Cadmium Accumulation in Plants: Insights from Phylogenetic Variation into the Evolution and Functions of Membrane Transporters

Abstract

Rapid industrialization during recent decades has resulted in the widespread contamination by cadmium (Cd) of agricultural soils, which has become a ubiquitous environmental problem and poses great risk to human health via the food chain. Cd accumulation greatly varies among different plant species and even within different genotypes of the same species across the plant kingdom. A better understanding of the physiological and molecular mechanisms underlying Cd uptake, translocation, sequestration, and (re)distribution in plants will shed light on developing strategies to minimize Cd in crops. Moreover, analysis of molecular evolution of the key transporters reveals that Cd transporters were highly conserved throughout the evolutionary lineage of the entire plant kingdom and underwent lineage-specific expansion as the result of gene duplication. However, different Cd transporters may experience different evolutionary lineages from algae to angiosperms, suggesting the divergence of their roles in plant adaptation to metalliferous soil. In summary, all the knowledge in the present review can be used to predict the transfer of Cd from soils to plants, to further understand the origins of Cd-accumulating phenotypes, and to discover the plant genetic resources for the breeding of low-Cd crops and the phytoremediation of Cd-contaminated soils.

1. Introduction

Cadmium (Cd) is a toxic non-essential transition metal that has detrimental impacts on all kinds of organisms. With the rapid industrialization in recent decades, Cd contamination in agricultural soils has become one of the most important environmental issues worldwide [1], as a result of the unreasonable discharge of industrial wastes, metal smelting, and ore mining as well as the extensive use of chemical fertilizers and pesticides containing Cd over the years [2]. It is estimated that approximately 30,000 tons of Cd pollutants are discharged into the environment annually, more than 43% of which are produced from human activity [3]. Due to its strong solubility and mobility, Cd in the environment is easily absorbed and accumulated by plants. The absorption of Cd by plants may occur through soil–plant, water–plant, and air–plant systems, with the soil–plant system being the most prominent [4]. There is a close association between the soil Cd level and the plant Cd content [5]. In the soil–plant system, Cd in the rhizosphere enters plant roots mainly in ionic forms, and it is then translocated to the above-ground tissues, including grains [6]. Once excessively accumulated in plants, Cd damages the cell membrane permeability, disrupts the cellular redox homeostasis, interferes with the intercellular biochemical and physiological processes, disturbs the nutrient uptake and water relations, and reduces the photosynthetic and respiratory capacities, which thereby results in the inhibition of plant growth and development and eventually plant death [4,7]. Due to consuming agricultural products with excessive levels of Cd, the Cd accumulated in plants is easily absorbed and stored in humans through the food chain. Such excessive intake of Cd causes severely detrimental effects on human health, including cancer diseases in kidney and breast; cardiovascular and cerebrovascular diseases like hypertension, anemia, heart failure, and cerebral infarction; and renal diseases like proteinuria and renal insufficiency [2]. It is well known that humans are exposed to Cd mainly through the consumption of crops, which account for 90% of the total dietary Cd intake [5]. Therefore, it is of great importance to develop crop cultivars with reduced Cd accumulation (especially in the edible parts like grains) to minimize the harm caused by Cd to human health. To reach this goal, it is imperative to obtain a comprehensive understanding of the mechanisms underlying Cd accumulation in plants.

Over the last decade, a great deal of research has focused on revealing the physiological and molecular mechanisms of Cd uptake and accumulation [2,6]. A series of genes associated with Cd transport in plants have been identified and functionally verified [4,6,7,8]. Meanwhile, specific plant genotypes or lines with extremely low or high Cd accumulation potential have been developed by genetic transformation and gene editing [9]. Furthermore, there is increasing evidence indicating that Cd accumulation and tolerance in some plant species might have evolved under the selection pressure for their colonization in metalliferous soils or to protect them against herbivores or pathogens [10]. However, there are still many questions far from being clearly answered, such as the following: How does Cd accumulation vary across the plant kingdom? How is the trait of Cd accumulation evolved? What is the molecular basis driving the evolution of Cd accumulation in plants? Is there any relationship between the physiological/molecular mechanisms of Cd accumulation and evolutionary adaptation? To answer these questions, the present review firstly analyzed the phylogenetic variation in Cd accumulation among plant species with the available research data. Then, it summarized the research advances concerning the physiological and molecular mechanisms of Cd accumulation in plants and highlighted the function of the critical transporters in Cd accumulation, and finally, it discussed the functional evolution of Cd accumulation in plants in terms of the most important membrane transporter families mediating Cd accumulation. All these knowledge may be helpful for predicting the transfer of Cd from soils to plants, further understanding the origins of Cd-accumulating phenotypes, and discovering the plant genetic resources for the breeding of low-Cd crops and the phytoremediation of the Cd-contaminated soils.

2. Phylogenetic Variations in Cd Accumulation in Plants

Plants have adapted to a variety of environments, from deserts to oceans. They are gathered into five major groups: algae, bryophytes, pteridophytes, gymnosperms, and angiosperms. The characteristics of Cd accumulation have been described in many crop species like rice [11], barley [12], and rapeseed [13]. All these studies demonstrate that Cd accumulation varies among different plant species and even within different genotypes of the same species [10]. To further understand the characteristics of Cd accumulation across the plant kingdom, the phylogenetic variation in Cd accumulation was summarized for different phyla based on the previously published research data.

Algae constitute important components of our environment and ecosystem as primary producers and contribute nearly 40% of the global biomass productivity. A comparative study on the biosorption of Cd by 30 strains of algae demonstrated that algae species have the ability to absorb Cd from the environment in a species-dependent manner [14]. Chandrashekharaiah et al. [15] reported that two freshwater microalgae, Chlorella pyrenoidosa and Scenedesmus acutus, showed 45.45% and 57.14% removal of Cd from initial Cd concentration of 1.5 ppm, and Chlorella pyrenoidosa exhibited higher Cd bioaccumulation (3%) capacity than Scenedesmus acutus (1.5%). In addition, the marine brown algae (Ochrophyta) Ecklonia radiata was shown to uptake 660–1634 mg/g Cd of dry biomass [16]. Such high capacity of algae for Cd²⁺ uptake might be attributed to the high alginate content in the cell wall matrix [17]. On the other hand, the ability of algae to uptake Cd²⁺ is strongly influenced by water pH. For instance, the marine brown algae Ecklonia radiata exhibited the highest uptake of Cd²⁺ at pH 4.0 (1634 mg/g Cd of dry biomass) but the lowest uptake at pH 1.5 (660 mg/g Cd of dry biomass) when the strains were grown under Cd condition at pH 1.5–6.15 [16]. However, two green algae, Pseudokirchneriella subcapitata and Scenedesmus obliquus, were found to show the highest Cd uptake capacity at pH 7.0 [18,19]. These results suggest that different algae species may require the specific optimal pH value for their uptake of Cd.

Bryophytes are among the earliest green plants to colonize the terrestrial environment and have achieved great ecological success in various habitats, from desert to aquatic and from tropical to arctic, since evolving ~500 Mya [20], and they are usually divided into three large clades: liverworts (Marchantiophyta), mosses (Bryophyta), and hornworts (Anthocerophyta) [21]. Mosses and liverworts possess the characteristics of lacking the cuticle layer and having pronounced ion-exchange abilities and large surface-to-weight ratio, which endow them with the ability to absorb heavy metal ions through the entire surface (reviewed by [22]). Thus, both mosses and liverworts are used as biomonitors or bioindicators of heavy metal pollution in both terrestrial and aquatic environments due to the characteristics mentioned above [22,23]. Nevertheless, such abilities are highly species-specific. Vukojević et al. [24] reported that two moss species, Bryum capillare and Ceratodon purpureus, could accumulate Cd up to 0.1% of their dry weight (approximately 1125–1250 mg/kg) in their shoots, whereas Macedo-Miranda et al. [25] detected a very small amount of Cd (ranging from 0.1 to 7.3 mg/kg, with an average value of 1.3 mg/kg) in two other moss species, Fabronia ciliaris and Leskea angustata [25]. In addition, the biosorption of Cd by mosses is also strongly dependent on the environmental conditions. It was found that low pH values not only reduced the extracellular binding of Cd but also inhibited the intracellular uptake of Cd in the moss species Rhytidiadelphus squarrosus [26].

Ferns and lycophytes constitute the largest vascular plant groups besides seed plants and consist of approximately 10,578 and 1338 species, respectively [27]. They have played important roles in early land plant evolution and been remarkably adapted to a wide range of environments, including both tropical and cold temperate climates, alpine and lowland regions, and aquatic and xeric conditions [28]. In general, although both ferns and lycophytes can accumulate large amounts of aluminum (Al) and arsenic (As), they accumulate much less Cd [29,30]. However, Talebi et al. [31] measured the Cd concentrations in two Azolla species (small aquatic ferns) and found that Azolla pinnata would accumulate 4673.8 mg/kg Cd and Azolla filiculoides would accumulate 2747 mg/kg Cd when exposed to 500 µM of CdCl₂ for 3 days, thereby suggesting the potential of Azolla species in Cd hyperaccumulation.

The Gymnosperm clade is ancient and widespread, and it includes four of the five main seed plant lineages, including cycads, ginkgos, gnetophytes, and conifers [32]. Gymnosperm lineages diverged from each other during the period from the Late Carboniferous to the Late Triassic (311–212 Mya), earlier than the appearance of the earliest extant angiosperms (around 300 Mya) [33]. Living gymnosperms consist only of a little more than 1000 species, with conifers (pines, cypresses, and relatives) being the largest group [34]. The available researches on Cd accumulation in gymnosperms are mostly focused on conifers [35,36,37,38,39,40,41,42]. Kim et al. [35] reported that the Cd accumulation in the shoots of Pinus sylvestris could amount to about 33.2 mg/kg when plants are treated with 10 mg/kg Cd for 11 weeks. In a subsequent study, however, another Pinus species, Pinus pinaster, was found to accumulate only 11.9 mg/kg Cd in its shoots when exposed to 15 mg/kg Cd for six months [36]. Hashemi and Farajpour [37] found that Picea abies grown near the metal production factory (Cd at the depth of 0–10 cm beneath the level of the soil: 6.8 mg/kg) accumulates 1.1 and 1.5 mg/kg Cd in its leaves and stems. On the other hand, Österås and Greger [38] found that Picea abies could accumulate up to 72.4 mg/kg and 26.1 mg/kg Cd in its barks and wood, respectively, after being exposed to 0.5 μM Cd for three months. Guo et al. [39] showed that Platycladus orientalis (containing 28.0, 69.3, and 406 mg/kg Cd in leaf, stem, and root, respectively) accumulated a 2.30-fold higher amount of Cd in leaves compared with Juniperus chinensis (containing 12.2, 49.4, and 455 mg/kg Cd in leaf, stem, and root respectively), after being exposed to 100 mg/kg of Cd for 220 days. Recently, Zeng et al. [40] reported that the growth of Platycladus orientalis was significantly inhibited in soil containing 9.6 mg/kg Cd, while it accumulated more than 2 mg/kg and approximately 41.5 mg/kg Cd in its stems and roots, respectively. Although the Cd accumulation results were not consistent between the above two studies [39,40], the authors of both studies suggest that Platycladus orientalis is a promising plant for the phytostabilization of Cd-contaminated soil. In addition, the hybrid larch was found to show an extraordinary potential for the phytoextraction of Cd from the contaminated soils to the above-ground tissues, as it could tolerate a 4-week exposure to 0.25 mM Cd (or a 1-week exposure to 1.5 mM Ca) and accumulate as much as 208–220 mg/kg Cd in its shoots [41,42].

Angiosperms first appeared in northern Gondwana during the Early Cretaceous (approximately 135 Mya) and dominated the species composition worldwide within 10–30 Mya [43]. They are the largest and most diverse group within the plant kingdom, representing approximately 80% of all known living plants and including five major groups: eudicots, monocots, magnoliids, chloranthales, and ceratophyllales [44]. Angiosperms are able to occupy any environment on earth, such as high mountaintops, deep oceans, freezing tundras, and warm, wet rainforests. It is well documented that the angiosperm ecological incursion is highly driven by environmental circumstances and biotic factors [45]. Broadley et al. [46] summarized the early records of shoot Cd content in angiosperms and found a variation of 44.3% in shoot Cd content among the 108 angiosperm species that were tested. Up to the present, approximately 20 angiosperm species (~10 families) have already been identified as Cd hyperaccumulators [10,47,48,49]. These Cd accumulators are not randomly distributed among the angiosperm families but mainly belong to the Brassicaceae, Crassulaceae, and Asteraceae families [47,49]. Two Brassicaceae species, Arabidopsis halleri and Thlaspi caerulescens, which are the most intensively studied Cd-hyperaccumulating plant species, demonstrate an extraordinary ecotypic variation (ca. 100 > 1000 µg/g among sites and populations) in terms of their Cd accumulation potential [50,51]. Likewise, the Cd hyperaccumulator Sedum alfredii exhibits fixed ecotypic differences in Cd sequestration (hyperaccumulating ecotype preferentially accumulates Cd in the pith and cortex, whereas non-hyperaccumulating ecotype restricts Cd to its vascular bundles) [52], root-to-shoot translocation (>10 times higher in hyperaccumulating ecotype than in non-hyperaccumulating ecotype) [53], and shoot Cd concentration (7.5–1090 mg/kg among different populations) [54]. These results indicate the local adaptation of the above species to natural habitats [50] or, probably, to the anthropogenic metal pollution [55].

Taken together, plants, particularly angiosperms, have developed widespread adaptations to their habitat, with contrasting Cd contents throughout the course of their evolution [46,49]. It has been proposed that the trait of Cd accumulation is broadly distributed over the plant phylogeny and has evolved independently many times and under selection [10]. Similar convergent patterns of evolution in angiosperms have also been observed in terms of the accumulation of other metals like selenium (Se) and nickel (Ni) [10,56]. Hence, the question regarding which selection pressures (for example, biotic and abiotic environmental factors as well as physiological processes) favor the evolution of Cd accumulation in plants arises [10]. It is well known that the emergence of the Cd hyperaccumulation trait in some plant species like Arabidopsis halleri, Thlaspi caerulescens, and Sedum alfredii coincided with the appearance of anthropogenic metal-polluted sites in mining regions [47,48,49]. Therefore, it is believed that the colonization of plant species in heavy metal-contaminated soils resulting from human activities represents a relatively recent event in their evolutionary history [57]. However, a comparison between Arabidopsis halleri (Cd accumulator) and Arabidopsis lyrate (non Cd accumulator) suggests that these two close relatives of the model species Arabidopsis thaliana diverged about 5 Mya or earlier, and the trait of Cd hyperaccumulation had evolved well in Arabidopsis halleri long before anthropogenic activities fostered the spread of Cd-polluted areas [58]. Therefore, Cd accumulation may have evolved in calamine outcrops prior to metal pollution by mining activities or in nonmetalliferous soils prior to the colonization of metalliferous soils and an increase in metal tolerance [57]. The other hypotheses proposed on the selective factors in Cd accumulation by plants include protection against herbivores or pathogens, allelopathy, and positive physiology effect [10].

3. Physiological Processes Mediating Cadmium Accumulation in Plants

During recent decades, great efforts have been made to elucidate the mechanisms of Cd accumulation in plants [2,4,6,7]. It is generally accepted that Cd accumulation in plants entails four processes, including (i) root Cd uptake from soil, (ii) root-to-shoot translocation via the xylem, (iii) intracellular Cd sequestration, and (iv) Cd (re)distribution in above-ground tissues like stems, leaves, and grains through the phloem.

3.1. Cd Uptake by Roots

Plants take up Cd from the external environment mainly through their roots; the rate and amount of Cd uptake by roots are dependent on the Cd bioavailability or concentration in the soil [59] and controlled by plant genetic factors. The uptake of Cd by plant roots consists of two phases: apoplastic binding and symplastic uptake [60]. In the first phase, the positively charged Cd²⁺ initiates an electrostatic interaction with the negatively charged carboxyl groups (-COOH) of polygalacturonic and hydroxycinnamic acids on cell walls, resulting in Cd accumulation in the root apoplast [61]. This phase is rapid and spontaneous, thereby indicating that it has no energy requirements (a passive system). In the second phase, Cd is symplastically taken up in an ATP-driven active process, which requires a great deal of energy and is highly dependent on the metabolic activity of plants [62]. The transport of Cd across the root cell plasma membrane is the initial step for its symplastic uptake. It is a concentration-dependent process, reflected by the saturation kinetics in the relationship between the uptake velocity and concentration of Cd in the external environment, thereby indicating that Cd is taken up via a carrier (transporter)-mediated system [63]. For instance, many transporters from the natural resistance-associated macrophage protein (NRAMPs, such as OsNRAMP1, OsNRAMP5, and AtNRAMP6) and zinc/iron-regulated transporter-like protein (ZIPs, such as AtIRT1) families are responsible for Cd transport in the root cells of Arabidopsis plants and rice [64,65,66]. In addition, Cd may also enter root cells through the transport pathway for Ca [67,68,69], because of their similarities in charge (c.a. +2) and ionic radius (Ca²⁺, 0.97 Å; Cd²⁺, 0.99 Å). Adding La³⁺ and Gd³⁺ (the potent Ca channel inhibitors) or increasing the Ca concentration in the culture solution suppresses the metabolically dependent Cd uptake substantially in the Zn hyperaccumulator Thlaspi caerulescens [60], the halophyte Suaeda salsa [68], and rice [67]. Recently, one member of the major facilitator superfamily (MFS), OsCd1, was found to be associated with root Cd uptake in rice [70]. However, although the involvement of a great number of membrane proteins in Cd uptake has been demonstrated, limited knowledge exists regarding the transport of Cd across plant root plasma membranes at the molecular level.

3.2. Root-to-Shoot Cd Translocation

After its uptake by the root epidermis or exodermis, Cd is radially transported across the cortical, endodermal, and pericycle cells and then loaded into the root xylem for its subsequent root-to-shoot translocation via the xylem [71,72,73]. Loading Cd into the root xylem is a crucial step for Cd translocation to the aerial plant parts [4,73]. The radial movement of Cd toward the root xylem occurs via its symplastic and/or apoplastic transport in the form of free Cd²⁺ or Cd-complexes with various chelates [7]. Symplastic transport is considered an energy-consuming (positive) pathway involving both influx and efflux transporters [6]. Apoplastic transport is a passive pathway, usually driven by transpiration [48]. It has been suggested that symplastic transport plays a dominant role in the radial transport of Cd [71,72]. However, Tao et al. [74] reported that the apoplastic pathway contributed up to 37% of the Cd transported to the xylem in Sedum alfredii. After crossing the root epidermis to the root cortex barrier, free Cd²⁺ or Cd chelates may enter the symplasm and are then loaded into the root xylem [3]. The xylem loading of Cd in roots is an energy-consuming process because it occurs against the membrane potential (xylem parenchyma cells, −110 to −150 mV; xylem vessels, −90 mV) [6] and is mediated by heavy metal P1B-ATPases (HMAs) and possibly also by yellow stripe-like (YSL) proteins [4,64].

3.3. Intracellular Cd Sequestration

Cd sequestration contributes significantly to Cd translocation prevention, thereby reducing Cd accumulation in cereal grains. The main Cd sequestration sites in plants are cell walls and vacuoles. The cell wall is the first line of defense against toxic metals from the external environment. When captured by root cells, metal ions are largely bound by some cell wall components, such as cellulose, hemicellulose, lignin, and pectin [8]. As a result, the highest Cd concentration in root tissues occurs in the apoplast, particularly in the walls of rhizodermal and cortical cells [64]. Such binding of Cd to cell walls can efficiently prevent 36–70% of Cd from being transported across the plasma membrane into protoplasts [75], thereby reducing the Cd translocation between cells and tissues.

Within the root cells, Cd is mainly located in vacuoles, which are generally considered to be the main storage sites for metals [76]. Wu et al. [77] found that 51% of Cd in barley roots was present in the soluble fraction of the vacuole. After entering the aerial plant parts, Cd is also mainly sequestered in the vacuoles of parenchyma cells in the leaf mesophyll, stem pith, and cortex [78,79]. Vacuolar sequestration of Cd has been demonstrated to reduce Cd concentration in the cytoplasm of roots and leaves as well as alleviate Cd toxicity in plants [80]. Several transporter families, such as HMAs, Ca²⁺ exchangers (CAXs), NRAMPs, metal tolerance proteins (MTPs), and adenosine triphosphate (ATP)-binding cassette subfamily C proteins (ABCCs), have been identified as being responsible for the vacuolar sequestration of Cd [80,81,82]. It is generally assumed that Cd²⁺ in the cytosol first forms low-molecular-weight (LMW) complexes by binding with metal ligands, such as glutathione (GSH), nicotianamine (NA), and organic and amino acids; subsequently, the complexes are transported into the vacuoles, where more Cd²⁺ and thiol-containing chelators such as phytochelatins (PCs) are incorporated to produce high-molecular-weight (HMW) complexes like Cd-PC [83]. In addition, the formation of the Cd-malate complex in vacuoles has also been found to reduce the subsequent Cd efflux from the vacuoles to the cytoplasm [84].

3.4. Cd Accumulation in Shoots and Grains

In the final stage of Cd transportation, which consists of three processes, including xylem unloading, phloem translocation, and intervascular transfer, Cd accumulates in shoots and grains (the edible parts of cereals) [6]. Cd xylem unloading by effluxing Cd from vessels to parenchyma cells is the first step for the distribution of Cd in the shoot, as well as for the re-distribution of Cd via the phloem [85]. Phloem translocation is the main pathway of Cd accumulation in plant shoots and grains. Mendoza-Cózatl et al. [86] reported that the main complexes of Cd in phloem sap are with NA, GSH, and PCs. However, the manner in which these Cd chelates are loaded into the phloem is poorly understood, and no responsible transporter has been identified up to the present [87].

Intervascular transfer in nodes is closely associated with the movement of Cd toward the grains [6]. In cereals, nodes are complex but well-organized vascular systems that consist of two major types of vascular bundles: enlarged vascular bundles (EVBs) and diffuse vascular bundles (DVBs) [6,88]. EVBs stem from the lower nodes and are connected to leaves, whereas DVBs surrounding the EVBs start at the node and are connected to the upper nodes or panicles [6]. Fujimaki et al. [89] performed a noninvasive detection of Cd in rice plants and found higher Cd concentrations in nodes than in internodes. Increasing evidence suggests that phloem loading by intervascular transfer from EVBs to DVBs in node I is a major pathway for Cd movement toward grains in rice plants [88]. In rice, OsHMA2 and low-affinity cation transporter 1 (OsLCT1) are involved in intervascular transfer, with OsHMA2 loading Cd into the phloem of EVBs and DVBs and OsLCT1 exporting Cd from phloem parenchyma cells into the sieve tubes [90,91]. However, the molecular mechanisms underlying Cd distribution in eudicot shoots are still unclear.

4. Evolution of Cadmium Transporters and Their Functions in Cadmium Accumulation

As Cd is a non-essential element for plants and interferes with the uptake of other ions, it is likely to enter plant cells through the transporters of essential elements such as Fe²⁺, Zn²⁺, and Mn²⁺, owing to the similarity between Cd²⁺ and these ions in terms of their chemical and physical properties. In recent decades, many genes conferring Cd transportation across membranes in plants have been identified and functionally verified [4,6,7,8]. These transporters belong to NRAMPs, HMAs, ZIPs (ZRT1/IRT1-like protein), cation diffusion facilitators (CDFs), the oligopeptide transporter family (OPTs), ABCCs, and CAXs. In this section, the evolution of these transporters and their function in plant Cd accumulation are extensively discussed and summarized.

4.1. Evolution of Cd Transporters

The evolution of membrane transporters may have played an important role in the adaptation of plants to metalliferous environments. The ancestors of modern land plants colonized the terrestrial habitat approximately 500–470 Mya. Since then, dramatic changes, including large fluctuations in water availability, illumination, light intensity, temperature, and the carbon-dioxide-to-oxygen concentration ratio, have occurred in the living environments of land plants compared to the aquatic environments of seagrasses grown in the ocean [92]. Have the membrane transporters of plants experienced any persistent adaptive and stepwise evolution for their adaptation to changing environments? To answer this question, a comparative genomics analysis of NRAMP, HMA, ABCC, ZIP, CDF, CAX, and OPT families was performed using 41 plant species ranging from rhodophytes to eudicots. The results showed that the vascular plants (including lycophytes, ferns, gymnosperms, and angiosperms) contain many more Cd transporter family members in their genomes than do algae and bryophytes (liverworts and mosses) (

Table 1. Number of predicted and published cadmium (Cd) transporter families in 41 plant and algal species.

Clade	Plant Species	NRAMPs	HMAs	ZIPs	CDFs	ABCCs	CAXs	OPTs
Eudicots	Arabidopsis thaliana	6	8	15	12	15	11	9
	Brassica rapa	9	15	28	18	17	12	11
	Gossypium raimondii	10	9	21	15	26	10	8
	Theobroma cacao	6	7	12	13	15	6	10
	Eucalyptus grandis	9	9	20	18	33	5	25
	Malus domestica	10	17	16	26	24	12	18
	Medicago truncatula	7	10	13	13	38	8	13
	Glycine max	13	18	19	23	37	17	13
	Populus trichocarpa	8	13	16	22	24	8	13
	Vitis vinifera	6	5	13	11	23	5	11
	Solanum lycopersicum	4	8	11	12	14	7	8
	Chenopodium quinoa	10	18	18	15	45	6	10
Monocots	Spirodela polyrhiza	3	8	9	11	12	7	6
	Zostera marina	5	7	13	9	10	6	4
	Phoenix dactylifera	10	21	14	19	32	16	11
	Triticum aestivum	18	26	25	19	53	15	22
	Hordeum vulgare	7	11	12	8	16	9	9
	Brachypodium distachyon	7	9	11	10	20	8	8
	Phyllostachys heterocycla	7	9	11	10	11	10	9
	Zea mays	8	12	10	11	12	11	8
	Sorghum bicolor	8	11	12	8	16	10	8
	Oryza sativa	7	8	11	9	16	7	9
Basal angiosperms	Amborella trichopoda	3	7	8	8	14	4	8
Gymnosperms	Pinus taeda	13	8	11	8	11	11	34
	Pinus lambertiana	9	11	13	10	10	7	11
	Picea abies	5	5	10	2	6	4	9
Ferns	Azolla filiculoides	9	18	11	16	22	11	10
	Salvinia cucullata	2	9	3	7	16	2	3
Lycophytes	Selaginella moellendorffii	6	12	5	8	23	2	6
Mosses	Physcomitrella patens	6	18	7	12	15	6	2
	Sphagnum fallax	6	8	5	9	16	4	11
Liverworts	Marchantia polymorpha	5	6	5	5	15	3	5
Streptophytes	Mesotaenium endlicherianum	3	7	1	3	8	0	0
	Spirogloea muscicola	9	19	3	17	11	3	0
	Chara braunii	1	3	2	2	2	2	0
	Klebsormidium flaccidum	3	6	2	5	3	1	0
Chlorophyta	Chlamydomonas reinhardtii	2	4	2	4	4	3	0
	Volvox carteri	2	4	0	1	4	3	0
	Ostreococcus sp.	1	5	1	3	2	2	0
Rhodophyta	Cyanidioschyzon merolae	2	1	1	1	1	0	0
	Porphyra yezoensis	1	3	0	2	2	1	0

Note: Numbers are based on both literature search and bioinformatic analysis. Query Cd transporter genes are listed in Table S1. Genome sequence data were downloaded from the oneKP database. Genesis software (Version 2.0) was used to estimate the similarity between protein sequences. Candidate protein sequences were selected by BLASTP searches that satisfied the E value < 10⁻¹⁰ and query coverage > 50% criteria. NRAMPs: natural resistance-associated macrophage proteins; HMAs: heavy metal ATPases; ZIPs: zinc/iron-regulated transporter-like proteins (ZRT1/IRT1-like protein); CDFs: cation diffusion facilitators; OPTs: oligopeptide transporter family; ABCCs: adenosine triphosphate-binding cassette subfamily C proteins; CAXs: cation/H⁺ exchangers.

), implying that the Cd transporter families may have undergone a lineage-specific expansion as a result of gene duplication (whole-genome duplication or duplications of large chromosomal regions) and/or tandem duplication [93,94,95]. Such expansion of membrane transporter families in vascular plants could have provided an adaptive advantage for the colonization of new habitats, such as metalliferous soils, before significant vascular development occurred in early land plants. Furthermore, most Cd transporter families except the OPT family can be identified across all plant species (Table 1), suggesting that these Cd transporter families seem to be highly conserved throughout the evolutionary lineage of the entire plant kingdom, thereby indicating an evolutionarily conserved function of them in the metal homeostasis of these transporters (Figure 1). However, the OPT protein family was not detected in all algal species (Table 1). Hanikenne et al. [96] attempted to find the YS1-like proteins in the genome sequences of green alga Chlamydomonas reinhardtii and red alga Cyanidioschizon merolae but failed to identify any homologs. Likewise, a previous phylogenetic analysis of 325 OPT family members, ranging from prokaryotes to eukaryotes, revealed that OPT family members in eukaryotes were found only in fungi and land plants [97]. Thus, it may be suggested that the OPT family evolved after the emergence of land plants.

In order to understand the origin of Cd transporters, a further phylogenetic analysis was performed on the orthologs of two famous Cd transporters, NRAMP5 and HMA2, across the entire plant kingdom using the One Thousand Plant Transcriptome (oneKP) database [98]. It is interesting that these two transporters underwent significantly different evolution courses. In the phylogenetic analysis, the orthologs of NRAMP5 proteins were grouped into two clusters (Figure 2A), with cluster 1 including green algae, Glaucophyta, Rhodophyta, and mosses, and cluster 2 containing green algae, hornworts, liverworts, lycophytes, ferns, gymnosperms, and angiosperms. Furthermore, cluster 2 was deeply branching. In the three sub-clusters of cluster 2, green algae formed a distinct sub-sub-cluster. The other two sub-clusters included lycophytes, ferns, and gymnosperms, while only the last sub-cluster included hornworts, liverworts, and angiosperms. In addition, the plants were clearly divided into two groups, monocots and eudicots, containing their specific ancestral gymnosperms and basal angiosperms (Figure 2A). It could be suggested that the orthologs of NRAMP5 in plants descended from a polyphyletic evolutionary lineage that originated in different ancestors. Furthermore, the divergent presence of green algae, lycophytes, ferns, and gymnosperms in the phylogenetic tree indicates a rampant occurrence of horizontal gene transfer during the evolution of the NRAMP5 orthologs, which has been previously evidenced in studies on the evolution of NRAMPs in bacteria [99] and OPTs in plants [97]. However, further studies are necessary to provide insight into the molecular mechanisms and adaptive roles of horizontal gene transfer events in the evolution of Cd transporters in plants. The evolution course of HMA2 is much simpler. The orthologs of HMA2 from algae, mosses, ferns, lycophytes, gymnosperms, and angiosperms formed a distinct cluster, but they were all basal to Rhodophyta species Rhodochaete parvula (Figure 2B), thereby indicating that this transporter experienced an early evolution in plants. The angiosperm cluster was grouped into two sub-clusters consisting of monocots and eudicots, all of which were basal to Amborella trichopoda, which is the only living species in the sister lineage to all other flowering plants [100]. In addition, eudicots had a closer orthologous relationship with Myristica fragrans (Figure 2B). These results indicate that the origin of the HMA2 transporter in both monocots and eudicots can be traced back to Amborella trichopoda, which dates back to approximately 130 Mya. However, these species evolved separately thereafter and formed a monophyletic evolutionary lineage.

4.2. Function of Cd Transporters

The functions of NRAMP, HMA, ABCC, ZIP, CDF, CAX, and OPT transporters, including uptake, translocation, sequestration, and distribution of Cd in plants, and their tissue-specific localization and substrate specificity are summarized in Table 2.

4.2.1. NRAMPs

NRAMPs represent a family of metal transporters that are located at the membranes of root cells and are evolutionarily conserved in a wide range of organisms, including bacteria, fungi, plants, and animals [95]. In plants, NRAMP genes participate in the uptake of divalent cations, such as Fe²⁺, Mn²⁺, Cu²⁺, Zn²⁺, and Cd²⁺ (Table 2). In Arabidopsis plants, six NRAMP family members have been identified (Table 1), with AtNRAMP3 and AtNRAMP4 being localized at the tonoplast and responsible for the Cd²⁺ efflux from vacuoles to the cytosol [103]. In rice, OsNRAMP1 and OsNRAMP5 are plasma membrane (PM)-localized transporters involved in taking up Cd from the external solution to root cells [65,104]. The knockdown or CRISPR/Cas9-mediated editing of OsNRAMP5 causes a dramatic reduction in Cd and Mn concentrations in both rice roots and shoots [65,105]. However, OsNRAMP3, OsNRAMP4, and OsNRAMP6 have no Cd transportation capacity [106,107,108]. In barley, HvNRAMP5, which is 84% identical with OsNRAMP5, can also mediate Cd uptake [109]. In the hyperaccumulator Noccaea caerulescens, NcNRAMP1 is one of the main transporters involved in the influx of Cd²⁺ across the endodermal PM, thus playing a key role in Cd²⁺ influx into the stele and contributing to Cd root-to-shoot transport [110].

4.2.2. HMAs

HMAs, also known as P1B-ATPases, are involved in transporting cations across the membrane by consuming energy from ATP hydrolysis [111]. Eight HMA members have been identified in Arabidopsis plants, with AtHMA1-AtHMA4 being able to transport divalent cations such as Cd²⁺, Zn²⁺, and Pb²⁺ [111]. Of these eight HMAs, AtHMA2 and AtHMA4 are predominately expressed in the tissues surrounding the vascular vessels of roots, and they mediate Cd²⁺ efflux from xylem parenchyma cells to xylem vessels, which is necessary for the root-to-shoot Cd translocation [112]. Hyperaccumulators Thlaspi caerulescens, Sedum plumbizincicola, and Arabidopsis halleri have much higher HMA4 gene copy numbers and transcript levels than Arabidopsis thaliana [113,114,115]; additionally, HMA4 is a candidate gene for determining the evolution of the Cd hyperaccumulator phenotype [114]. In rice, OsHMA2 is localized mainly at the PM of the root cells, and it has been proven to play a crucial role in Cd xylem loading and root-to-shoot translocation. The loss of its function decreases significantly the Cd accumulation in leaves and grains [116,117]. OsHMA3 is a tonoplast-localized transporter and involved in restricting Cd translocation by mediating Cd sequestration into the vacuoles [118]. A loss-of-function allele of OsHMA3 could cause high Cd accumulation in rice shoots and grains [119], whereas its overexpression presents a great opportunity to produce Cd-free rice through reducing the Cd concentration in brown rice by 94–98% [120]. In addition, several recent studies have suggested that the natural variation in the promoter or coding region of HMA3 contributes to the genotypic difference in Cd accumulation in rice and Brassica rapa [121].

4.2.3. ZIPs

ZIP family members are generally involved in Cd uptake and translocation in plants [122]. IRT1 is the first identified member of the ZIP family in Arabidopsis plants and participates in the uptake of Fe²⁺, Zn²⁺, Cu²⁺, Ni²⁺, and Cd²⁺ from the soil [123]. Seventeen ZIP transporters have been identified in rice. OsIRT1, which is highly homologous to AtIRT1, is predominantly expressed in roots and up-regulated by Fe deficiency and Cd exposure [124]. Recently, Zheng et al. [125] demonstrated a distinct difference between Arabidopsis and rice plants in terms of their respective expression profiles of ZIP genes in response to Cd stress. In addition, other ZIP transporters, such as OsZIP1 and OsZIP3, have also been shown to be involved in Cd uptake in rice [125]. It is noteworthy that the involvement of ZIP genes in Cd uptake has also been detected in hyperaccumulators. In Noccaea caerulescens, NcZNT1, a homolog of AtZIP4, mediates low-affinity Cd uptake when expressed in Saccharomyces cerevisiae ZHY3 cells [126]. Recently, it was reported that NcZNT1 is a PM-localized Zn²⁺/Cd²⁺ transporter and its promoter is mainly active in the cortical, endodermal, and pericycle cells of Noccaea caerulescens [127].

4.2.4. CDFs

CDF proteins, also known as MTPs, are a family of heavy metal transporters involved in the transport of Zn²⁺, Cd²⁺, and Co²⁺ [128]. They have been identified in diverse organisms, including bacteria, fungi, animals, and plants. The CDFs of plant cells, generally known as MTPs [129], are known to mediate the heavy metal efflux from the cytoplasm either to the extracellular space or to vacuoles and organelles [130]. MTPs consist of seven phylogenetic groups, with Zn-CDFs in groups 1 (MTP1-MTP4), 5 (MTP5), and 12 (MTP12); Fe/Zn-CDFs in groups 6 (MTP6) and 7 (MTP7); and Mn-CDFs in groups 8 (MTP8) and 9 (MTP9-MTP11) [131]. In rice, OsMTP1 has been demonstrated to be a PM-localized transporter involved in the translocation of Cd and other heavy metals in both roots and shoots [132]. Other MTPs, such as TgMTP1 from Thlaspi goesingense [133], CsMTP1 and CsMTP4 from Cucumis sativus [134], and CitMTP1 from Citrus sinensis [129], have been proven to be involved in the sequestration of Cd into vacuoles or its efflux from root cells.

4.2.5. OPTs

The OPT family, which contains YSL transporters, is involved in transporting metal–nicotianamine (NA) complexes through the plant cell membrane. Thus, when Cd²⁺ is chelated, it can be taken up through the OPT or YSL proteins [135]. In order to enhance the availability of metal ions in the rhizosphere, plant roots secrete LMW organic acids, such as mugineic acids (Mas) and phytosiderophores (PS), to form metal–ligand complexes, which are then transported by YSL transporters [135]. This strategy is very efficient, as it facilitates the uptake of Fe from Fe-deficient soils by some Poaceae species [136]. In addition, YSLs also play an important role in Cd transport. In Zea mays, ZmYS1 has been suggested to transport the Cd-PS and Cd-NA complexes at a low rate [66]. Two ZmYS1 orthologs isolated from rice and Cd hyperaccumulator Solanum nigrum, OsYSL2 and SnYSL3, have been found to transport Cd-NA complexes when expressed heterologously in yeast [137,138].

4.2.6. ABCCs

The family of ABC proteins is one of the largest in organisms [139] and has various substrates, including carbohydrates, lipids, xenobiotics, antibiotics, drugs, and heavy metals [140]. Unlike NRAMP3/4, HMA3, and CAX2/4, which transport free Cd²⁺ ions, ABCCs transport Cd-PC complexes [141]. In Arabidopsis plants, AtABCC1 and AtABCC2 are responsible for the transport of Cd-PCs into the vacuoles [82]. Likewise, AtABCC3 has been suggested to mediate the transport of Cd-PC complexes [142,143]. Apart from ABCCs, other ABC transporters also confer Cd tolerance. AtABCG36/AtPDR8, a member of the pleiotropic drug resistance (PDR) subfamily of ABC transporters in Arabidopsis plants, has been suggested to play a role in Cd tolerance by pumping Cd²⁺ or Cd complexes out of root epidermal cells [144]. AtATM3, which belongs to the mitochondria subfamily of Arabidopsis ABC proteins, contributes to Cd tolerance by mediating the transport of glutamine synthetase-conjugated Cd across the mitochondrial membrane [145]. In rice, OsABCG36/OsPDR9 has been recently demonstrated to be involved in Cd tolerance through exporting Cd²⁺ or Cd conjugates from the root cells [146].

4.2.7. CAXs

CAXs are tonoplast-localized transporters that export cations out of the cytosol to maintain ion homeostasis across biological membranes [147]. Most CAX members are calcium (Ca²⁺)-specific. However, two CAXs identified in Arabidopsis plants, AtCAX1 and AtCAX2, have been found to be capable of pumping Ca²⁺ and other cations such as Cd²⁺, Zn²⁺, and Mn²⁺ into vacuoles [81,148]. Wu et al. [149] reported that the ectopic expression of AtCAX1 in petunia plants increased significantly their Cd tolerance and accumulation. In the Cd hyperaccumulator Arabidopsis halleri, Cd tolerance is highly associated with the expression of AhCAX1 [150], suggesting an involvement of AhCAX1 in conferring Cd tolerance in this plant. SaCAX2, a CAX2-like protein in the hyperaccumulator Sedum alfredii, confers Cd tolerance and accumulation when it is expressed heterologously in yeast and tobacco [151].

4.2.8. Other Transporters Involved in the Uptake and Intervascular Transfer of Cd

In addition to the well-known transporter families like NRAMPs, HMAs, ABCCs, ZIPs, CDFs, CAXs and OPTs, there are other transporters also described to be involved in Cd transport. OsLCT1, a PM-localized protein that is strongly expressed in nodes and leaf blades during the reproductive stage of rice, mediates the efflux of Cd²⁺ from phloem into grains [90]. It has been estimated that the RNA interference-mediated knockdown of OsLCT1 resulted in 30–50% reduction in grain Cd of rice plants grown in contaminated soil [90]. Recently, OsCd1, a member of the major facilitator superfamily, was shown to be involved in the root uptake and grain accumulation of Cd in rice [70]. Furthermore, phylogenetic analysis using 950 rice accessions demonstrated the natural variation in OsCd1 diverged between indica (genotype OsCd1^D449) and japonica (genotype OsCd1^V449) subspecies, and genotype OsCd1^V449 displayed a relatively lower Cd accumulation in rice grain than the genotype OsCd1^D449, suggesting the great potential of OsCd1^V449 in the breeding of low-Cd rice [70]. In addition, Luo et al. [152] found that Cd accumulation in leaf 1 (CAL1), which encodes a defensin-like protein, is preferentially expressed in root exodermis and xylem parenchyma cells and acts by chelating Cd in the cytosol and facilitating Cd secretion into the apoplast, thus reducing the cytosolic Cd concentration.

Given the similarities in charge and ionic radius between Ca²⁺ and Cd²⁺, it is possible for Cd²⁺ to enter the plant symplast via passive transport through channel proteins transporting Ca²⁺ [67,68,153]. Indeed, several kinds of calcium-permeable channels, such as depolarization-activated calcium channels (DACCs), hyperpolarization-activated calcium channels (HACCs), and voltage-insensitive cation channels (VICCs), are capable of transporting Cd²⁺. Increasing evidence obtained using channel blockers and flux measurements confirms the effects of Ca on Cd uptake and accumulation in plants [67,68,153]. However, the function of these channels in the facilitation of Cd transport is poorly understood. In a recent study on rice, the expression of the genes belonging to the two Ca channel families, annexins and glutamate receptors (GLRs), was shown to co-segregate with Cd²⁺ influx and uptake by root cells [67], thereby suggesting the possibility of identifying the candidate channels responsible for Cd transport.

Table 2. The identified transporters mediating cadmium (Cd) uptake, translocation, sequestration, and distribution in plants.

Family	Gene Symbol	Expression Organ and Localization	Plant Species	Possible Properties	References
NRAMPs	AtNRAMP3/4	root, leaf (tonoplast)	Arabidopsis thaliana	Cd, Fe, Mn	[103,154]
	NcNRAMP1	root, shoot (PM, tonoplast)	Noccaea caerulescens	Cd	[110]
	TcNRAMP3/4	root, shoot (tonoplast)	Thlaspi caerulescens	Cd, Fe, Mn	[155]
	OsNRAMP1	root, shoot (PM)	Oryza sativa	Cd	[104]
	OsNRAMP5	root (PM)	Oryza sativa	Cd, Fe, Mn	[156,157]
	HvNRAMP5	root (PM)	Hordeum vulgare	Cd, Mn	[109]
HMAs	AtHMA1	root, shoot (chloroplast envelope)	Arabidopsis thaliana	Cd, Zn, Cu	[158,159]
	AtHMA3	root, collar, leaf (tonoplast)	Arabidopsis thaliana	Cd, Zn, Pb, Co	[160,161,162]
	AtHMA2/4	root, stem, leaf (PM)	Arabidopsis thaliana	Cd, Zn	[163,164,165,166]
	OsHMA2	root (PM)	Oryza sativa	Cd, Zn	[91,117]
	OsHMA3	root (tonoplast)	Oryza sativa	Cd, Zn	[118,167,168]
	OsHMA9	vascular bundle and anther (PM)	Oryza sativa	Cd, Cu, Zn, Pb	[169]
	TaHMA2	root, shoot (PM)	Triticum aestivum	Cd, Zn	[170]
	GmHMA3	root (ER)	Glycine max	Cd, Zn	[171]
	TcHMA3	root, shoot (tonoplast)	Thlaspi caerulescens	Cd	[172]
	SaHMA3	root, shoot (tonoplast)	Sedum alfredii	Cd	[173]
	SpHMA3	root, shoot (tonoplast)	Sedum plumbizincicola	Cd, Zn	[115]
ABCCs	AtABCC1/2	root, shoot (tonoplast)	Arabidopsis thaliana	Cd-PC, Hg-PC, As(III)-PC	[82,174]
	AtABCC3	root, shoot (tonoplast)	Arabidopsis thaliana	Cd-PC	[143]
	AtPDR8	root, shoot (PM)	Arabidopsis thaliana	Cd	[144]
	OsABCG36	root, shoot (PM)	Arabidopsis thaliana	Cd	[146]
CDFs	OsMTP1	root, leaf (tonoplast)	Oryza sativa	Cd, Ni, Fe	[132]
	TgMTP1	root, leaf (tonoplast)	Thlaspi goesingense	Cd, Zn, Co, Ni	[133,175]
	CitMTP1	root, leaf (tonoplast)	Citrus sinensis	Cd, Zn, Mn, Cu	[129]
	CsMTP1/4	root, hypocotyl, cotyledon, petiole, leaf (tonoplast)	Cucumis sativus	Cd, Mn, Zn	[134]
OPTs	ZmYS1	leaf blade and sheath, crown, seminal root (PM)	Zea mays	Cu, Ni, Cd, Fe, Zn, Mn	[66]
	OsYSL2	shoot phloem (PM)	Oryza sativa	Fe(II)-NA, Mn-NA, Cd-NA	[138,176]
	AtOPT3	root, shoot (PM)	Arabidopsis thaliana	Cd, Zn, Fe	[177]
	SnYSL3	root, shoot (PM)	Sedum nigrum	Fe(II)-NA, Mn-NA, Cd-NA	[137]
ZIPs	OsZIP1	root, shoot (ER, PM)	Oryza sativa	Cd, Zn, Cu	[178]
	TcIRT1	root (PM)	Thlaspi caerulescens	Cd, Zn, Fe(II, III); Mn	[179]
	TcZNT1	root, shoot (PM)	Thlaspi caerulescens	Cd, Zn	[126,127]
CAXs	AtCAX2/4	root (tonoplast)	Arabidopsis thaliana	Cd, Zn, Mn	[81,148]
	AhCAX1	root, shoot (tonoplast)	Arabidopsis halleri	Cd	[150]
	SaCAX2	root, shoot (tonoplast)	Sedum alfredii	Cd	[151]
Others	OsLCT1	leaf, node, phloem parenchyma (PM)	Oryza sativa	Cd	[90,180]
	OsLCD	root, shoot (cytoplasm, nucleus)	Oryza sativa	Cd	[181]
	OsCd1	root (PM)	Oryza sativa	Cd	[70]
	CAL1	root, leaf sheath, internode (CW)	Oryza sativa	Cd	[152]

Note: PS: phytosiderophore; PC: phytochelatin; NA: nicotianamine; PM: plasma membrane; ER: endoplasmic reticulum; CW: cell wall; NRAMPs: natural resistance-associated macrophage proteins; HMAs: heavy metal P1B-ATPases; ABCCs: adenosine triphosphate-binding cassette subfamily C proteins; ZIPs: zinc/iron-regulated transporter-like proteins; CDFs: cation diffusion facilitators; CAXs: Ca²⁺ exchangers; OPTs: oligopeptide transporters.

Table 2. The identified transporters mediating cadmium (Cd) uptake, translocation, sequestration, and distribution in plants.

Family	Gene Symbol	Expression Organ and Localization	Plant Species	Possible Properties	References
NRAMPs	AtNRAMP3/4	root, leaf (tonoplast)	Arabidopsis thaliana	Cd, Fe, Mn	[103,154]
	NcNRAMP1	root, shoot (PM, tonoplast)	Noccaea caerulescens	Cd	[110]
	TcNRAMP3/4	root, shoot (tonoplast)	Thlaspi caerulescens	Cd, Fe, Mn	[155]
	OsNRAMP1	root, shoot (PM)	Oryza sativa	Cd	[104]
	OsNRAMP5	root (PM)	Oryza sativa	Cd, Fe, Mn	[156,157]
	HvNRAMP5	root (PM)	Hordeum vulgare	Cd, Mn	[109]
HMAs	AtHMA1	root, shoot (chloroplast envelope)	Arabidopsis thaliana	Cd, Zn, Cu	[158,159]
	AtHMA3	root, collar, leaf (tonoplast)	Arabidopsis thaliana	Cd, Zn, Pb, Co	[160,161,162]
	AtHMA2/4	root, stem, leaf (PM)	Arabidopsis thaliana	Cd, Zn	[163,164,165,166]
	OsHMA2	root (PM)	Oryza sativa	Cd, Zn	[91,117]
	OsHMA3	root (tonoplast)	Oryza sativa	Cd, Zn	[118,167,168]
	OsHMA9	vascular bundle and anther (PM)	Oryza sativa	Cd, Cu, Zn, Pb	[169]
	TaHMA2	root, shoot (PM)	Triticum aestivum	Cd, Zn	[170]
	GmHMA3	root (ER)	Glycine max	Cd, Zn	[171]
	TcHMA3	root, shoot (tonoplast)	Thlaspi caerulescens	Cd	[172]
	SaHMA3	root, shoot (tonoplast)	Sedum alfredii	Cd	[173]
	SpHMA3	root, shoot (tonoplast)	Sedum plumbizincicola	Cd, Zn	[115]
ABCCs	AtABCC1/2	root, shoot (tonoplast)	Arabidopsis thaliana	Cd-PC, Hg-PC, As(III)-PC	[82,174]
	AtABCC3	root, shoot (tonoplast)	Arabidopsis thaliana	Cd-PC	[143]
	AtPDR8	root, shoot (PM)	Arabidopsis thaliana	Cd	[144]
	OsABCG36	root, shoot (PM)	Arabidopsis thaliana	Cd	[146]
CDFs	OsMTP1	root, leaf (tonoplast)	Oryza sativa	Cd, Ni, Fe	[132]
	TgMTP1	root, leaf (tonoplast)	Thlaspi goesingense	Cd, Zn, Co, Ni	[133,175]
	CitMTP1	root, leaf (tonoplast)	Citrus sinensis	Cd, Zn, Mn, Cu	[129]
	CsMTP1/4	root, hypocotyl, cotyledon, petiole, leaf (tonoplast)	Cucumis sativus	Cd, Mn, Zn	[134]
OPTs	ZmYS1	leaf blade and sheath, crown, seminal root (PM)	Zea mays	Cu, Ni, Cd, Fe, Zn, Mn	[66]
	OsYSL2	shoot phloem (PM)	Oryza sativa	Fe(II)-NA, Mn-NA, Cd-NA	[138,176]
	AtOPT3	root, shoot (PM)	Arabidopsis thaliana	Cd, Zn, Fe	[177]
	SnYSL3	root, shoot (PM)	Sedum nigrum	Fe(II)-NA, Mn-NA, Cd-NA	[137]
ZIPs	OsZIP1	root, shoot (ER, PM)	Oryza sativa	Cd, Zn, Cu	[178]
	TcIRT1	root (PM)	Thlaspi caerulescens	Cd, Zn, Fe(II, III); Mn	[179]
	TcZNT1	root, shoot (PM)	Thlaspi caerulescens	Cd, Zn	[126,127]
CAXs	AtCAX2/4	root (tonoplast)	Arabidopsis thaliana	Cd, Zn, Mn	[81,148]
	AhCAX1	root, shoot (tonoplast)	Arabidopsis halleri	Cd	[150]
	SaCAX2	root, shoot (tonoplast)	Sedum alfredii	Cd	[151]
Others	OsLCT1	leaf, node, phloem parenchyma (PM)	Oryza sativa	Cd	[90,180]
	OsLCD	root, shoot (cytoplasm, nucleus)	Oryza sativa	Cd	[181]
	OsCd1	root (PM)	Oryza sativa	Cd	[70]
	CAL1	root, leaf sheath, internode (CW)	Oryza sativa	Cd	[152]

Note: PS: phytosiderophore; PC: phytochelatin; NA: nicotianamine; PM: plasma membrane; ER: endoplasmic reticulum; CW: cell wall; NRAMPs: natural resistance-associated macrophage proteins; HMAs: heavy metal P1B-ATPases; ABCCs: adenosine triphosphate-binding cassette subfamily C proteins; ZIPs: zinc/iron-regulated transporter-like proteins; CDFs: cation diffusion facilitators; CAXs: Ca²⁺ exchangers; OPTs: oligopeptide transporters.

5. Conclusions and Perspectives

Cd is one of the most hazardous and toxic heavy metals in the environment and can lead to Cd-related diseases such as cancer, renal tubular dysfunction, and bone disease in humans. Cd accumulation in plants is controlled by both genetic (such as the expression level and the natural variation in Cd transporters) and environmental factors (such as pH, organic matter content, redox potential, water content, etc., in soils) that affect the entire process, including Cd uptake from the soil, root-to-shoot translocation, sequestration, and (re)distribution in shoots. In essence, all these processes are governed by membrane metal transporters, including NRAMPs, HMAs, ZIPs, CDFs, OPTs, ABCCs, CAXs, and some other transporters or channels such as OsLCT1, OsCd1, CAL1, annexin, and GLRs. Cd accumulation varies among different plant species and even within different genotypes of the same species across the plant kingdom. To gain insights into the functional evolution of Cd accumulation in plants, we performed a comparative genomics analysis of the above Cd transporter gene families using 41 plant species and a further phylogenetic analysis across the entire plant kingdom using the oneKP database. Our analysis suggests that Cd accumulation in plants is a derived and polyphyletic trait that has evolved convergently several times. During the course of evolution from algae to angiosperms, membrane transporter families, such as NRAMPs, HMAs, ABCCs, ZIPs, CDFs, and CAXs, have been conserved throughout the evolutionary lineage of entire plant kingdom (ranging from Rhodophytes to eudicots), thereby indicating that their functions are evolutionarily conserved for metal homeostasis. However, the OPT protein family is absence in all algal species of Streptophytes, Chlorophyta and Rhodophyta, thereby suggesting that this family may have evolved after the emergence of land plants. Moreover, the genomes of vascular plants contain many more family members of NRAMPs, HMAs, ZIPs, CDFs, OPTs, ABCCs, CAXs, and OPTs than do algae and bryophytes, thereby suggesting that these transporter families underwent lineage-specific expansion, which could have been conferred by gene duplication owing to segmental duplication and/or tandem duplication. Furthermore, the orthologs of HMA2 (a transporter involved in intervascular transfer of Cd) in plants originated from an early common ancestor (c.a. Rhodophyta) and experienced a monophyletic evolutionary lineage from algae to angiosperms. On the contrary, the orthologs of NRAMP5 evolved from a polyphyletic evolutionary lineage with different ancestors. In addition, different phylogenetic clusters of NRAMP5 showed the divergent presence of green algae, lycophytes, ferns and gymnosperms, indicating a rampant occurrence of horizontal gene transfer during the evolution of NRAMP5. All these evolutionary patterns may provide an adaptive advantage for plants to colonize new habitats, such as metalliferous soil. Unfortunately, the oneKP database lacks specific transcriptomic data on the known Cd hyperaccumulators, and the extent to which such evolutionary patterns of membrane transporters contribute to Cd (hyper)accumulation in plants remains unclear. Ideally, the non-accumulator, accumulator, and hyperaccumulator plants should be compared comprehensively, in terms of genome sequencing, ecological distribution patterns, and their Cd uptake and transport abilities, so as to fully elucidate the evolutionary mechanisms associated with the accumulation and adaptive response of Cd in plants.