond to four independent positions within the genome (Table 3), and ranged in size from ∼15 to 30 cM (Figure 4).

Figure 4.

Logarithm of odds scores from the QTL analysis. (a) Total dry weight control: linkage group III. (b) Total dry weight Cd: linkage group XVI. (c) Total dry weight control–Cd: linkage group III. (d) Total dry weight control–Cd: linkage group XVI.

Figure 4.

Logarithm of odds scores from the QTL analysis. (a) Total dry weight control: linkage group III. (b) Total dry weight Cd: linkage group XVI. (c) Total dry weight control–Cd: linkage group III. (d) Total dry weight control–Cd: linkage group XVI.

Table 3.

Linkage group (LG), QTL positions, logarithm of odds ratio (LOD) scores, percent variation explained (PVE) and direction of effect of each QTL are given under the control (C) and cadmium-treated (Cd) conditions.

Locus Trait Linkage group Position LOD PVE Left marker Right marker T effects1 D effects2 
cd1 Total dry weight C–Cd3 III 109.267 3.33 8.9 *CCCGA_8 *CACAC_218  − 0.047 0.043 
cd1 Root volume C–Cd III 110.267 3.04 *CCCGA_8 *CACAC_218  − 0.025 0.022 
cd1 Leaf dry weight C–Cd III 111.267 3.06 7.7 *CCCGA_8 *CACAC_218  − 0.046 0.039 
cd1 Total dry weight C III 111.676 2.51 6.1 *CCCGA_8 *CACAC_218  − 0.025 0.022 
cd1 Root volume C III 114.726 2.5 6.8 *CCCGA_1 *CCCAC_22  − 0.014 0.013 
cd1 Root dry weight C III 110.267 2.66 6.6 *CCCGA_8 *CACAC_218  − 0.021 0.022 
cd2 Root dry weight C–Cd XVI 22.784 2.99 7.4 G_3141 *CCCCT_202R 0.045  − 0.030 
cd2 Total dry weight C–Cd XVI 26.609 2.81 6.8 *CCCCT_202R *TCCGT_8R 0.047  − 0.033 
cd2 Leaf dry weight C–Cd XVI 26.609 4.26 10 *CCCCT_202R *TCCGT_8R 0.056  − 0.042 
cd2 Total dry weight Cd XVI 26.609 3.14 7.8 *CCCCT_202R *TCCGT_8R  − 0.029 0.021 
cd2 Root dry weight Cd XVI 26.609 2.99 7.3 *CCCCT_202R *TCCGT_8R  − 0.027 0.017 
cd2 Leaf dry weight Cd XVI 26.609 4.6 11.6 *CCCCT_202R *TCCGT_8R  − 0.037 0.028 
cd3 Leaf dry weight C XVI 56.812 2.9 5.9 *CCCCT_246R *CTCAG_356R 0.033  − 0.016 
cd3 Total dry weight C–Cd XVI 56.812 3.61 7.4 *CCCCT_246R *CTCAG_356R 0.059  − 0.028 
cd3 Leaf dry weight C–Cd XVI 56.812 5.41 11 *CCCCT_246R *CTCAG_356R 0.072  − 0.036 
cd4 Leaf dry weight Cd XVI 68.558 4.61 11.3 *CGCCA_308R *CTCTC_100R  − 0.040 0.025 
Locus Trait Linkage group Position LOD PVE Left marker Right marker T effects1 D effects2 
cd1 Total dry weight C–Cd3 III 109.267 3.33 8.9 *CCCGA_8 *CACAC_218  − 0.047 0.043 
cd1 Root volume C–Cd III 110.267 3.04 *CCCGA_8 *CACAC_218  − 0.025 0.022 
cd1 Leaf dry weight C–Cd III 111.267 3.06 7.7 *CCCGA_8 *CACAC_218  − 0.046 0.039 
cd1 Total dry weight C III 111.676 2.51 6.1 *CCCGA_8 *CACAC_218  − 0.025 0.022 
cd1 Root volume C III 114.726 2.5 6.8 *CCCGA_1 *CCCAC_22  − 0.014 0.013 
cd1 Root dry weight C III 110.267 2.66 6.6 *CCCGA_8 *CACAC_218  − 0.021 0.022 
cd2 Root dry weight C–Cd XVI 22.784 2.99 7.4 G_3141 *CCCCT_202R 0.045  − 0.030 
cd2 Total dry weight C–Cd XVI 26.609 2.81 6.8 *CCCCT_202R *TCCGT_8R 0.047  − 0.033 
cd2 Leaf dry weight C–Cd XVI 26.609 4.26 10 *CCCCT_202R *TCCGT_8R 0.056  − 0.042 
cd2 Total dry weight Cd XVI 26.609 3.14 7.8 *CCCCT_202R *TCCGT_8R  − 0.029 0.021 
cd2 Root dry weight Cd XVI 26.609 2.99 7.3 *CCCCT_202R *TCCGT_8R  − 0.027 0.017 
cd2 Leaf dry weight Cd XVI 26.609 4.6 11.6 *CCCCT_202R *TCCGT_8R  − 0.037 0.028 
cd3 Leaf dry weight C XVI 56.812 2.9 5.9 *CCCCT_246R *CTCAG_356R 0.033  − 0.016 
cd3 Total dry weight C–Cd XVI 56.812 3.61 7.4 *CCCCT_246R *CTCAG_356R 0.059  − 0.028 
cd3 Leaf dry weight C–Cd XVI 56.812 5.41 11 *CCCCT_246R *CTCAG_356R 0.072  − 0.036 
cd4 Leaf dry weight Cd XVI 68.558 4.61 11.3 *CGCCA_308R *CTCTC_100R  − 0.040 0.025 

1The relative effect of the allele originating from P. trichocarpa.

2The relative effect of the allele originating from P. deltoides.

3Difference in BLUP between control and Cd conditions (Cd effect).

Table 4.

Genetic (above diagonal) and phenotypic correlations among the variables measured.

Traits all correlations Total dry weight C Root dry weight C Leaf dry weight C Stem dry weight C Root volume C Total dry weight Cd Root dry weight Cd Leaf dry weight Cd Stem dry weight Cd Root volume Cd Total dry weight C–Cd Root dry weight C–Cd Leaf dry weight C–Cd Stem dry weight C–Cd Root volume C–Cd 
Total dry weight C 0.8838 0.9765 0.9463 0.7703 0.268 0.2562 0.2242 0.2822 0.1859 0.9197 0.7918 0.8783 0.8825 0.5893 
Root dry weight C 0.9022 0.796 0.8016 0.7862 0.2106 0.2084 0.1769 0.2149 0.1401 0.8235 0.9271 0.7186 0.757 0.6323 
Leaf dry weight C 0.973 0.8317 0.8841 0.7102 0.2648 0.2556 0.2259 0.2669 0.1816 0.8968 0.7036 0.901 0.8232 0.537 
Stem dry weight C 0.9431 0.8152 0.8895 0.7313 0.2674 0.2454 0.2148 0.3103 0.1941 0.8646 0.7131 0.7899 0.9278 0.5485 
Root volume C 0.6294 0.626 0.6059 0.5817 0.2165 0.205 0.1793 0.2338 0.1961 0.7043 0.7131 0.6315 0.6758 0.7931 
Total dry weight Cd  − 0.3557  − 0.3091  − 0.2795  − 0.2959  − 0.0793 0.8005 0.961 0.8843 0.5168  − 0.1319  − 0.0947  − 0.1627  − 0.0663  − 0.1229 
Root dry weight Cd  − 0.2765  − 0.3656  − 0.1976  − 0.2234  − 0.0563 0.8157 0.655 0.6965 0.4847  − 0.0628  − 0.1733  − 0.0356  − 0.0157  − 0.1135 
Leaf dry weight Cd  − 0.3188  − 0.2572  − 0.3007  − 0.2286  − 0.0624 0.9129 0.6886 0.7598 0.4666  − 0.161  − 0.0729  − 0.219  − 0.0727  − 0.1258 
Stem dry weight Cd  − 0.3146  − 0.2638  − 0.2208  − 0.356  − 0.0569 0.8559 0.721 0.689 0.4719  − 0.0701  − 0.0505  − 0.071  − 0.0667  − 0.0792 
Root volume Cd  − 0.079  − 0.1164  − 0.0457  − 0.0305  − 0.1963 0.4798 0.4724 0.4145 0.46  − 0.0194  − 0.0446  − 0.0259 0.0186  − 0.4417 
Total dry weight C–Cd 0.8291 0.7409 0.7669 0.7582 0.4363  − 0.8175  − 0.6584  − 0.7426  − 0.706  − 0.3353 0.8533 0.9699 0.935 0.6565 
Root dry weight C–Cd 0.7198 0.8329 0.6294 0.6347 0.4196  − 0.6752  − 0.8196  − 0.5678  − 0.591  − 0.3515 0.8475 0.7372 0.7683 0.6802 
Leaf dry weight C–Cd 0.7977 0.6723 0.803 0.69 0.412  − 0.7424  − 0.5519  − 0.8099  − 0.5664  − 0.2869 0.9358 0.742 0.8569 0.5939 
Stem dry weight C–Cd 0.7824 0.6716 0.6941 0.8424 0.405  − 0.682  − 0.558  − 0.5429  − 0.8034  − 0.2839 0.8902 0.7452 0.7662 0.6068 
Root volume C–Cd 0.474 0.4948 0.4376 0.4119 0.796  − 0.3495  − 0.3295  − 0.2978  − 0.3221  − 0.7498 0.5016 0.5003 0.4553 0.4488 
Traits all correlations Total dry weight C Root dry weight C Leaf dry weight C Stem dry weight C Root volume C Total dry weight Cd Root dry weight Cd Leaf dry weight Cd Stem dry weight Cd Root volume Cd Total dry weight C–Cd Root dry weight C–Cd Leaf dry weight C–Cd Stem dry weight C–Cd Root volume C–Cd 
Total dry weight C 0.8838 0.9765 0.9463 0.7703 0.268 0.2562 0.2242 0.2822 0.1859 0.9197 0.7918 0.8783 0.8825 0.5893 
Root dry weight C 0.9022 0.796 0.8016 0.7862 0.2106 0.2084 0.1769 0.2149 0.1401 0.8235 0.9271 0.7186 0.757 0.6323 
Leaf dry weight C 0.973 0.8317 0.8841 0.7102 0.2648 0.2556 0.2259 0.2669 0.1816 0.8968 0.7036 0.901 0.8232 0.537 
Stem dry weight C 0.9431 0.8152 0.8895 0.7313 0.2674 0.2454 0.2148 0.3103 0.1941 0.8646 0.7131 0.7899 0.9278 0.5485 
Root volume C 0.6294 0.626 0.6059 0.5817 0.2165 0.205 0.1793 0.2338 0.1961 0.7043 0.7131 0.6315 0.6758 0.7931 
Total dry weight Cd  − 0.3557  − 0.3091  − 0.2795  − 0.2959  − 0.0793 0.8005 0.961 0.8843 0.5168  − 0.1319  − 0.0947  − 0.1627  − 0.0663  − 0.1229 
Root dry weight Cd  − 0.2765  − 0.3656  − 0.1976  − 0.2234  − 0.0563 0.8157 0.655 0.6965 0.4847  − 0.0628  − 0.1733  − 0.0356  − 0.0157  − 0.1135 
Leaf dry weight Cd  − 0.3188  − 0.2572  − 0.3007  − 0.2286  − 0.0624 0.9129 0.6886 0.7598 0.4666  − 0.161  − 0.0729  − 0.219  − 0.0727  − 0.1258 
Stem dry weight Cd  − 0.3146  − 0.2638  − 0.2208  − 0.356  − 0.0569 0.8559 0.721 0.689 0.4719  − 0.0701  − 0.0505  − 0.071  − 0.0667  − 0.0792 
Root volume Cd  − 0.079  − 0.1164  − 0.0457  − 0.0305  − 0.1963 0.4798 0.4724 0.4145 0.46  − 0.0194  − 0.0446  − 0.0259 0.0186  − 0.4417 
Total dry weight C–Cd 0.8291 0.7409 0.7669 0.7582 0.4363  − 0.8175  − 0.6584  − 0.7426  − 0.706  − 0.3353 0.8533 0.9699 0.935 0.6565 
Root dry weight C–Cd 0.7198 0.8329 0.6294 0.6347 0.4196  − 0.6752  − 0.8196  − 0.5678  − 0.591  − 0.3515 0.8475 0.7372 0.7683 0.6802 
Leaf dry weight C–Cd 0.7977 0.6723 0.803 0.69 0.412  − 0.7424  − 0.5519  − 0.8099  − 0.5664  − 0.2869 0.9358 0.742 0.8569 0.5939 
Stem dry weight C–Cd 0.7824 0.6716 0.6941 0.8424 0.405  − 0.682  − 0.558  − 0.5429  − 0.8034  − 0.2839 0.8902 0.7452 0.7662 0.6068 
Root volume C–Cd 0.474 0.4948 0.4376 0.4119 0.796  − 0.3495  − 0.3295  − 0.2978  − 0.3221  − 0.7498 0.5016 0.5003 0.4553 0.4488 

Three QTL for dry weight and its components under control conditions were co-located on LG III along with three QTL for Cd effects on dry weight, collectively referred to as locus cd1 (Table 3). Two clusters of QTL were mapped to LG XVI at two discrete positions, the first of which (cd3) contained QTL for leaf dry weight under control conditions and Cd effects on total dry weight and its components (Table 3). For both cd1 and cd3, the allelic effects for dry weight measures under control conditions were in the same direction as those for QTL for Cd effects. However, the relative effect of the allele originating from P. trichocarpa was in opposite directions for these two loci, decreasing Cd tolerance at cd1 and enhancing Cd tolerance at cd3. The second cluster (cd2) on LG XVI contained QTL for total dry weight and its components under Cd conditions, as well as three QTL for Cd effects. In contrast to the other two loci, the allelic effects were opposite for dry weight measures and Cd effects. In this case, the allele from P. trichocarpa conferred reduced Cd tolerance on average. Finally, an apparently independent locus on LG XVI (cd4) explained 11% of the variation in leaf Cd concentration under Cd conditions (Table 3).

Candidate gene identification

The Cd response QTL intervals encompassed 1571 predicted genes from v2.0 of the P. trichocarpa genome annotation. In addition, across all genotypes and treatments there were 1748 different genes that showed up-regulation, and 672 genes that showed down-regulation in response to Cd treatment (see Figures S2 and S3 available as Supplementary Data at Tree Physiology Online). Nine of these differentially expressed genes were also present in the Cd response QTL intervals (Table 5).

Table 5.

Cd tolerance candidate genes in QTL intervals that also showed altered expression levels in the microarray experiment. Corresponding Arabidopsis genes were identified based on best reciprocal BLAST hits and the putative functions were derived from the TAIR 9 annotation. Fold change is the ratio of normalized expression levels under Cd versus control conditions. ns, not significant with a false discovery rate of 0.05.

Populus gene Arabidopsis gene QTL LG Gene symbol Fold change Pfp Clone Time (h) Putative function 
gw1.XVI.2928.1 AT1G70280.2 XVI None 12.3 3E-04 Both 24 NHL repeat-containing membrane protein 
fgenesh4_pg.C_LG_III001134 AT3G59140.1 III MRP14 4.7 0.003 183 72 Multidrug resistance-associated protein; ATP-type transporter 
gw1.XVI.2910.1 AT2G28305.1 XVI LOG1 3.3 0.013 182 72 Unknown function, expressed during 4 leaf senescence stage 
eugene3.00031003 AT4G11450.1 III F25E4.70 3.1 0.021 183 24,72 Unknown function, expressed during leaf senescence 
estExt_fgenesh4_pg.C_LG_III1085 AT4G23496.1 III SP1L5 0.32 0.016 182 72 Unknown function; potentially involved in anisotropic cell expansion 
estExt_fgenesh4_pg.C_LG_III1187 AT4G24015.1 III None 0.27 0.041 182 24 Zinc finger/unknown function 
fgenesh4_pg.C_LG_III001060 No Hits III None 0.27 0.028 182 24 None 
gw1.XVI.2786.1 AT5G03150.1 XVI JKD 0.24 0.017 182 24,72 Zinc finger transcription factor; root development, regulation of cell division, cell differentiation, meristem growth 
fgenesh4_pm.C_LG_XVI000330 AT2G37220.1 XVI F3G5.1 0.13 0.001 182 24 Chloroplast RNA binding protein, response to ABA stimulus, response to cold 
Populus gene Arabidopsis gene QTL LG Gene symbol Fold change Pfp Clone Time (h) Putative function 
gw1.XVI.2928.1 AT1G70280.2 XVI None 12.3 3E-04 Both 24 NHL repeat-containing membrane protein 
fgenesh4_pg.C_LG_III001134 AT3G59140.1 III MRP14 4.7 0.003 183 72 Multidrug resistance-associated protein; ATP-type transporter 
gw1.XVI.2910.1 AT2G28305.1 XVI LOG1 3.3 0.013 182 72 Unknown function, expressed during 4 leaf senescence stage 
eugene3.00031003 AT4G11450.1 III F25E4.70 3.1 0.021 183 24,72 Unknown function, expressed during leaf senescence 
estExt_fgenesh4_pg.C_LG_III1085 AT4G23496.1 III SP1L5 0.32 0.016 182 72 Unknown function; potentially involved in anisotropic cell expansion 
estExt_fgenesh4_pg.C_LG_III1187 AT4G24015.1 III None 0.27 0.041 182 24 Zinc finger/unknown function 
fgenesh4_pg.C_LG_III001060 No Hits III None 0.27 0.028 182 24 None 
gw1.XVI.2786.1 AT5G03150.1 XVI JKD 0.24 0.017 182 24,72 Zinc finger transcription factor; root development, regulation of cell division, cell differentiation, meristem growth 
fgenesh4_pm.C_LG_XVI000330 AT2G37220.1 XVI F3G5.1 0.13 0.001 182 24 Chloroplast RNA binding protein, response to ABA stimulus, response to cold 

Discussion

We observed substantial phenotypic variation in Cd effects on plant biomass and leaf Cd accumulation in Family 52-124, a pseudo-backcross pedigree derived from P. trichocarpa × deltoides hybrids. The effects of genotype, Cd treatment and their interactions were significant for all response variables measured in this study, including total dry weight, root dry weight, leaf dry weight, stem dry weight and root volume. Substantial variation in Cd tolerance is commonly observed within and among species. For example, Pietrini et al. (2010) used 50 µM of Cd (i.e., twice as high as in our experiment) to compare responses of multiple hybrid Populus clones involving six species. They observed significant variation among clones in the effects of Cd on biomass accumulation, photosynthetic responses, Cd content of plant parts and concentration of photosynthetic pigments. Interestingly, the variation in Cd responses observed among full siblings in our inter-specific cross is of magnitude similar to the variation observed among highly diverse clones that were exposed to a higher concentration of Cd. This illustrates the tremendous phenotypic variation captured by our crossing design and experimental conditions. The Cd-tolerant and Cd-susceptible genotypes identified in this study are a useful resource for future experiments targeted at physiological and molecular dissection of Cd tolerance.

Heritability estimates

Our broad-sense heritability estimate for total dry weight was somewhat lower than those for growth and yield traits measured for Populus in recent field studies (Marron et al. 2006, 2010, Rae et al. 2008, Dillen et al. 2009), though certainly within the range of estimates in earlier studies (Riemenschneider et al. 1996). The broad-sense heritabilities calculated for total dry weight and its components under control conditions were not substantially different from the heritabilities calculated for these traits under Cd exposure. Root volume had the lowest heritabilities under both conditions, whereas root dry weight had heritabilities that were comparable to those for other dry weight components. This discrepancy may reflect difficulties in accurately measuring root volume through water displacement.

Quantitative trait loci

Although we detected 16 QTL in this study, these occurred at only four different chromosomal locations. Co-localization of QTL for total dry weight and dry weight components was not surprising, given that these traits were highly intercorrelated and were co-segregating in the pedigree. However, it is potentially more interesting and informative that dry weight under control and Cd conditions mapped to different QTL, and had opposing relationships to the Cd response QTL (C–Cd dry weight). Dry weight under control conditions was positively correlated with Cd effect and allelic effects were in the same direction for these traits when they were co-located, suggesting that these QTL were driven by the size of plants under control conditions (i.e., large plants showed more sensitivity to the Cd treatment). In contrast, dry weight under Cd conditions was negatively correlated with Cd effect, and allelic effects were opposite for these traits at co-located QTL, suggesting that Cd tolerance at this locus was associated with growth in the presence of Cd irrespective of performance under control conditions.

Quantitative trait loci for different traits may also be co-located even when they are not correlated (Wu et al. 1997, Zhang et al. 2006). For example, in this study leaf dry weight Cd and leaf dry weight C–Cd were mapped to the same position on LG XVI, but were not significantly correlated. The same was observed between root dry weight Cd and root dry weight C–Cd and between total dry weight Cd and total dry weight C–Cd (Tables 3 and 4). This could be due to a single gene or a regulatory element having pleiotropic effects on these traits and/or different genes within the QTL interval that might be independently controlling these traits, as one QTL encompasses hundreds of candidate genes (Wullschleger et al. 2005, Novaes et al. 2009). Identification of different QTLs on different linkage groups under control and Cd conditions for total dry weight and root dry weight suggested that the genes governing those traits were differentially controlled (Rae et al. 2006), and that the control exerted by these QTLs or genomic regions was dependent on Cd exposure. Finally, it is important to note that QTLs identified in this study account for a relatively small proportion of the total variation in Cd responses. Furthermore, since the population used for the current study was relatively small, QTL effects were probably overestimated (Beavis 1998), so much work remains to identify further mechanisms of Cd tolerance in Populus.

Candidate genes from QTL and microarray analyses

Transcriptional responses to experimental stimuli are one possible indicator of gene function. We therefore used microarray analyses to gain further insights into the potential roles of the many candidate genes within the QTL intervals. Of the 1571 genes contained within the Cd response QTL intervals, only 12 showed differential expression under Cd treatment. Among these were homologs of AT3G59140.1 (also known as multidrug resistance-associated protein 14), which were significantly overexpressed in the Cd-tolerant genotype (1–183) relative to the control after 72 h of treatment. Two closely related sequences matching this gene occur in the Cd tolerance QTL on LG III, presumably the result of recent tandem duplication. This gene belongs to the MRP (multidrug resistance-associated protein) subfamily of the superfamily of ATP binding cassette (ABC) transporters, which has been implicated in Cd sequestration (Klein et al. 2006). ATP binding cassette transporters facilitate the translocation of chelates including glutathione with Cd and Pb by ATP-driven processes (Lu et al. 1998, Tommasini et al. 1998, Kolukisaoglu et al. 2002, Martinoia et al. 2002, Klein et al. 2003). Interestingly, expressions of AtMRPs 3, 4, 6, 7 and 14 were up-regulated by Cd in Arabidopsis (Kolukisaoglu et al. 2002, Bovet et al. 2003, Kim et al. 2007). Wojas et al. (2009) demonstrated the role of AtMRP7 in Cd tolerance and vacuolar sequestration in tobacco. Many yeast mutant complementation studies have demonstrated the role of AtMRPs in Cd tolerance, and these MRPs could be used as targets of genetic engineering to increase Cd tolerance and accumulation (Wojas et al. 2009). This MRP is therefore an excellent candidate for further functional characterization in Populus.

The remaining genes with differential expression in the QTL intervals have no known role in Cd homeostasis or transport. Four of these genes showed significant up-regulation, and five showed down-regulation (Table 6). Six of these have no known function for their closest homologs in model plant species. Interestingly, the gene that was most strongly up-regulated (12.3-fold) in response to 24-h Cd treatment in both genotypes encodes a putative membrane protein. The predicted protein sequence contains an NHL-repeat domain, which is a six-bladed propeller structure characteristic of a large number of proteins across all major kingdoms (Chaudhuri et al. 2008), including 35 predicted proteins in Populus and 30 in Arabidopsis. For example, the beta propeller is a prominent feature of receptor proteins like TolB, which is involved in antibiotic translocation in Escherichia coli (Kleanthous 2010) and of phosphate-chelating phytases found in a wide range of bacteria and fungi (Lung et al. 2005). The vast majority of plant proteins containing this domain have no known function, although some may be involved in alkaloid biosynthesis (Ma et al. 2006). Although no known connection with Cd tolerance has been demonstrated, it is intriguing that this protein is localized in cell membranes and is known to function in pore formation in microbes. Functional characterization could well lead to the identification of a novel metal tolerance mechanism in Populus and other plants.

Table 6.

Candidate genes identified from the QTL intervals with no detected change in expression in response to Cd treatment.

Populus gene Arabidopsis gene QTL LG Gene symbol Annotation 
estExt_fgenesh4_pm.C_LG_III0405 AT4G23100.1 III GSH1, CAD2 Glutamate-cysteine ligase 
estExt_Genewise1_v1.C_LG_III2138 AT1G63440.1 III HMA5 Heavy metal ATPase 
fgenesh4_pm.C_LG_III000450 AT5G41610.1 III CHX18 Monovalent cation:proton antiporter 
estExt_fgenesh4_pg.C_LG_III1026 AT5G41210.1 III GST10 Glutathione-S-transferase 
fgenesh4_kg.C_LG_III000037 AT4G11600.1 III GPX Glutathione peroxidase 
Populus gene Arabidopsis gene QTL LG Gene symbol Annotation 
estExt_fgenesh4_pm.C_LG_III0405 AT4G23100.1 III GSH1, CAD2 Glutamate-cysteine ligase 
estExt_Genewise1_v1.C_LG_III2138 AT1G63440.1 III HMA5 Heavy metal ATPase 
fgenesh4_pm.C_LG_III000450 AT5G41610.1 III CHX18 Monovalent cation:proton antiporter 
estExt_fgenesh4_pg.C_LG_III1026 AT5G41210.1 III GST10 Glutathione-S-transferase 
fgenesh4_kg.C_LG_III000037 AT4G11600.1 III GPX Glutathione peroxidase 

Candidate genes from the QTL intervals

It is quite possible that genes responsible for interspecific differences in Cd tolerance would not show transcriptional responses that would be detectable in our experiments. In fact, some candidate genes with potential roles in Cd tolerance were present in the QTL intervals, but did not show differential expression in the microarray experiment. These include CAD2 (cadmium sensitive 2), HMA5 (heavy metal ATPase5), CHX18 (cation/hydrogen exchanger 18), GST10 (glutathione-S-transferase1) and GPX6 (glutathione peroxidase 6) (Table 6).

Phytochelatins (PCs) are one of the major chelators produced by plants upon exposure to heavy metals. These peptides chelate heavy metals and facilitate their sequestration in vacuoles, thus limiting heavy metal toxicity (Koprivova et al. 2002). Phytochelatins are synthesized from glutathione by phytochelatin synthase in the presence of heavy metal ions (Grill et al. 1986, Rauser 1999). Phytochelatins are induced most strongly by Cd, and plants without PCs showed hypersensitivity toward Cd (Howden et al. 1995b) and other heavy metals. The Cd–PC complexes are stored primarily in the root vacuoles, although translocation to shoots also occurs (Kim et al. 2007, Saathoff et al. 2011). The transporter responsible for vacuolar sequestration of Cd–PC complexes is yet to be definitively identified in plants, but ABC transporters like the MRPs have been implicated in plants (Klein et al. 2006, Wojas et al. 2009) and yeast (Mendoza-Cozatl et al. 2010). Cadmium sensitive 2 is a γ-glutamylcysteine synthetase, which is involved in glutathione biosynthesis, a precursor for PCs (Howden et al. 1995a). In Arabidopsis, cad2-1 seedlings were more sensitive to Cd than wild-type plants, presumably because the mutants were unable to synthesize PCs (Howden et al. 1995a, 1995b). In this study, the presence of CAD2 in the Cd tolerance QTL on LG III suggests a role for this gene in Cd tolerance in Populus.

Populus contains three homologs of HMA5 (heavy metal ATPase5) (Migeon et al. 2010), including the one present within the LG III QTL. Some members of this family have been shown to transport Cd and enhance Cd tolerance when overexpressed. For example, HMA 1 through 4 transport Zn/Cd/Pb/Co in Arabidopsis and other plants (Cobbett et al. 2003). AtHMA4 enhanced Cd and Zn tolerance, transport, accumulation and root–shoot partitioning when overexpressed in tobacco (Siemianowski et al. 2011). Furthermore, overexpression of HMA3 enhanced Cd accumulation and tolerance in rice (Oryza sativa L.) (Miyadate et al. 2010). Although HMA5 is not known to transport Cd (Andres-Colas et al. 2006), it is possible that the specificities of these transporters are different in Populus, especially since the expression profiles of the putative homologs differ markedly between Populus and Arabidopsis (Migeon et al. 2010). This gene is therefore a viable candidate for functional characterization for Cd transport in Populus.

Another gene from the QTL interval on LG III has homology to CHX18, a cation transporter that belongs to a group of vacuolar membrane proteins in the cation exchanger (CAX) subfamily of proteins. These transport cations utilizing an H + /ATPase pump that generates proton gradients across the vacuolar membrane, and are involved in calcium transport and metal homeostasis (Hirschi et al. 2000, Mei et al. 2009, Migeon et al. 2010). Cation exchangers have a broad range of metal substrates in plants, including significant roles in transporting Cd in roots and shoots (Korenkov et al. 2009). Overexpression of CAX genes enhances Cd tolerance and accumulation in Arabidopsis (Hirschi et al. 2000, Mei et al. 2009, Migeon et al. 2010), tobacco (Korenkov et al. 2007) and petunia (Wu et al. 2011).

A third gene from the QTL interval on LG III has homology to GST10, which belongs to the theta group of the superfamily of glutathione-S-transferases (GSTs). Glutathione-S-transferases are broadly important in protection against and detoxification of harmful xenobiotics, including heavy metals (Dixon et al. 2005, Brentner et al. 2008). GPX6 (glutathione peroxidase), together with reduced glutathione, reduces hydrogen peroxide and other organic peroxides to water. AtGPX6 was up-regulated under Cd exposure in Arabidopsis (Sarry et al. 2006), and is strongly induced under a wide range of abiotic stresses (Milla et al. 2003). Thus, homologs of AtGSTT1 and AtGPX6 might contribute to Cd tolerance in Populus through glutathione-mediated oxidative stress tolerance.

Conclusion

This study is the first to map QTL loci for Cd tolerance in the commercially and ecologically important genus Populus. This is an important initial step toward enhancing Cd tolerance in this genus for phytoremediation applications, either through future marker-assisted selection strategies, or through genetic engineering. We refined the list of potential candidate genes from the QTL intervals using whole-genome microarray analysis. This revealed putative metal transporters and genes of unknown function that can now be further characterized for their role in Cd transport and vacuolar sequestration. Further work should be geared toward establishing finer-scale associations for these genes by assaying polymorphisms in association populations within P. trichocarpa and P. deltoides, and by assaying informative candidate polymorphisms that differentiate these species. Furthermore, expression data should be gathered for a larger number of genotypes, and QTLs should be validated in different genetic and environmental backgrounds under both greenhouse and field conditions. Ultimately, this research could lead to the development of heavy metal-tolerant and/or hyperaccumulating Populus clones for reclaiming and remediating the extensive marginal lands contaminated with heavy metals.

Supplementary data

Supplementary data for this article are available at Tree Physiology Online.

Funding

Portions of this work were supported by the Department of Energy Bioenergy Science Center under Contract Number DE-AC05-00OR22725, and support from the Eberly College of Arts and Sciences at West Virginia University and the WVU Research Corporation. This was a key project for Jiangsu Universities (10KJA180018).

Conflict of interest

None declared.

Acknowledgments

We greatly appreciate logistical and intellectual input from Carina Barth, Alex Lastinger, Chase Kempinski, Ryan Imperio, Jonathan Cumming, Dorothy Vesper and Louis McDonald.

References

Almeida
A.-A.F.d.
Valle
R.l.R.
Mielke
M.S.
Gomes
F.b.P.
.
2007
.
Tolerance and prospection of phytoremediator woody species of Cd, Pb, Cu and Cr. Braz. J
.
Plant Physiol. 19
 :
83
98
.
Andres-Colas
N.
Sancenon
V.
Rodriguez-Navarro
S.
Mayo
S.
Thiele
D.J.
Ecker
J.R.
Puig
S.
Penarrubia
L.
.
2006
.
The Arabidopsis heavy metal P-type ATPase HMA5 interacts with metallochaperones and functions in copper detoxification of roots
.
Plant J. 45
 :
225
236
.
Beavis
W.D.
1998
.
QTL analyses: power, precision, and accuracy. In Molecular Dissection of Complex Traits
.
Paterson
A.H.
.
CRC Press
,
Boca Raton, FL
, pp
145
162
.
Bolstad
B.M.
Irizarry
R.A.
Gautier
L.
Wu
Z.
.
2005
.
Preprocessing high-density oligonucleotide arrays. In Bioinformatics and Computational Biology Solutions Using R and Bioconductor
.
Gentleman
R.
Carey
V.J.
Huber
W.
Irizarry
R.A.
Dudoit
S.
.
Springer
,
New York
, pp
13
32
.
Bovet
L.
Eggmann
T.
Meylan-Bettex
M.
Polier
J.
Kammer
P.
Marin
E.
Feller
U.
Martinoia
E.
.
2003
.
Transcript levels of AtMRPs after cadmium treatment: induction of AtMRP3
.
Plant Cell Environ
 .
26
:
371
381
.
Breitling
R.
Armengaud
P.
Amtmann
A.
Herzvk
P.
.
2004
.
Rank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments
.
FEBS Lett
 .
573
:
83
92
.
Brentner
L.B.
Mukherji
S.T.
Merchie
K.M.
Yoon
J.M.
Schnoor
J.L.
Van Aken
B.
.
2008
.
Expression of glutathione S-transferases in poplar trees (Populus trichocarpa) exposed to 2,4,6-trinitrotoluene (TNT). Chemosphere 73
:
657
662
.
Burken
J.G.
Schnoor
J.L.
.
1997
.
Uptake and metabolism of atrazine by poplar trees
.
Environ. Sci. Technol
 .
31
:
1399
1406
.
Chaudhuri
I.
Soding
J.
Lupas
A.N.
.
2008
.
Evolution of the beta-propeller fold
.
Proteins Struct. Funct. Bioinformat
 .
71
:
795
803
.
Cobbett
C.S.
Hussain
D.
Haydon
M.J.
.
2003
.
Structural and functional relationships between type 1(B) heavy metal-transporting P-type ATPases in Arabidopsis
.
New Phytol
 .
159
:
315
321
.
Coles
N.D.
McMullen
M.D.
Balint-Kurti
P.J.
Pratt
R.C.
Holland
J.B.
.
2010
.
Genetic control of photoperiod sensitivity in maize revealed by joint multiple population analysis
.
Genetics
 
184
:
799
U301
.
Cunningham
S.D.
Ow
D.W.
.
1996
.
Promises and prospects of phytoremediation
.
Plant Physiol
 .
110
:
715
719
.
Di Baccio
D.
Tognetti
R.
Minnocci
A.
Sebastiani
L.
.
2009
.
Responses of the Populus × euramericana clone I-214 to excess zinc: carbon assimilation, structural modifications, metal distribution and cellular localization
.
Environ. Exp. Bot
 .
67
:
153
163
.
Dillen
S.Y.
Marron
N.
Sabatti
M.
Ceulemans
R.
Bastien
C.
.
2009
.
Relationships among productivity determinants in two hybrid poplar families grown during three years at two contrasting sites
.
Tree Physiol
 .
29
:
975
987
.
Dixon
D.P.
Lapthorn
A.J.
Edwards
R.
.
2005
.
Synthesis and analysis of chimeric maize glutathione transferases
.
Plant Sci
 .
168
:
873
881
.
Eapen
S.
D'Souza
S.F.
.
2005
.
Prospects of genetic engineering of plants for phytoremediation of toxic metals
.
Biotechnol. Adv
 .
23
:
97
114
.
Fodor
F.
Gaspar
L.
Morales
F.
Gogorcena
Y.
Lucena
J.J.
Cseh
E.
Kropfl
K.
Abadia
J.
Sarvari
E.
.
2005
.
Effects of two iron sources on iron and cadmium allocation in poplar (Populus alba) plants exposed to cadmium
.
Tree Physiol
 .
25
:
1173
1180
.
Grill
E.
Gekeler
W.
Winnacker
E.L.
Zenk
H.H.
.
1986
.
Homo-phytochelatins are heavy metal-binding peptides of homo-glutathione containing Fabales
.
FEBS Lett
 .
205
:
47
50
.
Hirschi
K.D.
Korenkov
V.D.
Wilganowski
N.L.
Wagner
G.J.
.
2000
.
Expression of Arabidopsis CAX2 in tobacco. Altered metal accumulation and increased manganese tolerance
.
Plant Physiol
 .
124
:
125
133
.
Howden
R.
Andersen
C.R.
Goldsbrough
P.B.
Cobbett
C.S.
.
1995
a.
A cadmium-sensitive, glutathione-deficient mutant of Arabidopsis thaliana
.
Plant Physiol
 .
107
:
1067
1073
.
Howden
R.
Andersen
C.R.
Goldsbrough
P.B.
Cobbett
C.S.
.
1995
b.
Cadmium-sensitive, cad1 mutants of Arabidopsis thaliana are phytochelatin deficient
.
Plant Physiol
 .
107
:
1059
1066
.
Kim
D.Y.
Bovet
L.
Maeshima
M.
Martinoia
E.
Lee
Y.
.
2007
.
The ABC transporter AtPDR8 is a cadmium extrusion pump conferring heavy metal resistance
.
Plant J
 .
50
:
207
218
.
Kleanthous
C.
2010
.
Swimming against the tide: progress and challenges in our understanding of colicin translocation
.
Nat. Rev. Microbiol
 .
8
:
843
848
.
Klein
M.
Perfus-Barbeoch
L.
Frelet
A.
Gaedeke
N.
Reinhardt
D.
Mueller-Roeber
B.
Martinoia
E.
Forestier
C.
.
2003
.
The plant multidrug resistance ABC transporter AtMRP5 is involved in guard cell hormonal signalling and water use
.
Plant J
 .
33
:
119
129
.
Klein
M.
Burla
B.
Martinoia
E.
.
2006
.
The multidrug resistance-associated protein (MRP/ABCC) subfamily of ATP-binding cassette transporters in plants
.
FEBS Lett
 .
580
:
1112
1122
.
Kolukisaoglu
H.U.
Bovet
L.
Klein
M.
Eggmann
T.
Geisler
M.
Wanke
D.
Martinoia
E.
Schulz
B.
.
2002
.
Family business: the multidrug-resistance related protein (MRP) ABC transporter genes in Arabidopsis thaliana
.
Planta
 
216
:
107
119
.
Koprivova
A.
Kopriva
S.
Jager
D.
Will
B.
Jouanin
L.
Rennenberg
H.
.
2002
.
Evaluation of transgenic poplars over-expressing enzymes of glutathione synthesis for phytoremediation of cadmium
.
Plant Biol
 .
4
:
664
670
.
Korenkov
V.
Hirschi
K.
Crutchfield
J.D.
Wagner
G.J.
.
2007
.
Enhancing tonoplast Cd/H antiport activity increases Cd, Zn, and Mn tolerance, and impacts root/shoot Cd partitioning in Nicotiana tabacum L
.
Planta
 
226
:
1379
1387
.
Korenkov
V.
King
B.
Hirschi
K.
Wagner
G.J.
.
2009
.
Root-selective expression of AtCAX4 and AtCAX2 results in reduced lamina cadmium in field-grown Nicotiana tabacum L
.
Plant Biotechnol. J
 .
7
:
219
226
.
Lu
Y.P.
Li
Z.S.
Drozdowicz
Y.M.
Hortensteiner
S.
Martinoia
E.
Rea
P.A.
.
1998
.
AtMRP2, an Arabidopsis ATP binding cassette transporter able to transport glutathione S-conjugates and chlorophyll catabolites: functional comparisons with AtMRP1
.
Plant Cell
 
10
:
267
282
.
Lung
S.C.
Chan
W.L.
Yip
W.
Wang
L.J.
Yeung
E.C.
Lim
B.L.
.
2005
.
Secretion of beta-propeller phytase from tobacco and Arabidopsis roots enhances phosphorus utilization
.
Plant Sci
 .
169
:
341
349
.
Ma
X.
Panjikar
S.
Koepke
J.
Loris
E.
Stöckigt
J.
.
2006
.
The structure of Rauvolfia serpentina strictosidine synthase is a novel six-bladed β-propeller fold in plant proteins
.
Plant Cell
 
18
:
907
920
.
Marron
N.
Bastien
C.
Sabatti
M.
Taylor
G.
Ceulemans
R.
.
2006
.
Plasticity of growth and sylleptic branchiness in two poplar families grown at three sites across Europe
.
Tree Physiol
 .
26
:
935
946
.
Marron
N.
Ricciotti
L.
Bastien
C.
et al
.
2010
.
Plasticity of growth and biomass production of an intraspecific Populus alba family grown at three sites across Europe during three growing seasons. Can
.
J. For. Res
 .
40
:
1887
1903
.
Martinoia
E.
Klein
M.
Geisler
M.
Bovet
L.
Forestier
C.
Kolukisaoglu
U.
Muller-Rober
B.
Schulz
B.
.
2002
.
Multifunctionality of plant ABC transporters—more than just detoxifiers
.
Planta
 
214
:
345
355
.
Mei
H.
Cheng
N.H.
Zhao
J.
Park
S.
Escareno
R.A.
Pittman
J.K.
Hirschi
K.D.
.
2009
.
Root development under metal stress in Arabidopsis thaliana requires the H + /cation antiporter CAX4
.
New Phytol
 .
183
:
95
105
.
Mendoza-Cozatl
D.G.
Zhai
Z.Y.
Jobe
T.O.
et al
.
2010
.
Tonoplast-localized Abc2 transporter mediates phytochelatin accumulation in vacuoles and confers cadmium tolerance
.
J. Biol. Chem
 .
285
:
40416
40426
.
Migeon
A.
Richaud
P.
Guinet
F.
Chalot
M.
Blaudez
D.
.
2009
.
Metal accumulation by woody species on contaminated sites in the north of France
.
Water Air Soil Pollut
 .
204
:
89
101
.
Migeon
A.
Blaudez
D.
Wilkins
O.
Montanini
B.
Campbell
M.M.
Richaud
P.
Thomine
S.
Chalot
M.
.
2010
.
Genome-wide analysis of plant metal transporters, with an emphasis on poplar
.
Cell. Mol. Life Sci
 .
67
:
3763
3784
.
Milla
M.A.R.
Maurer
A.
Huete
A.R.
Gustafson
J.P.
.
2003
.
Glutathione peroxidase genes in Arabidopsis are ubiquitous and regulated by abiotic stresses through diverse signaling pathways
.
Plant J
 .
36
:
602
615
.
Milone
M.T.
Sgherri
C.
Clijsters
H.
Navari-Izzo
F.
.
2003
.
Antioxidative responses of wheat treated with realistic concentration of cadmium
.
Environ. Exp. Bot
 .
50
:
265
276
.
Miyadate
H.
Adachi
S.
Hiraizumi
A.
et al
.
2010
.
OsHMA3, a P-1B-type of ATPase affects root-to-shoot cadmium translocation in rice by mediating efflux into vacuoles
.
New Phytol
 .
189
:
190
199
.
Newman
L.A.
Strand
S.E.
Choe
N.
et al
.
1997
.
Uptake and biotransformation of trichloroethylene by hybrid poplars
.
Environ. Sci. Technol
 .
31
:
1062
1067
.
Novaes
E.
Osorio
L.
Drost
D.R.
et al
.
2009
.
Quantitative genetic analysis of biomass and wood chemistry of Populus under different nitrogen levels
.
New Phytol
 .
182
:
878
890
.
Ouariti
O.
Boussama
N.
Zarrouk
M.
Cherif
A.
Ghorbal
M.H.
.
1997
.
Cadmium- and copper-induced changes in tomato membrane lipids
.
Phytochemistry 45
:
1343
1350
.
Pietrini
F.
Zacchini
M.
Iori
V.
Pietrosanti
L.
Bianconi
D.
Massacci
A.
.
2010
.
Screening of poplar clones for cadmium phytoremediation using photosynthesis, biomass and cadmium content analyses
.
Int. J. Phytoremediat
 .
12
:
105
120
.
Pilon-Smits
E.A.H.
de Souza
M.P.
Lytle
C.M.
Shang
C.
Lugo
T.
Terry
N.
.
1998
.
Selenium volatilization and assimilation by hybrid poplar (Populus tremula × alba)
.
J. Exp. Bot
 .
49
:
1889
1892
.
Rae
A.M.
Ferris
R.
Tallis
M.J.
Taylor
G.
.
2006
.
Elucidating genomic regions determining enhanced leaf growth and delayed senescence in elevated CO2
.
Plant Cell Environ
 .
29
:
1730
1741
.
Rae
A.M.
Pinel
M.P.C.
Bastien
C.
et al
.
2008
.
QTL for yield in bioenergy Populus: identifying GxE interactions from growth at three contrasting sites
.
Tree Genet. Genomes
 
4
:
97
112
.
Rauser
W.E.
1999
.
Structure and function of metal chelators produced by plants—the case for organic acids, amino acids, phytin, and metallothioneins
.
Cell Biochem. Biophys
 .
31
:
19
48
.
Riemenschneider
D.E.
Stelzer
H.E.
Foster
G.S.
.
1996
.
Quantitative genetics of poplars and poplar hybrids. In Biology of Populus and its Implications for Management and
Conservation
 .
Stettler
R.F.
Bradshaw
H.D.
Jr
Heilman
P.E.
Hinckley
T.M.
.
NRC Research Press
,
Ottawa, ON
, pp
159
181
.
Robinson
B.H.
Mills
T.M.
Petit
D.
Fung
L.E.
Green
S.R.
Clothier
B.E.
.
2000
.
Natural and induced cadmium-accumulation in poplar and willow: implications for phytoremediation
.
Plant Soil
 
227
:
301
306
.
Rockwood
D.L.
Naidu
C.V.
Carter
D.R.
Rahmani
M.
Spriggs
T.A.
Lin
C.
Alker
G.R.
Isebrands
J.G.
Segrest
S.A.
.
2004
.
Short-rotation woody crops and phytoremediation: opportunities for agroforestry?
Agrofor. Syst
 .
61–2
:
51
63
.
Rodriguez-Serrano
M.
Romero-Puertas
M.C.
Pazmino
D.M.
Testillano
P.S.
Risueno
M.C.
del Rio
L.A.
Sandalio
L.M.
.
2009
.
Cellular response of pea plants to cadmium toxicity: cross talk between reactive oxygen species, nitric oxide, and calcium
.
Plant Physiol
 .
150
:
229
243
.
Saathoff
A.J.
Ahner
B.
Spanswick
R.M.
Walker
L.P.
.
2011
.
Detection of phytochelatin in the xylem sap of Brassica napus
.
Environ. Eng. Sci
 .
28
:
103
111
.
Sandalio
L.M.
Dalurzo
H.C.
Gomez
M.
Romero-Puertas
M.C.
del Rio
L.A.
.
2001
.
Cadmium-induced changes in the growth and oxidative metabolism of pea plants
.
J. Exp. Bot
 .
52
:
2115
2126
.
Sarry
J.E.
Kuhn
L.
Ducruix
C.
et al
.
2006
.
The early responses of Arabidopsis thaliana cells to cadmium exposure explored by protein and metabolite profiling analyses
.
Proteomics
 
6
:
2180
2198
.
Shao
G.S.
Chen
M.X.
Wang
D.Y.
Xu
C.M.
Mou
R.X.
Cao
Z.Y.
Zhang
X.F.
.
2008
.
Using iron fertilizer to control Cd accumulation in rice plants: a new promising technology
.
Sci. China Ser. C Life Sci
 .
51
:
245
253
.
Shapiro
S.S.
Wilk
M.B.
.
1965
.
An analysis of variance test for normality (complete samples)
.
Biometrika
 
52
:
591
611
.
Siddique
K.H.M.
Tennant
D.
Perry
M.W.
Belford
R.K.
.
1990
.
Water-use and water-use efficiency of old and modern wheat cultivars in a Mediterranean-type environment. Aust
.
J. Agric. Res
 .
41
:
431
447
.
Siedlecka
A.
Krupa
Z.
.
1999
.
Cd/Fe interaction in higher plants - its consequences for the photosynthetic apparatus
.
Photosynthetica
 
36
:
321
331
.
Siemianowski
O.
Mills
R.F.
Williams
L.E.
Antosiewicz
D.M.
.
2011
.
Expression of the P-1B-type ATPase AtHMA4 in tobacco modifies Zn and Cd root to shoot partitioning and metal tolerance
.
Plant Biotechnol. J
 .
9
:
64
74
.
Smeets
K.
Cuypers
A.
Lambrechts
A.
Semane
B.
Hoet
P.
Van Laere
A.
Vangronsveld
J.
.
2005
.
Induction of oxidative stress and antioxidative mechanisms in Phaseolus vulgaris after Cd application
.
Plant Physiol. Biochem
 .
43
:
437
444
.
Solti
A.
Gaspar
L.
Meszaros
I.
Szigeti
Z.
Levai
L.
Sarvari
E.
.
2008
.
Impact of iron supply on the kinetics of recovery of photosynthesis in Cd-stressed poplar (Populus glauca)
.
Ann. Bot
 .
102
:
771
782
.
Tommasini
R.
Vogt
E.
Fromenteau
M.
Hortensteiner
S.
Matile
P.
Amrhein
N.
Martinoia
E.
.
1998
.
An ABC-transporter of Arabidopsis thaliana has both glutathione-conjugate and chlorophyll catabolite transport activity
.
Plant J
 .
13
:
773
780
.
Tuskan
G.A.
Gunter
L.E.
Yang
Z.M.K.
Yin
T.M.
Sewell
M.M.
DiFazio
S.P.
.
2004
.
Characterization of microsatellites revealed by genomic sequencing of Populus trichocarpa. Can
.
J. For. Res
 .
34
:
85
93
.
Van Ooijen
J.
2004
.
MapQTL 5, software for the mapping of quantitative trait loci in experimental populations. Kyazma B.V
.
Wageningen
,
Netherlands
.
Van Ooijen
J.
Voorrips
R.E.
.
2001
.
JoinMap 3.0, software for the calculation of genetic linkage maps
.
Plant Research International, Wageningen
,
Netherlands
.
Vanassche
F.
Clijsters
H.
.
1990
.
Effects of metals on enzyme-activity in plants
.
Plant Cell Environ
 .
13
:
195
206
.
Wojas
S.
Hennig
J.
Plaza
S.
Geisler
M.
Siemianowski
O.
Sklodowska
A.
Ruszczynska
A.
Bulska
E.
Antosiewicz
D.M.
.
2009
.
Ectopic expression of Arabidopsis ABC transporter MRP7 modifies cadmium root-to-shoot transport and accumulation
.
Environ. Pollut
 .
157
:
2781
2789
.
Wu
R.
Bradshaw
H.D.
Stettler
R.F.
.
1997
.
Molecular genetics of growth and development in Populus (Salicaceae). 5. Mapping quantitative trait loci affecting leaf variation
.
Am. J. Bot
 .
84
:
143
153
.
Wu
Q.
Shigaki
T.
Williams
K.A.
Han
J.S.
Kim
C.K.
Hirschi
K.D.
Park
S.
.
2011
.
Expression of an Arabidopsis Ca2 + /H+ antiporter CAX1 variant in petunia enhances cadmium tolerance and accumulation
.
J. Plant Physiol
 .
168
:
167
173
.
Wullschleger
S.
Yin
T.M.
DiFazio
S.P.
Tschaplinski
T.J.
Gunter
L.E.
Davis
M.F.
Tuskan
G.A.
.
2005
.
Phenotypic variation in growth and biomass distribution for two advanced-generation pedigrees of hybrid poplar. Can
.
J. For. Res
 .
35
:
1779
1789
.
Yin
T.M.
DiFazio
S.P.
Gunter
L.E.
Jawdy
S.S.
Boerjan
W.
Tuskan
G.A.
.
2004
.
Genetic and physical mapping of Melampsora rust resistance genes in Populus and characterization of linkage disequilibrium and flanking genomic sequence
.
New Phytol
 .
164
:
95
105
.
Yin
T.M.
Zhang
X.Y.
Gunter
L.E.
Li
S.X.
Wullschleger
S.D.
Huang
M.R.
Tuskan
G.A.
.
2009
.
Microsatellite primer resource for Populus developed from the mapped sequence scaffolds of the Nisqually-1 genome
.
New Phytol
 .
181
:
498
503
.
Zhang
D.
Zhang
Z.
Yang
K.
.
2006
.
QTL analysis of growth and wood chemical content traits in an interspecific backcross family of white poplar (Populus tomentosa × P. bolleana) × P. tomentosa
.
Can. J. For. Res
 .
36
:
2015
2023
.

Supplementary data