Functional analysis of the three HMA4 copies of the metal hyperaccumulator Arabidopsis halleri.

In Arabidopsis halleri, the HMA4 gene has an essential function in Zn/Cd hypertolerance and hyperaccumulation by mediating root-to-shoot translocation of metals. Constitutive high expression of AhHMA4 results from a tandem triplication and cis-activation of the promoter of all three copies. The three AhHMA4 copies possess divergent promoter sequences, but highly conserved coding sequences, and display identical expression profiles in the root and shoot vascular system. Here, an AhHMA4::GFP fusion was expressed under the control of each of the three A. halleri HMA4 promoters in a hma2hma4 double mutant of A. thaliana to individually examine the function of each AhHMA4 copy. The protein showed non-polar localization at the plasma membrane of the root pericycle cells of both A. thaliana and A. halleri. The expression of each AhHMA4::GFP copy complemented the severe Zn-deficiency phenotype of the hma2hma4 mutant by restoring root-to-shoot translocation of Zn. However, each copy had a different impact on metal homeostasis in the A. thaliana genetic background: AhHMA4 copies 2 and 3 were more highly expressed and provided higher Zn tolerance in roots and accumulation in shoots than copy 1, and AhHMA4 copy 3 also increased Cd tolerance in roots. These data suggest a certain extent of functional differentiation among the three A. halleri HMA4 copies, stemming from differences in expression levels rather than in expression profile. HMA4 is a key node of the Zn homeostasis network and small changes in expression level can have a major impact on Zn allocation to root or shoot tissues.

Introduction

Plants possess a complex and tightly regulated metal homeostasis network enabling appropriate metal supply to tissues throughout development (Puig and Peñarrubia, 2009). This allows plants to colonize environments that substantially differ in metal availability, ranging from severe deficiency to toxicity in polluted areas. For instance, so-called metal-hyperaccumulating plants, which are represented by ~500 species, establish populations on soils heavily polluted by metals and are characterized by their capacity to accumulate extremely high concentrations of metals in shoots (e.g. >0.3% of their dry biomass of Zn and 0.01% of Cd) without any toxicity symptoms (Verbruggen ; Krämer, 2010; Hanikenne and Nouet, 2011). Metal-hyperaccumulating plants are instrumental in unravelling the molecular and adaptive mechanisms underlying a naturally selected extreme trait (Hanikenne ; Verbruggen ; Krämer, 2010; Verbruggen ) and possibly in the development of technologies for phytoremediation (soil remediation by plants), phytomining (metal extraction from the soil), or biofortification (increasing the amount of essential micronutrients in edible plant parts) (Zhao and McGrath, 2009).

Arabidopsis halleri is a model for studying Zn and Cd hyperaccumulation in plants (Clauss and Koch, 2006; Hanikenne ; Roosens ; Krämer, 2010). A. halleri is closely related to the non-accumulator and non-tolerant species A. thaliana and A. lyrata, with divergence estimated at 3–5.8 and 0.4–2 million years ago, respectively (Yogeeswaran ; Clauss and Koch, 2006; Roux ). Transcriptomics and quantitative genetics studies allowed the identification of candidate genes for Zn and Cd hypertolerance and hyperaccumulation. These genes contribute to metal homeostasis, namely metal transport, chelation, or detoxification, and are highly expressed in A. halleri compared to A. thaliana and A. lyrata (Becher ; Weber ; Filatov ; Talke ; Courbot ; Willems ; Frérot ; Willems ).

Among the metal homeostasis genes, Heavy Metal ATPase 4 (HMA4) has been identified as one of the key components contributing to Zn and Cd hyperaccumulation and hypertolerance in A. halleri (Talke ; Courbot ; Willems ; Hanikenne ; Frérot ; Willems ). First characterized in A. thaliana, HMA4 encodes a IB-2 P-type ATPase acting as a Zn/Cd efflux pump (Williams and Mills, 2005; Palmgren and Nissen, 2011; Pedersen ; Hanikenne and Baurain, 2014). Together with its paralog HMA2, it is localized at the plasma membrane and is expressed in vascular tissues in roots and shoots (Hussain ; Verret ; Siemianowski ). HMA2 and HMA4 are responsible for the translocation of Zn from roots to shoots: an hma2hma4 double mutant experiences severe Zn deficiency in shoots and cannot develop normally and set seeds. This phenotype is reversed by massive external Zn supply (Hussain ). HMA2 and HMA4 are also responsible for Cd export to the shoots (Wong and Cobbett, 2009).

In A. halleri, HMA4 displays 6- and 30-fold higher transcript levels in roots and shoots, respectively, compared to A. thaliana (Talke ). The HMA4 gene co-localizes with major quantitative trait loci for Zn and Cd hyperaccumulation and hypertolerance (Courbot ; Willems ; Frérot ; Willems ). High expression of HMA4 in A. halleri is required for high rates of root-to-shoot translocation of Zn by mediating xylem loading in roots and possibly intercellular distribution in leaves (Hanikenne ). Increased expression of HMA4 in A. halleri results from tandem triplication and cis-regulatory changes activating the promoters of all three HMA4 copies (Hanikenne ; Hanikenne ). The promoter of HMA4 copy 1 (pAhHMA4-1) shares ~73% identity with the AtHMA4 promoter, whereas the promoters of copies 2 and 3 (pAhHMA4-2 and pAhHMA4-3) are highly similar (~82% of identity), but share only limited similarity with the AtHMA4 promoter (Hanikenne ). Each of the three promoters contributes to the elevated expression of HMA4 and mediate a similar spatial profile of expression in vascular tissues (Hanikenne ). Promoter GUS reporter studies established that HMA4 is expressed in the pericycle and xylem parenchyma in A. halleri roots, whereas it is expressed in the xylem parenchyma and cambium in shoots (Hanikenne ). An analysis of nucleotide polymorphism patterns at the AhHMA4 locus provided evidence for positive selection on cis-regulatory sequences and/or copy number expansion during the evolutionary history of A. halleri (Hanikenne ). In addition, coding sequences of the three AhHMA4 copies are almost identical (>99% nucleotide sequence identity) which results from ectopic gene conversion among gene copies (Hanikenne ). Together, this complex nucleotide polymorphism pattern at the AhHMA4 locus substantiates selection for increased gene product.

HMA4 is also constitutively more highly expressed in Noccaea caerulescens, another Zn and Cd hyperaccumulator species in the Brassicaceae that diverged from A. thaliana about 20 million years ago (Verbruggen ; Krämer, 2010; Hanikenne and Nouet, 2011). As in A. halleri, high expression of HMA4 in N. caerulescens is associated with copy number expansion and regulatory changes (Ò Lochlainn ; Craciun ), suggesting parallel evolution in the two hyperaccumulator species. Moreover, differences in HMA4 expression levels between N. caerulescens populations exhibiting contrasted metal accumulation and tolerance were associated with gene copy number variations (Craciun ).

In this study, genetic constructs were generated to express the AhHMA4 cDNA in fusion with GFP under the control of each of the three AhHMA4 promoters for transformation in A. halleri and in the hma2hma4 double mutant of A. thaliana. This allowed determination of the expression pattern of the HMA4 protein and examination of the contribution of each AhHMA4 gene copy to Zn/Cd accumulation and tolerance. The reported data suggest functional specialization among AhHMA4 gene copies.

Materials and methods

Plant material, cultivation, and transformation

A. halleri (L.) O’Kane and Al-Shehbaz ssp. halleri (accession Langelsheim) or A. thaliana L. Heynhold (accession Columbia-0, Col-0) and the A. thaliana hma2hma4 double mutant (Col-0 background, described in Hussain ) were used in all experiments. For physiological experiments, A. thaliana plants were cultivated in liquid or on solid modified Hoagland medium supplemented with 0.8% (w/v) agar (Agar type M; Sigma-Aldrich) in plastic Petri dishes as previously described (Talke ; Hanikenne ), either under a photoperiod of 16h (long days) or 8h (short days) light (100 µmol photon m−2 s−1) in a climate-controlled growth chamber at 21/19°C (day/night).

Before transformation, the hma2hma4 mutant was cultivated on soil supplied daily with 1mM ZnSO4.7H2O solution for 7 weeks and then 3mM ZnSO4.7H2O for 5 weeks. The hma2hma4 mutant was transformed by floral dip (Clough and Bent, 1998). Homozygous lines were obtained after selection on hygromycin B (20 µg/ml) on half-strength Murashige and Skoog solid medium (Duchefa Biochimie) supplemented with 1% sucrose. For the phenotyping on soil, seeds of the complemented lines, wild-type, and hma2hma4 mutant were germinated on half-strength Murashige and Skoog solid medium supplemented with 1% sucrose in long days. Then, 18-day-old seedlings were transferred in soil watered with tap water and grown in long days for 6 weeks.

Agrobacterium tumefaciens-mediated (GV3101, pMP90) stable transformation of A. halleri was performed using a tissue-culture based procedure (Hanikenne ).

Generation of the pAhHMA4-x::HMA4::GFP construct

The pAhHMA4-x::HMA4::GFP constructs for transformation of the hma2hma4 mutant and A. halleri were generated as follows: (i) the eGFP gene was amplified using primers 5′-tatagtcgacatggtgagcaagggcgaggag-3′ containing a SalI restriction site (italic) and 5′-tcttaattaattacttgtacagctcgtccatgccgagagtgat-3′ containing a PacI restriction site (italic). The full-length AhHMA4 cDNA was amplified from an A. halleri cDNA library using primers 5′-tatagtcgacagcactcacatggtgatggtgg-3′ containing a SalI restriction site (italic) and 5′-caccccgaaaatggcgtcacaaaacaaag-3′ containing a BsaHI restriction site (italic). The promoter fragment and cDNA were fused by ligation at the SalI site. Then the AhHMA4::GFP fragment was cloned into the BsaHI/PacI sites of the pBluescript II KS+ vector carrying the promoter of AhHMA4 copy 1 (pAhHMA4-1) (2296bp) as described in Hanikenne . (ii) The pAhHMA4-2 (2284bp) and pAhHMA4-3 (2030bp) promoters were amplified from A. halleri genomic DNA (Langelsheim Lan-3.1; Talke ; Hanikenne ) using primers 5′-atatgtcgactttctcttcttctttgttttgtgacgcc-3′ containing a SalI restriction site (italic) and 5′-atatgaattc ggcgcgccgctctctatcctcctttgtaagttcacc-3′ containing SalI (italic) and AscI (bold) restriction sites and cloned by replacing the pAhHMA4-1 into the BsaHI/AscI sites of the pBluescript II KS+ vector carrying AhHMA4::GFP. (iii) The pAhHMA4::AhHMA4::GFP cassettes were AscI/PacI-excised from pBluescript II KS+ and cloned into the corresponding sites of the pMDC32 binary vector (Curtis and Grossniklaus, 2003) from which the 35S promoter had been removed using ApaI and HindIII. All constructs were verified by sequencing.

Fluorescence confocal microscopy

The roots of 18-day-old seedlings of three independent complemented A. thaliana lines and A. halleri lines expressing pAhHMA4::AhHMA4::GFP were analysed for each of the three promoters. Images were collected using a SP2 confocal microscope (Leica, Mannheim, Germany) as previously described (Tillemans ; Rausin ). An Argon/Ion laser (488nm) for excitation of the GFP protein and a Helium/Neon laser (543nm) for excitation of propidium iodide (cell walls) were used. The emission light was dispersed and recorded at 500–540nm for GFP and 600–700nm for propidium iodide. Plasmolysis of the root cells was performed by incubating roots in a 4% (w/v) NaCl solution.

Tolerance assays

Col-0 (wild-type) and hma2hma4 mutant plants as well as the complemented mutants (four to five independent T3 homozygous lines per promoter construct) were analysed. The seedlings were cultivated in short days on solid Hoagland medium containing 1 µM ZnSO4 for 10 days and then transferred on solid Hoagland medium containing either 1 µM ZnSO4, 150 µM ZnSO4, or 40 µM CdSO4 (Hanikenne ). The root growth was monitored every 2 days and was measured on day 7.

Analysis of metal accumulation in plants

Col-0 (wild-type) and hma2hma4 mutant plants as well as the complemented mutants (four to five independent T3 homozygous lines per promoter construct) were analysed. Eighteen days after germination on modified Hoagland medium in short days, the seedlings were transferred to hydroponic trays (Araponics, Tocquin ) with Hoagland medium containing 1 μM ZnSO4 and grown in short days for 2 weeks before initiating the treatments. Plants were then cultivated on either 0.2 μM ZnSO4 or 0.05 μM CdSO4 for 4 weeks under long days. Root and rosette tissues of two to three plants were harvested separately and desorbed as described (Talke ). Plant tissues were dried at 60°C for 3 days. Tissues (10–50mg) were then acid-digested in a DigiPrep tube with 3ml 65% HNO3 (Sigma-Aldrich) on a DigiPrep Graphite Block Digestion System (SCP Science) as follows: 15min at 45°C, 15min at 65°C, and 90min at 105°C. After cooling, sample volumes were adjusted to 10ml with milliQ water, and 200 µl >65% HNO3 was added. Metal concentration was measured using inductively coupled plasma atomic emission spectroscopy with a Vista-AX instrument (Varian, Melbourne, Australia).

Analysis of gene expression in A. thaliana plants

Col-0 (wild-type) and hma2hma4 mutant plants as well as the complemented mutants (three to four independent T3 homozygous lines per promoter construct) were analysed. Eighteen days after germination on modified Hoagland medium in short days, seedlings were transferred to hydroponic trays with Hoagland medium containing 0.2 μM ZnSO4 and grown for 5 weeks in short days. Root and rosette tissues were harvested separately from two to four plants per genotype. Total DNase-treated RNA was extracted with RNeasy Plant Mini Kit and DNase set (Qiagen). Quality and quantity of RNA was checked visually by denaturing gel electrophoresis and by photometric analysis (A260 and A280). Syntheses of cDNA were performed with 500ng of total RNA using Oligo(dT) and the RevertAid H Minus First Strand cDNA Synthesis Kit (Fisher Scientific). Quantitative PCR reactions were performed in 384-well plates with an ABI Prism 7900HT system (Applied Biosystems) using Mesa Green qPCR MasterMix (Eurogentec). A total of three technical repeats were run for each combination of cDNA and primer pair (Supplementary Table S1). Equal amounts of cDNA, corresponding to approximately 6ng of total RNA, were used in each reaction. In addition, each reaction contained 5 µl of Mesa Green qPCR MasterMix and 2.5 pmol of forward and reverse primers in a total volume of 10 µl. The following standard thermal profile was used: 2min at 50°C, 10min at 95°C, 40 repeats of 15 s at 95°C and 60 s at 60°C, and a final stage of 15 s at 95°C, 15 s at 60°C, and 15 s at 95°C to determine dissociation curves of the amplified products. The quality of the PCRs was checked visually through analysis of dissociation and amplification curves, and reaction efficiencies were determined for each PCR using the LinRegPCR software v2013 (Ruijter ). Mean reaction efficiencies were then determined for each primer pair from all reactions (>100 reactions; Supplementary Table S1) and used to calculate relative gene expression levels by normalization using multiple reference genes with the qBase software (Biogazelle; Hellemans ). Two reference genes (UBQ10, EF1a) were selected from the literature (Czechowski ). Their adequacy to normalize gene expression in the experimental conditions was verified using the geNorm software in qBase (gene stability measure M = 0.464, pairwise variation CV = 0.161) (Vandesompele ).

Statistical analysis

All data evaluation and statistics were done using GraphPad Prism 5 (GraphPad Software) and the GLM procedure (two-way ANOVA) in SAS 9.3 (SAS Software).

Accession numbers

A. halleri sequences are available through EBI (http://www.ebi.ac.uk), accession numbers EU382072, EU382073.

Results

Expression of the three AhHMA4::GFP copies complemented the hma2hma4 phenotype

Based on sequence similarity, it was postulated that if functional differences exist among the three AhHMA4 copies, they are most likely determined by the promoters. To individually examine the function of each AhHMA4 gene copy, homozygous transgenic lines (T3 generation) were generated of the hma2hma4 loss-of-function A. thaliana mutant expressing the same AhHMA4 cDNA under the control of each of the three AhHMA4 promoters. Owing to technical constraints, relatively short versions of the pAhHMA4-2 and pAhHMA4-3 promoters (about 1000bp upstream of the ATG translation initiation codon) were cloned previously for a promoter GUS reporter study (Hanikenne ). Here, longer DNA fragments of 2284 and 2030bp upstream of the ATG start codon were cloned for pAhHMA4-2 and pAhHMA4-3, respectively, based on available A. halleri BAC sequences in the HMA4 locus (Hanikenne ). For the pAhHMA4-1 promoter, the 2296bp fragment described in Hanikenne was used. Because no antibody raised against the HMA4 protein is available, an AhHMA4::GFP fusion was cloned in order to localize the HMA4 protein in planta.

In the described growth conditions on standard soil, the hma2hma4 mutant developed its typical phenotype: stems were very small, which led to a bushy aspect with chlorotic leaves and inflorescences presented no pollen and sterile flowers, in contrast to the wild type (Fig. 1A, B), as previously described (Hussain ; Wong and Cobbett, 2009; Mills ). Expression of AhHMA4::GFP under the control of any of the three AhHMA4 promoters fully complemented the mutant, restoring a wild-type-like phenotype: the plants developed normally and were able to flower and set seeds without additional Zn supply in the soil (Fig. 1C–E). HMA4 transcript levels were determined in root and shoot tissues of these plants by quantitative RT-PCR analysis. As expected, AtHMA4 transcripts were detected in wild-type plants only (Supplementary Fig. S1). AhHMA4 transcripts were detected consistently in the complemented mutants only and the steady-state levels were higher in roots than in shoots. In roots, AhHMA4 transcript levels were not significantly different between constructs (Fig. 2A). In contrast, in shoots, the AhHMA4 transcript levels were 3-fold higher in pAhHMA-3 lines than in pAhHMA4-1 and pAhHMA4-2 lines (Fig. 2B).

Fig. 1.

The hma2hma4 Zn deficiency phenotype can be rescued by AhHMA4::GFP. (A) Wild-type, and (B) hma2hma4 mutant A. thaliana plants as well as transgenic homozygous plants expressing AhHMA4::GFP under the control of (C) pAhHMA4-1, (D) pAhHMA4-2, and (E) pAhHMA4-3 were grown on soil watered with tap water (this figure is available in colour at JXB online).

Fig. 2.

Expression of AhHMA4 in roots and shoots of the complemented lines. Transgenic homozygous plants expressing AhHMA4::GFP under the control of pAhHMA4-1, pAhHMA4-2, and pAhHMA4-3 were grown hydroponically with 0.2 µM Zn. Relative expression levels of AhHMA4 in (A) roots and (B) shoots are means±SEM of three to four independent lines from one experiment, representative of two independent biological experiments, each including two to four plants per line. The data were analysed with an unpaired t-test. Statistically significant differences (P < 0.05) between means are indicated by different superscripted letters. RTL: relative transcript level.

The AhHMA4 protein showed non-polar localization in pericycle cells in roots

Because the AhHMA4::GFP fusion protein was functional (Fig. 1), the complemented mutant transgenic lines were used to examine if each of the three AhHMA4 promoters determined an identical expression profile at the protein level (Fig. 3). In all cases, the AhHMA4::GFP protein was expressed in the pericycle in roots. The protein localized in the plasma membrane of cells in a non-polar fashion. Performing a plasmolysis of the root cells confirmed the plasma membrane localization (Fig. 3I, J).

Fig. 3.

Localization of AhHMA4::GFP fusion protein in A. thaliana. GFP fluorescence was imaged by confocal microscopy in roots of 18-day-old seedlings of homozygous hma2hma4 mutants expressing AhHMA4::GFP under the control of (A, B) pAhHMA4-1, (C, D) pAhHMA4-2, and (E, F) pAhHMA4-3 promoters. (G, H) Wild-type A. thaliana seedlings used as control. (I, J) Plasmolysis was performed on a pAhHMA4-1 line, resulting in shrinkage of the plasma membrane. Propidium iodide was used to stain the cell walls (red). Images are representative of a minimum of three independent lines. Dotted arrows identify the different root tissues: c, cortex; e, endodermis; ep, epidermis; p, pericycle; x, xylem. Cross sections: A, C, E, G; longitudinal sections: B, D, F, H, I, J. Scale bars: 30 µm (A, C, E, G, H, J), 20 μm (B, D, F) and 10 µm (I).

Localization of AhHMA4::GFP in A. halleri transgenics was also studied. Because transforming A. halleri is a long and complex process, only A. halleri lines expressing AhHMA4::GFP under the control of pAhHMA4-1 and pAhHMA4-3 could be generated as representatives of the two extreme modifications of the HMA4 promoter compared to A. thaliana (Fig. 4). For these two promoters, the expression profile and sub-cellular localization was identical to that observed in A. thaliana. It can be reasonably assumed that pAhHMA4-2 is also active in the pericycle of A. halleri roots. In addition, GFP was observed in cells within the vascular bundle (Fig. 4B).

Fig. 4.

Localization of AhHMA4::GFP fusion protein in A. halleri. GFP fluorescence was imaged by confocal microscopy in roots of A. halleri expressing AhHMA4::GFP under the control of (A, B) pAhHMA4-1 and (C) pAhHMA4-3 promoters and of (D) wild-type A. halleri as control. Propidium iodide was used to stain the cell walls (red). Images are representative of a minimum of two independent lines. Dotted arrows identify the different root tissues: c, cortex; e, endodermis; ep, epidermis; p, pericycle; x, xylem. Cross sections: B; longitudinal sections: A, C, D. Scale bars: 30 µm (A, B, D) and 50 μm (C).

Differential Zn and Cd tolerance in complemented lines

To determine how the expression of AhHMA4 under the control of each AhHMA4 promoter contributed to Zn and Cd tolerance, root growth was measured in 10-day-old seedlings (wild-type, hma2hma4 mutant, and complemented lines) exposed for 7 days on agar medium plates to the control condition (1 µM Zn), Zn excess (150 µM), and Cd (40 µM). In control conditions, the root growth was identical for all genotypes (Fig. 5A). Upon exposure to excess Zn, the root growth of all genotypes was reduced but the hma2hma4 mutant was more affected, with a 59% growth reduction compared to control conditions (Fig. 5B). The root growths of wild-type and pAhHMA4-1 lines were similar and ~16% higher than hma2hma4. pAhHMA4-2 and pAhHMA4-3 lines were the most tolerant with a ~39% higher root growth compared to the double mutant (Fig. 5B). In the presence of Cd, the root growth drastically decreased for all genotypes, by about 70–90% compared to control conditions (Fig. 5C). Again, the hma2hma4 mutant was the most affected and its growth almost stopped (90% growth reduction compared to control conditions). The root growths of wild-type, pAhHMA4-1, and pAhHMA4-2 lines were similar and 62–90% higher than for hma2hma4. The root growth of pAhHMA4-3 lines was even higher: 173% higher than the double mutant (Fig. 5C).

Fig. 5.

Zn and Cd tolerance in A. thaliana complemented lines. Wild-type and hma2hma4 mutant A. thaliana seedlings as well as transgenic homozygous seedlings expressing AhHMA4::GFP under the control of pAhHMA4-1, pAhHMA4-2, and pAhHMA4-3 were exposed to (A) 1 µM ZnSO4 (control), (B) 150 µM ZnSO4, and (C) 40 µM CdSO4. Root growth values (in cm) are means±SEM and are representative of one experiment out of a total of three independent biological experiments, each including 15–20 seedlings per line/treatment and four to five independent lines. The data were analysed with a two-way ANOVA test with log-transformed values followed by Tukey and Kramer’s multiple comparison tests (P < 0.05). Statistically significant differences between means within treatments are indicated by different superscripted letters.

Differential Zn and Cd accumulation in complemented lines

How the expression of AhHMA4 under the control of each AhHMA4 promoter contributed to metal accumulation was also determined. Zn and Cd concentrations were measured in the roots and rosette leaves of wild-type, hma2hma4 mutant, and complemented plants after cultivation for 4 weeks in Hoagland hydroponic medium containing either 0.2 µM Zn or 0.05 µM Cd. Hoagland medium classically contains 1 µM Zn (Becher ; Talke ), and at this Zn concentration, the growth of the hma2hma4 mutant was rescued: it could flower and set seeds similarly to the wild type (data not shown). A concentration of 0.2 µM Zn was optimal to distinguish the mutant from the wild type in these growth conditions and assess complementation by the AhHMA4::GFP constructs. Indeed, at 0.2 µM Zn, the hma2hma4 mutant displayed a Zn-deficiency phenotype, whereas all complemented lines displayed a wild-type phenotype (Supplementary Fig. S2). The hma2hma4 mutant accumulated about 6-fold higher Zn in roots and 4.4-fold lower Zn in shoots than the wild type, respectively (Fig. 6A, B). In roots, the complemented lines accumulated less Zn than the mutant: root Zn accumulation was intermediate between wild type and mutant in pAhHMA-1 and pAhHMA-2 lines while the pAhHMA-3 lines accumulated the smallest Zn amount, at a level similar to the wild type (Fig. 6A). In contrast, Zn accumulation in shoots was 4.4-fold higher in the pAhHMA-1 lines than in the hma2hma4 mutant and similar to the wild type. The pAhHMA-2 and pAhHMA-3 lines accumulated even more Zn: ~8-fold more than the hma2hma4 mutant and 1.8-fold more than the wild type (Fig. 6B). Identical observations were made upon 0.05 µM Cd exposure (in the presence of 1 µM Zn): expression of AhHMA4::GFP in the hma2hma4 mutant restored Cd accumulation in shoots and, in pAhHMA-2 and pAhHMA-3 lines, the Cd levels were similar to wild-type levels in roots and shoots (Fig. 6C, D).

Fig. 6.

Zn and Cd accumulation in A. thaliana complemented lines. Wild-type and hma2hma4 mutant A. thaliana plants as well as transgenic homozygous plants expressing AhHMA4::GFP under the control of pAhHMA4-1, pAhHMA4-2, and pAhHMA4-3 were cultivated hydroponically with (A, B) 0.2 µM ZnSO4 and (C, D) 0.05 µM CdSO4/1 µM ZnSO4. Zn and Cd concentrations (mg kg−1 dry weight) were measured in root and rosette tissues collected from two to three plants per line. Values are means±SEM of four to five independent lines from one experiment and are representative of three independent experiments. The data were analysed with a two-way ANOVA test with log-transformed values followed by Tukey and Kramer’s multiple comparison tests (P < 0.05). Statistically significant differences (P < 0.05) between means within each figure panel are indicated by different superscripted letters.

Differential expression of Zn status-responsive genes in complemented lines

To further examine how the expression of AhHMA4 under the control of each AhHMA4 promoter contributed to Zn translocation from roots to shoots, the expression level of four genes that respond to Zn status in tissues was measured: ZIP4, ZIP9, and IRT3 encoding Zn uptake transporters, and NAS2 encoding an enzyme involved in the synthesis of the Zn chelator nicotianamine (NA). The four genes are up-regulated by Zn deficiency and down-regulated by Zn excess in A. thaliana (Talke ). Gene expression in roots and shoots of plants hydroponically cultivated at 0.2 µM Zn for 5 weeks was analysed by quantitative RT-PCR. In roots, for all four genes, the steady-state transcript levels were lower (up to 75%) in the hma2hma4 mutant than in the wild type (Fig. 7), as expected: the increased Zn accumulation in roots of the mutant (Fig. 6A) resulted in a down-regulation of the genes. In roots, ZIP4 steady-state transcript levels were higher in all complemented lines than in the wild type (3.4-fold) and the hma2hma4 mutant (8-fold) (Fig. 7B). In addition, the steady-state transcript levels of IRT3, ZIP9, and NAS2, were 2- to 5.4-fold higher in the pAhHMA4-2 and pAhHMA4-3 lines compared to the pAhHMA4-1, wild type, and hma2hma4 mutant lines (Fig. 7A, C, D), indicative of Zn depletion in root tissues.

Fig. 7.

Expression of Zn-responsive genes in roots of A. thaliana complemented lines. Wild-type and hma2hma4 mutant A. thaliana plants as well as transgenic homozygous plants expressing AhHMA4::GFP under the control of pAhHMA4-1, pAhHMA4-2, and pAhHMA4-3 were grown hydroponically with 0.2 µM Zn. Relative expression levels of (A) IRT3, (B) ZIP4, (C) ZIP9, and (D) NAS2 are presented as means±SEM of three to four independent lines from one experiment representative of two independent biological experiments, each including two to four plants per line. The data were analysed with an unpaired t-test. Statistically significant differences (P < 0.05) between means are indicated by different superscripted letters. RTL: relative transcript level.

In contrast, in shoots, all genes were more highly expressed in the hma2hma4 mutant (5 to 137-fold) compared to the wild type (Fig. 8), highlighting the major Zn deficiency in shoots of the mutant (Hussain ). Steady-state transcript levels of all genes were massively down-regulated in complemented lines compared to the mutant. The steady-state transcript levels of IRT3 and ZIP4 in all complemented lines were similar to the wild type (Fig. 8A, B). Interestingly, the steady-state transcript levels of ZIP9 and NAS2 were higher, although not significantly, in shoots of the pAhHMA4-3 lines than in the other complemented lines (Fig. 8C, D).

Fig. 8.

Expression of Zn-responsive genes in shoots of A. thaliana complemented lines. Wild-type and hma2hma4 mutant A. thaliana plants as well as transgenic homozygous plants expressing AhHMA4::GFP under the control of pAhHMA4-1, pAhHMA4-2, and pAhHMA4-3 were grown hydroponically with 0.2 µM Zn. Relative expression levels of (A) IRT3, (B) ZIP4, (C) ZIP9, and (D) NAS2 are presented as means±SEM of three to four independent lines from one experiment representative of two independent biological experiments, each including two to four plants per line. The data were analysed with an unpaired t-test. Statistically significant differences (P < 0.05) between means are indicated by different superscripted letters. RTL: relative transcript level.

Discussion

So far, no data has been available evaluating separately the contribution of each of the three AhHMA4 copies to Zn homeostasis. Here, the possible functional differentiation among the three AhHMA4 copies was examined and the sub-cellular localization and expression profile of the AhHMA4 protein determined when expressed under the control each AhHMA4 promoter.

The HMA4 protein is expressed in the root pericycle

The sub-cellular localization of the A. thaliana or A. halleri HMA4 protein in the plasma membrane has so far only been determined by ectopic over-expression in yeast, A. thaliana protoplasts, or tobacco (Verret ; Verret ; Courbot ; Siemianowski ). Here, using the AhHMA4 promoters allowed the determination of the sub-cellular localization of the reporter protein upon tissue-specific expression. When expressed under the control of the AhHMA4 promoters, the AhHMA4::GFP fusion was functional in planta (Fig. 1) and localized in the plasma membrane of pericycle cells in roots of both species (Figs. 3 and 4). As with promoter GUS reporter constructs (Hanikenne ), the three AhHMA4 promoters determined an identical pattern of localization. The AtHMA2 protein, the paralog of AtHMA4, has a similar expression pattern in root cells (Hussain ; Sinclair ). In A. halleri, GFP was also observed in cells within the vascular bundle, which likely correspond to the xylem parenchyma of roots, again consistent with GUS staining (Hanikenne ).

Zn hyperaccumulation is a directional process where Zn is very efficiently translocated from the root to the shoot. After cellular uptake in the root epidermis, Zn is transported radially towards the xylem. A polar localization of several membrane transporters has been observed in plant cells, providing directional substrate transport in tissues. For instance, the polar localization of PIN auxin efflux carriers is key for the control of plant growth (Tanaka ). In Oryza sativa (rice), the Lsi1 and Lsi2 transporters act together to provide polarized influx and efflux of silicon in root exodermal and endodermal cells, respectively (Ma ; Ma ). Current models representing the function of HMA4 in root cells hypothesize that the protein accumulates on one side of the pericycle cell membrane, directly adjacent to the xylem, allowing Zn loading (Hanikenne ; Palmer and Guerinot, 2009; Krämer, 2010; Hanikenne and Nouet, 2011; Claus ). In contrast, the HMA4 protein appears to be distributed uniformly, in a non-polar fashion, in the plasma membrane of pericycle cells. The observations presented here suggest that HMA4 contributes to the mobilization and efflux of cellular Zn within the root cells delimited by the Casparian strip. Once in the apoplast, Zn becomes available for xylem loading.

AhHMA4 expression in hma2hma4 complemented lines

In roots of the complemented hma2hma4 mutant lines, the AhHMA4 transcript levels were slightly, but not significantly, higher in pAhHMA4-3 lines compared to pAhHMA4-1 lines (Fig. 2A). This is consistent with previously published data on GUS activity in promoter GUS reporter lines, where no major differences between promoters were observed (Hanikenne ). However, in shoots, pAhHMA4-3 led to AhHMA4 steady-state levels 3-fold higher than pAhHMA4-1 and pAhHMA4-2, whereas GUS activity levels were similar for all three promoters (Fig. 2B, Hanikenne ). Two hypotheses can be suggested to explain this discrepancy in shoots. First, levels of AhHMA4::GFP transcript were measured, whereas the GUS assay measures protein activity. Disconnection between transcript levels and protein levels is not rare (Ghazalpour ; Kuersten ). These two measurements can thus provide different results as translation efficiency of the transcripts or protein stability can vary (Dvir ; Kim ; Remy ). Note that the AhHMA4::GFP protein fusion could not be detected in western blots to evaluate protein levels. Second, larger fragments for pAhHMA4-2 (2284bp) and pAhHMA4-3 (2030bp) were cloned in this study compared to Hanikenne . Pairwise comparisons (BLASTN, NCBI) of the three AhHMA4 promoters used in this study and the AtHMA4 promoter used in Hanikenne identified six regions with more than 70% identity in the promoters (A–F in Supplementary Fig. S3). This analysis confirmed that promoters pAhHMA4-2 and pAhHMA4-3 shared a high level of identity (83%) and diverged substantially from pAtHMA4 and pAhHMA-1. A conserved block (F in Supplementary Fig. S3) is specific to pAhHMA4-2 and pAhHMA4-3. These sequence differences may be responsible for differential AhHMA4 expression in the complemented lines and may determine the differences in Zn tolerance and accumulation between pAhHMA4-2/-3 and pAhHMA4-1 lines. Further work will be required to dissect the function of the promoters.

Expression of the AhHMA4 complements the hma2hma4 mutant phenotype

The expression of AhHMA4::GFP under the control of each of the three AhHMA4 promoters complemented the phenotypes of the hma2hma4 mutant: it restored root-to-shoot translocation of Zn, abolished Zn deficiency in shoots, and enabled the plants to complete their life cycle. The expression of AhHMA4 also increased Zn and Cd tolerance in roots and accumulation in shoots. This was reflected by changes in the expression levels of Zn-regulated genes, which were back to wild-type levels when compared to the mutant. These genes are up-regulated by Zn deficiency and down-regulated by Zn excess in A. thaliana (Talke ). This indicated that all three copies of the AhHMA4 gene are functional in A. thaliana. This is in contrast with a recent report by Iqbal . In their study, the expression of A. thaliana or N. caerulescens HMA4 cDNAs under the control of N. caerulescens HMA4 promoters, representing different gene copies from three different accessions of N. caerulescens, did not complement the hma2hma4 A. thaliana mutant and rather exacerbated the shoot Zn-deficiency phenotype. The aggravation of the phenotype was positively correlated with the HMA4 expression level. Because the two species are phylogenetically more distant, these observations may be linked to the non-conserved specificity of expression of N. caerulescens HMA4 promoters in A. thaliana tissues, with high expression in leaf mesophyll cells and ectopic expression in roots (Iqbal ).

Copy-specific phenotypes are linked to differences in HMA4 expression level

If the expression of AhHMA4::GFP under the control of each of the three AhHMA4 promoters complemented the phenotypes of the hma2hma4 mutant, there was substantial differences among copies. A subset of complemented lines displayed increased root tolerance to Zn (pAhHMA4-2 and pAhHMA4-3) and Cd (pAhHMA4-3 only) compared to wild-type, mutant, and pAhHMA4-1 lines.

In a previous study, the expression of AhHMA4 under the control of pAhHMA4-1 in Col-0 wild-type plants resulted in a marginal increase (~16%) of Zn accumulation in shoots (Hanikenne ). Expression of the same construct in Nicotiana tabacum and Solanum lycopersicum also resulted in moderate changes in shoot Zn contents, which were dependent on Zn supply in the medium, and in alterations of the metal homeostasis network (Barabasz ; Barabasz ; Antosiewicz ). Here, expression of AhHMA4::GFP under the control of pAhHMA4-1 in the hma2hma4 double mutant restored wild-type levels of shoot Zn accumulation (Fig. 5). Interestingly, pAhHMA4-2 and pAhHMA4-3 transgenic lines displayed ~800% and ~180% increased accumulation of Zn in shoots compared to the mutant and the Col-0 wild type, respectively (Fig. 6B). This represents a substantial increase in shoot accumulation. Indeed, summing the effect of each HMA4 copy shows that the increase in Zn shoot accumulation compared to the wild type was 4.5-fold, which is in the range of previous comparisons of A. thaliana and A. halleri grown at low Zn supply (Talke ). Note that the pAhHMA-2 and pAhHMA-3 promoters also had a higher impact on Cd accumulation than pAhHMA-1 (Fig. 6D).

The expression of Zn-responsive genes reflected changes in Zn accumulation in the complemented lines tissues. Indeed, IRT3, ZIP4, ZIP9, and NAS2 steady-state transcript levels in roots were highest in the pAhHMA4-2 and pAhHMA4-3 lines compared to the wild type (Fig. 7). This indicated that the pAhHMA4-2- and pAhHMA4-3-dependent expression of AhHMA4 resulted in a more efficient Zn translocation from roots to shoots and higher Zn depletion in roots compared to the wild-type, the hma2hma4 mutant, and the pAhHMA4-1 lines.

Claus modelled Zn transport in A. thaliana roots describing the spatio-temporal evolution of Zn concentration in the symplast and apoplast depending on Zn supply. The model predicted that, in addition to radially oriented advection and the cylindrical geometry of the roots, HMA4 protein level affects the overall Zn concentration in the pericycle, and suggested that a slight change in AtHMA4 transcript level could drastically modify the Zn gradient from roots to shoots. This is in agreement with the function of HMA4 in A. halleri where increased expression supports an enhanced Zn flux from the root symplasm into the xylem vessels and contributes to hyperaccumulation. The activity of AhHMA4 in roots determines Zn root to shoot ratios and regulates Zn-responsive gene expression in roots (Hanikenne ).

In this context, an examination was performed to determine whether the phenotypes observed in the complemented lines correlated with AhHMA4 transcript levels. To a certain extent, root tolerance correlated with the level of AhHMA4 transcripts in roots (Supplementary Fig. S4). Root tolerance to Cd also correlated with AhHMA4 expression level in shoots, where AhHMA4 might provide efficient xylem unloading of Cd, acting as a sink to decrease Cd accumulation in roots. In addition, root Zn accumulation negatively correlated with AhHMA4 expression levels in roots. AhHMA4 expression also moderately influenced shoot Zn and Cd accumulation (Supplementary Fig. S5). The expression levels of ZIP9, ZIP4, and NAS2, and to a lesser extent IRT3, correlated with AhHMA4 expression in roots, indicating a more efficient transport of Zn to shoots resulting in Zn depletion in roots (Supplementary Fig. S6A-D).

Altogether, these observations suggest that higher expression of AhHMA4 in roots of pAhHMA4-2 and pAhHMA4-3 lines was responsible for the better tolerance and accumulation of Zn, as well as differential expression of Zn-responsive genes.

Physiological impact of high AhHMA4 expression in shoots

In shoots of pAhHMA4-3 lines, the steady-state expression levels of ZIP9 and NAS2 were higher than in pAhHMA4-1 and pAhHMA4-2 and wild-type lines despite a higher accumulation of Zn in shoot tissues (Fig. 8C, D). Higher accumulation of Zn in shoots should result in decreased expression of these Zn-responsive genes (Talke ). Similarly, a negative correlation between AhHMA4 and Zn-responsive gene expression is expected (Hanikenne ). In contrast, a positive correlation between AhHMA4 expression level and the expression of ZIP9 and NAS2 in shoots was observed here (Fig. S6G, H). High expression of AhHMA4 in pAhHMA4-3 lines possibly depletes Zn in shoot cells and triggers a Zn-deficiency response in shoots in the presence of high Zn concentration (see also Antosiewicz ; Iqbal ). As a result, Zn may be mislocalized in the apoplast in shoots of pAhHMA4-3 lines.

Conclusion

The data presented here suggest a certain extent of functional differentiation among the three AhHMA4 copies when expressed in A. thaliana, stemming from differences in expression levels rather than in expression profile. AhHMA4 copies 2 and 3 provide higher Zn tolerance and accumulation than copy 1 (Hanikenne ). Interestingly, AhHMA4 copy 3 was subjected to the strongest, possibly most recent, positive selection during the evolutionary history of A. halleri (Hanikenne ). The present study thus links evolutive sequence diversity patterns and function in vivo. The data further suggest that AhHMA4 copies 2 and 3 are possibly better targets for phytoremediation and biofortification purposes (Clemens ; Palmgren ). Stacking up the different copies of AhHMA4 in A. thaliana may further increase tolerance and accumulation, although it may also result in a more pronounced deregulation of Zn homeostasis in shoots (Antosiewicz ).

Supplementary data

Supplementary data are available at JXB online.

Supplementary Fig. S1. Expression of AtHMA4 in roots and shoots of the complemented mutants.

Supplementary Fig. S2. Phenotype of the complemented lines grown in Hoagland hydroponic medium.

Supplementary Fig. S3. Sequence conservation in HMA4 promoters of A. thaliana and A. halleri.

Supplementary Fig. S4. Relationship between root growth and AhHMA4 transcript levels.

Supplementary Fig. S5. Relationship between Zn or Cd accumulation and AhHMA4 transcript levels.

Supplementary Fig. S6. Relationship between expression of Zn-responsive genes and AhHMA4.

Supplementary Table S1. Sequences and reaction efficiencies of primer pairs used for real-time RT-PCR.