Vvmrp1, Vvmt1, and Vvmt2 Co-Expression Improves Cadmium Tolerance and Reduces Cadmium Accumulation in Rice

Abstract

Cadmium (Cd) contamination in agricultural soils severely threatens rice production and food safety. To address this issue, this study developed transgenic rice lines co-expressing three Vitis vinifera genes: the ABCC transporter Vvmrp1 and metallothioneins Vvmt1 and Vvmt2. AlphaFold computational modeling confirmed the conserved ABCC-type transporter domain in VvMRP1. Under hydroponic conditions, transgenic rice showed remarkable Cd tolerance, surviving 30 mM Cd (lethal to wildtype, WT) without growth penalties, and exhibited 62.5% survival at 1 mM Cd vs. complete wild-type mortality. Field-relevant Cd exposure (1 mM) reduced Cd accumulation to 35.8% in roots, 83% in stems, and 76.8% in grains compared to WT. Mechanistic analyses revealed that Vvmrp1 mediates cellular Cd efflux while Vvmt1 and 2 chelate free Cd ions, synergistically inhibiting Cd translocation. Transgenic plants also maintained better Fe, P, and Mg homeostasis under Cd stress. This study pioneers the co-expression of a transporter with metallothioneins in rice, demonstrating their complementary roles in Cd detoxification without pleiotropic effects from endogenous gene modification. The findings provide an effective genetic strategy for cultivating low-Cd rice in contaminated soils, offering significant implications for food safety and sustainable agriculture.

1. Introduction

Cd, a naturally occurring heavy metal element, is ubiquitously distributed in the natural environment. Under natural conditions, soil Cd concentrations typically remain at relatively low levels, with global soil Cd mass fractions ranging from 0.01 to 2.00 mg/kg with a median value of 0.35 mg/kg. However, the expansion of industrial and agricultural activities has led to substantial increases in soil Cd content in certain regions due to anthropogenic discharges such as wastewater and effluent [1]. This phenomenon has resulted in a continuous global expansion of Cd-contaminated land areas, with agricultural soil contamination being particularly pronounced [2,3,4].

Cd exerts varying degrees of detrimental effects on rice plants across germination, seedling growth, and maturation stages. Cd significantly inhibits seed vigor and root elongation in seedlings. With prolonged Cd uptake, rice seedlings exhibit a marked decline in root biomass, fresh weight, plant height, and chlorophyll content in leaves. Moreover, Cd exposure during the seedling stage suppresses tiller number, plant height, and dry weight accumulation. These physiological impairments collectively reflect Cd-induced phytotoxicity in critical developmental phases of rice [5]. In paddy systems, excessive soil Cd not only reduces rice productivity but also induces Cd enrichment in grains. This contamination propagates through food chains, inducing multi-organ damage in humans, including renal, pulmonary, hepatic, testicular, and neurological impairments, as well as hematological disorders [5,6]. Recognized as a group 1 carcinogen by the WHO, the maximum tolerable daily Cd intake for adults is strictly limited to 70 micrograms [7]. Therefore, developing Cd-tolerant rice cultivars with low Cd accumulation not only mitigates cadmium-induced phytotoxicity but also enables safe rice production in contaminated fields, representing a crucial strategy to address soil Cd pollution and ensure food security.

In addition to screening existing rice varieties for natural low-Cd traits, researchers primarily employ genetic engineering approaches to develop low-Cd rice cultivars. These efforts focus on manipulating genes encoding transport proteins associated with Cd uptake and translocation. Since Cd lacks specific transporters in plants [8], it relies on competitive utilization of divalent metal ion transporters to facilitate its movement from the epidermis to vascular tissues and from roots to shoots [9]. Key transporter families involved in Cd stress responses include: zinc/iron-regulated transporter-like proteins (ZIPs), heavy metal ATPases (HMA), cation diffusion facilitators (CDFs), natural resistance-associated macrophage proteins (NRAMPs) and ABC transporters. Most of these genes exhibit positive regulatory effects on Cd accumulation. Targeted suppression or knockout of these genes through CRISPR/Cas9 or RNA interference technologies can significantly reduce Cd uptake and translocation in rice.

Among the seven NRAMP family members identified in rice (Oryza sativa L.), three isoforms—OsNRAMP1, OsNRAMP4, andOsNRAMP5—demonstrate functional involvement in Cd uptake and translocation [10,11]. Overexpression of these genes elevates Cd accumulation in rice tissues, whereas their knockout mutants exhibit 30–80% reductions in grain Cd content, as demonstrated in independent studies [12,13,14]. The ATP-binding cassette (ABC) transporter family, the largest group of transport proteins in plants, also plays multifaceted roles in Cd dynamics. Specific ABC transporters modulate Cd detoxification through vacuolar sequestration, apoplastic exclusion, and long-distance translocation. For instance, OsABCC1 facilitates vacuolar Cd storage in root cells [15], while OsABCG36 (PDR5) mediates Cd efflux at the plasma membrane [16] and RsPDR8 can promote Cd efflux in radish root [17,18].

The ABC transporter family widely exists in prokaryotes and eukaryotes in nature. It represents an important defense mechanism for the self-protection of organisms against environmental toxicants [19]. Due to the different order of nucleotide-binding domain (NBD) and transmembrane domain (TMD) arrangement, the ABC family is divided into different subfamilies, such as ABCA, ABCB, ABCC, etc. [20]. Recent studies focusing on cellular self-defense have discovered the vital role of the multidrug resistance-associated protein 1 (MRP1), which belongs to the ABCC subfamily, in excreting exogenous pollutants from cells [21,22].

The multidrug resistance-associated protein (MRP) family was first identified in humans, with HsMRP1 being a prototype member that confers multidrug resistance in cancer cells by exporting glutathione conjugates and organic acid anions extracellularly [23]. Plant MRP homologues were subsequently discovered through genomic homology analyses. In plants, MRP transporters are predominantly localized to the vacuolar membrane rather than the plasma membrane [24]. Their primary biochemical role involves compartmentalizing toxic substances—either absorbed from the environment or generated metabolically—into vacuoles. This sequestration mechanism minimizes cytoplasmic exposure to harmful compounds, thereby protecting the physiologically active protoplasm from toxicity and enhancing cellular detoxification capacity [25].

MRPs from multiple plant species contribute to vacuolar metal sequestration, thereby alleviating cellular metal toxicity. For example, AtMRP1 and AtMRP2 from Arabidopsis enhance plant tolerance to divalent Cd and mercury (Hg²⁺) [26]. Additionally, AtMRP3 and AtMRP6 in Arabidopsis contribute to heavy metal resistance: AtMRP3 reduces cadmium sensitivity in the yeast YCF1 mutant, while adeficiency in AtMRP6 leads to Cd-induced growth inhibition [27]. When exposed to Cd, wheat exhibits upregulated expression in four genes: TaABCC3, TaABCC4, TaABCC11, and TaABCC14. Research has further revealed that TaABCC13 in wheat participates in Cd detoxification by mediating membrane transport of glutathione-conjugated substrates [28,29].

Metallothioneins (MTs), characterized by their high cysteine residue content, bind free metal ions in plants through thiolate bonds to form stable complexes, serving as crucial heavy metal chelators in plant systems [30,31]. Additionally, MTs function as antioxidants, mitigating reactive oxygen species (ROS)-induced cellular damage [32,33]. In rice, the OsMT gene effectively sequesters intracellular Cd ions, substantially reducing phytotoxicity while maintaining normal growth and even enhancing Cd uptake from soil [34].

Furthermore, plant MT expression can be up regulated by exogenous inducers to alleviate Cd stress [35]. For instance, nitric oxide (NO) acts as a signaling molecule to elevate LeMT expression in tomato (Solanum lycopersicum L.), promoting efficient root-to-leaf Cd translocation and vacuolar compartmentalization of excess Cd. This spatial redistribution disperses cellular Cd accumulation, thereby alleviating toxicity and enhancing Cd resistance [36]. The overexpression of NtMT2F significantly enhanced resistance to Cd toxicity in transgenic Arabidopsis [37].

With the deepening understanding of Cd accumulation mechanisms in rice grains, preliminary attempts have been made in low-Cd rice breeding through multi-gene pyramiding strategies. However, all these genes are endogenous to rice. For instance, Yang successfully introgressed three japonica-derived genes (OsHMA3, OsNRAMP5, and OsNRAMP1) into indica rice cultivar 93-11, achieving a significant reduction in grain Cd accumulation [38]. Similarly, Tian et al. [39] demonstrated that co-expressing zinc transporter OsZIP3 with OsLCT1 and OsHMA2 in rice reduced Cd content in grains during the filling stage. Zinc application further enhanced Zn translocation to grains via OsZIP3, thereby competitively inhibiting Cd accumulation. These advancements highlight the promising potential of multi-gene combinatorial approaches in low-Cd rice breeding [40].

Vitis vinifera L., a perennial liana, has evolved unique heavy metal tolerance mechanisms through long-term evolution. Its root system is persistently exposed to soil heavy metals (e.g., Cd) and employs synergistic actions of ABC transporters (such as MRP1) and metallothioneins (MTs) to achieve vacuolar sequestration and metal chelation, a mechanism demonstrating superior efficiency compared to cereal crops in the Poaceae family.

Although cereals such as rice and wheat are phylogenetically closer to rice, their ABC transporters (e.g., OsABCC1) primarily participate in arsenic (As) detoxification, exhibiting relatively low transport efficiency for Cd [41]. In contrast, grapevine MRP1-type genes (e.g., VvABCC1) possess high affinity for Cd²⁺ and actively transport Cd into vacuoles [42], whereas the rice orthologue OsABCC1 preferentially transports arsenite (As (III)) with weak Cd transport activity [43].

Previous studies have primarily focused on introducing either MRP1-type genes or MT genes individually into plants to investigate their respective functions in Cd tolerance. To date, no corresponding research has been conducted on co-expressing these two gene classes in rice. Notably, current multi-gene breeding strategies for developing low-Cd rice exclusively utilize endogenous rice genes. The identification of novel genetic resources and subsequent functional studies of their synergistic effects hold significant importance for cultivating low-cadmium rice varieties while ensuring food safety. In this study, we integrated three genes from Vitis vinifera, named Vvmrp1, Vvmt1, and Vvmt2, into rice to systematically evaluate their synergistic effects on Cd tolerance and absorption. This strategy aimed to elucidate how coordinated metal chelation (via MTs) and vacuolar sequestration (via MRP) mechanisms collectively enhance rice resilience to Cd-contaminated environments while minimizing grain Cd deposition.

2. Materials and Methods

2.1. Materials, Vectors, and Chemicals

Agrobacterium tumefaciens (EHA105 and AT1) and Arabidopsis thaliana (ecotype Columbia L.) and rice (Zhonghua 11) were all prepared and maintained in our laboratory. Simple pMD-18 vector and Taq DNA polymerase, T4 DNA ligase, and other enzymes were purchased from Takara. An RNA isolation kit andreverse-transcription system kit was purchased from Sangon Biotech Co., Ltd. (Shanghai, China). Other reagents, except for special instructions, were all purchased from the China National Pharmaceutical Group.

2.2. Vvmrp1 Phylogeny

We retrieved 24 published mrp1 gene sequences from diverse species in GenBank, all belonging to the ABCC1 subfamily within the ABC transporter superfamily. Phylogenetic analysis was performed using the maximum-likelihood method in MEGA-X (2018) software. Nucleotide and amino acid sequence alignments between these 24 genes and Vvmrp1 were conducted with DNASTAR 17 software for homology assessment. Subsequent phylogenetic tree construction based on sequence homology was constructedusing MEGA.

2.3. Stereo View of the Structure of VvMRP1

First, the coding sequence of Vvmrp1 was translated into its amino acid sequence using DNAMAN software. The resulting sequence was submitted to SWISS-MODEL for template identification, aiming to locate proteins with known tertiary structures sharing > 30% sequence homology. If qualified templates were identified, homology modeling was performed to generate a 3D structure, followed by comparative analysis of functional domains (e.g., nucleotide-binding domains, transmembrane helices).

In the absence of suitable templates (30% homology), de novo structure prediction was conducted using AlphaFold. Due to computational constraints, the reduced AlphaFold database (reduced_ads) was downloaded to a local server. Custom scripts were executed with platform-default parameters to initiate structure prediction. After extensive computational processing, the predicted tertiary structure was saved as a PDB file. The PDB file was visualized and analyzed using PyMOLv2.5.7 software.

2.4. Design and Chemical Synthesis of Vvmrp1, Vvmt1, and Vvmt2 Genes of Vitis vinifera

To ensure high-efficiency expression of exogenous genes in rice, codon optimization was performed for the three target genes derived from grapevine (Vitis vinifera) based on rice codon usage preferences. According to the sequence on NCBI, the Vvmrp1 (GenBank accession No. XM002281034.4), Vvmt1 (GenBank accession No. XM010655984.3), and Vvmt2 (GenBank accession No. XM002285144.4) genes from Vitis vinifera were synthesized by continuous polymerase chain reaction (PCR) [44]. The design principles of gene synthesis include: avoiding polyA tailing signals such as ATTTA in genes, avoiding 6 or more continuous A + T sequences, avoiding 5 or more G + C sequences, maintaining the proportion of G + C at 40–60%, preventing intron cleavage sequences, reducing the two-stage structure hairpins within the gene, and avoiding the use of CG and TA double oligo nucleotides at the 2nd and 3rd positions (CG often causes methylation in plants). Primers were designed to cover the entire target gene sequence. Restriction enzyme sites were incorporated into the first and last primers. Excluding these terminal primers, all intermediate primers were 60 nucleotides in length and featured 20-nucleotide overlaps with adjacent primers. The final synthetic gene sequence was optimized for rice codon usage while preserving the original amino acid sequence. PCR was carried out as described by Tian et al. [45]. The primers are shown in Tables S1 and S2. The amplified fragment was digested by BamHI and SacI, cloned into Simple pMD-18, and sequenced. Errors in the synthetic gene were corrected by overlap extension PCR [46].

2.5. Plant Expression Vector Construction and Plant Transformation

We constructed the plant expression vector co-expressing three genes (Vvmrp1, Vvmt1, and Vvmt2) through a polycistronic expression cassette. The triple-gene co-expression vector was constructed with each gene flanked by independent regulatory elements: aCaMV35S promoter at the 5′ end and a Noster minator at the 3′ end. These three expression cassettes were sequentially ligated into the multiple cloning sites (MCSs) of the binary vector. The recombinant plasmid containing three genes was transformed into Agrobacterium EHA105 by electroporation.

The colony of the Agrobacterium strain containing the target plasmid was cultured at 28 °C for 2 days in 5 mL LB medium containing kanamycin, then 5 mL medium was transferred into 500 mL LB liquid medium and cultured at 28 °C for 16–24 h (OD = 1.5–2.0). The cells were collected by centrifugation at room temperature, centrifuged at 4000× g for 10 min, suspended in an equal volume of MS medium, then 25 mg/L acetosyringone solution was added and mixed thoroughly so as to obtain the transformed bacterial solution fully developed rice calli of uniform size were then immersed in the bacterial suspension and gently agitated for 8 min to facilitate optimal transformation. After infection, the calli were dried with sterile filter paper and transferred to MS medium containing 25 mg/L Acetosyringone. After 2 days of dark culture at 24 °C, the callis were washed with sterile water and transferred to N6 medium containing 50 mg/L hygromycin, 500 mg/L carbenicillin, and 1 mg/L 2, 4-D for screening. After 4 weeks, the positive calli were transferred to differentiation medium (0.2 mg/L ZT, 2 mg/L 6-BA, 0.2 mg/L NAA, 42 mg/L hygromycin) and differentiated into seedlings. The 3–4 cm rice seedlings were transferred to a rooting medium (42 mg/L hygromycin, 0.2 mg/L NAA, 1/2 MS). After the rice seedlings had completely rooted, they were planted in a greenhouse(25 °C, 16 h light/8 h dark cycle). All transgenic plants were carried for two more generations in order to obtain homozygous transgenic plants.

2.6. RT-PCR Detection of Transgenic Plants

The positive lines obtained by hygromycin screening were verified by RT-PCR analysis. Total RNA was extracted from the leaves of 3-week-old T2 lines using an RNA extraction kit (Tiangen, Beijing, China). The RT-PCR was programmed using the reverse transcription kit from Sangon Biotech Co., Ltd. (Shanghai, China).The first cDNA strand was synthesized in a 50 μL mixture with 10 ng RNA as template. Using cDNA as template, specific primer (Table S3) was added to amplify the specific fragment of the MRP1 gene. The rice actin gene (GenBank No. X16280) was used as internal reference (Table S3) to improve the reliability of RT-PCR. RT-PCR was carried out under the following conditions: 3 min at 94 °C, and 28 cycles of 30 s at 94 °C, 20 s at 60 °C, and 30 s at 72 °C, with a final extension 10 min at 72 °C in a PTC 200 thermal cycler (MJ Research, Reno, NV, USA). The PCR products were separated on 1% agarose gels. The DNA intensity ratio was quantified with a Model Gel Doc 1000 (Bio-Rad, USA). Three independent experiments were repeated with the same results, and only one of them is presented.

2.7. Tolerance of Transgenic Rice Overexpressing Vvmrp1, Vvmt1, and Vvmt2

Mature seeds of WT and transgenic rice lines were dehulled, surface-sterilized with 75% ethanol for 5 min and 2.5% sodium hypochlorite for 1 h, and then rinsed thoroughly with sterile water. Seeds were germinated on agarose plates supplemented with 0, 150, 300, 600, or 1000 µM CdCl₂, with eight seeds per genotype per plate. Survival rates were calculated after 14 days. After accelerating germination at 37 °C for 24 h in the dark, the WT and transgenic rice seedlings with uniform growth vigor were transferred into the solution containing 0, 50, and 100 µM CdCl₂ for 7 days (25 °C, 16 h light/8 h dark cycle).

One-month-old rice seedlings of WT and transgenic lines with uniform growth were selected in the greenhouse. Each pot containing 20 seedlings was bagged to isolate treatments and irrigated with 120 mL of Cd solutions at 10, 20, or 30 mM concentrations. Three biological replicates were maintained per line. During the experiment, plants were supplemented with Hoagland solution once a week. The treatments continued until observable phenotypic divergence emerged between WT and transgenic lines. Sixty days after sowing, the Cd application was terminated.

One-month-old WT and transgenic rice seedlings with uniform growth were selected for pot experiments. Soil was collected from the surface layer (0–20 cm depth) of paddy fields in Huacao Town, Shanghai, China, air-dried naturally, sieved (2 mm mesh) to remove debris, and homogenized. Each pot was filled with 4 kg of prepared soil, irrigated with 4 L water and 1 L CdCl₂ solution (0, 2, 5, or 10 mM), and maintained with a 2–3 cm water layer. After equilibrating under ambient conditions for 7 days, seedlings were transplanted.

Four seedlings per pot were transplanted, with three biological replicates per treatment. Plants were cultivated outdoors under natural light and temperature. Survival rates were recorded at 1 and 2 weeks post-transplantation. Tillering numbers were counted 7 days after panicle initiation. The effective tiller number of each plant was counted and the statistical values of tiller numbers for all plants at the same concentration were used for plotting. Chlorophyll content (SPAD values) of the penultimate fully expanded leaves was measured at tillering and booting stages using an SPAD-502 Plus chlorophyll meter (KO, Japan), and flag leaf SPAD values were assessed at maturity. At maturity, plant height, number of productive panicles, and fresh biomass of shoots and roots were measured. Root systems were carefully washed, blotted dry, and separated from shoots for fresh weight determination.

At maturity, roots, shoots, and grains were separately harvested.

To ensure unimpaired Cd absorption and translocation functions in plant tissues prior to phenotypic manifestation and the effect of Cd on Ca, Mg, P, K, and Fe content in various tissues of rice, we quantified Cd, Ca, Mg, P, K, and Fe concentrations in both WT and transgenic lines. The samples were oven-dried at 110 °C. The mineral content was determined according to the method of Meng [47].The dried rice materials were ground into powder, and 0.5 g of the homogenized sample was digested with 5 mL of 69.5% HNO₃ and 2 mL of 35% H₂O₂. The mineral content in the digested solution was determined using an axial-view inductively coupled plasma optical emission spectrometer (ICP-OES; PerkinElmer, Norwalk, CT, USA).The Cd content in roots and the Ca, Mg, P, K, and Fe content in shoots were determined.

2.8. Statistical Analysis

All analyses were performed using SPSS 26.0 (IBM, USA). Data normality was confirmed by Shapiro–Wilk tests (p > 0.05 for all groups). Homogeneity of variances was verified by Levene’s test (p = 0.186). One-way analysis of variance (ANOVA) was employed to assess intergroup differences. Where significant main effects were detected (p < 0.05), post hocmultiple comparisons were conducted using the Bonferroni method, controlling the family-wise error rate (FWER) at α = 0.05. Outliers were defined as observations beyond 1.5× the interquartile range (IQR) in box plots and treated by median imputation. Sensitivity analyses confirmed the robustness of conclusions. In the figures, significance is denoted as * p < 0.05, ** p < 0.01, *** p < 0.001, and error bars represent the standard error of the mean (SEM).

3. Results

3.1. Vvmrp1 Gene Phylogeny

Sequence alignments of mRNA and amino acid sequences from 25 MRP1 homologues were performed using AlignIR V2.0 and Clustal W v2.0.11 software. A molecular phylogenetic tree was subsequently constructed with MEGA-X (Figure 1). The analysis revealed that Vvmrp1 from Vitis vinifera (highlighted by a red ellipse) shared the highest homology (71.73%) with Arabidopsis thaliana (GenBank No. NM102777.3, marked by a blue rectangle), indicating the closest evolutionary relationship. This phylogenetic pattern suggests thatmrp1-type genes likely originated in ancestral marine organisms, followed by divergent evolutionary trajectories: one lineage advanced in higher animals, while the other diversified into fungi, algae, and higher plants.

3.2. The Tertiary Structure of Vvmrp1

The tertiary structure of VvMRP1 protein predicted by Deep Mind’s AlphaFold v2.0 AI system is shown in Figure 2. The overall architecture comprises three major components, consistent with the proposed topological model of MRP1 proteins: TMD0-L0-TMD1-NBD1-TMD2-NBD2 domain arrangement. The central core features a large transmembrane channel formed by 12 long α-helices, flanked by two symmetrical “lid-like” structures positioned above the channel. Each lid consists of seven short α-helices and nine β-sheets, interconnected by random coils. A smaller transmembrane channel composed of five moderately short α-helices is located on the left side of the protein, while the upper-right region contains a channel-like domain formed by four short α-helices.

3.3. Chemical Synthesis of Vvmrp1, Vvmt1, and Vvmt2

To ensure stable gene expression in transgenic plants, the Vvmrp1 gene was chemically synthesized using 121 primers (Table S1) according to the original sequence and plant codon usage preference. Except for the bases related to 18 enzyme cleavage sites, a total of 970 bases were altered in the chemically synthesized gene. The GC content of the optimal sequence was 50.09%, and the similarity with the WT gene was 80.06%. The Vvmt1 gene was chemically synthesized using seven primers (Table S2) according to the original sequence and plant codon usage preference. The chemically synthesized gene retains only two enzyme cleavage sites, BamHI and SacI, at both ends, and does not contain any other endonuclease sites. The GC content of the optimal sequence was 48%, and the similarity with the WT gene was 77.38%. The Vvmt2 gene was chemically synthesized using seven primers (Table S2). The GC content of the optimal Vvmt2 sequence was 50%, and the similarity with the WT gene was 79.67%.

The amino acid sequence of the gene was translated using DNAMAN. There was no difference between the optimized and wild amino acid sequences.

3.4. Construction of Triple-Gene Expression Vectors and Generation of Transgenic Rice Plants

The recombinant plasmid (Figure 3a) overexpressing Vvmrp1, Vvmt1, and Vvmt2 was transformed into EHA105 by electroporation. The three genes were transferred into Zhonghua 11 by EHA105. After continuously selfing the transgenic plants to produce the T2 generation seeds, these were then subjected to RT-PCR to detect the expression of Vvmrp1, Vvmt1, and Vvmt2. Based on the results, three rice overexpression lines (AH781-1, AH781-2, AH781-3)were selected for gene functional analysis (Figure 3b). Under normal growth conditions, gene transfer had no significant effect on plant growth or development.

3.5. Germination of Transgenic Rice on Cd Medium

Rice seed germination remained unaffected by Cd (≤1 mM), yet seedling survival exhibited dose-dependent inhibition. At 150 μM Cd, both WT and transgenic lines showed high survival rates (95–100%). At 300 μM Cd, transgenic lines maintained 95–97% survival vs. 75% in WT (p < 0.01). This tolerance gap widened at 600 μM (50–62.5% vs. 12.5%; p < 0.001), with complete WT mortality at 1 mM vs. residual transgenic survival (12.5–62.5%). Transgenic modifications conferred significant Cd resistance during post-germination stages (Figure 4).

3.6. Cadmium Tolerance of Transgenic Rice Seedlings

Compared to control, one-week-old aseptic rice seedlings exposed to 50 and 100 μM Cd for 7 days showed concentration-dependent growth inhibition. At 50 μM Cd, WT roots and leaves ceased elongation, while transgenic lines maintained slow growth with significant biomass differences vs. WT (p < 0.01). Under 100 μM Cd, both genotypes arrested growth, but only AH781-3 retained green leaves among transgenic lines, contrasting with WT’s complete mortality. Biomass disparities shifted from p < 0.01 to p < 0.05 between WT and transgenics under escalating Cd stress. Detailed phenotypic characteristics and plant height data are presented in Figure 5.

Two-week Cd treatment caused progressive growth inhibition in rice, with transgenic lines exhibiting superior tolerance. At 0 mM Cd, both WT and transgenic plants attained identical heights of 55 cm and fresh weights of 60 g. Under 10 mM Cd, transgenic plants reached 47 cm in height, 17.5% taller than WT (40 cm; p < 0.05), and achieved 52.5 g fresh weight, exceeding WT by 32.9% (39.5 g). At the highest Cd concentration (20 mM), transgenic plants maintained 44 cm height and 44.5 g fresh weight, surpassing WT (27 cm and 26.2 g) by 61.3% and 58.9%, respectively (p < 0.01) (Figure 6).

Cd accumulation in rice tissues exhibited concentration-dependent increases, with transgenic lines showing reduced uptake efficiency. At 0 mM Cd, no Cd was detected in transgenic roots or shoots, while WT roots contained trace levels (5 mg/kg), likely from soil contamination. Under 10 mM Cd, transgenic roots accumulated 1019 mg/kg Cd (51.9% of WT 1962 mg/kg; p < 0.05), with shoot content similarly reduced to 88 mg/kg vs. 156 mg/kg in WT (56.4% of WT). This pattern persisted at 20 mM Cd: transgenic roots and shoots contained 1977 mg/kg (65% of WT’s 3015 mg/kg) and 483 mg/kg (77.2% of WT’s 625 mg/kg), respectively. Shoot-to-root Cd ratios increased with concentration, rising from 8.6% (transgenics) and 8% (WT) at 10 mM to 24% and 20.8% at 20 mM (

Table 1. Cd content of WT and transgenic plants AH781 carrying the Vvmrp1, Vvmt1, and Vvmt2 genes.

Cd Concentration (mM)	Organs	Cd Content of AH781 (mg/kg DW)	Cd Content of WT (mg/kg DW)	Rate (%)
0	root	0 a	5 ± 1 b	0
	shoot	0 a	0 a	0
10	root	1019 ± 15.3 a	1962 ± 60.5 b	51.9
	shoot	88 ± 6.2 a	156 ± 15.32 b	56.4
20	root	1977 ± 35.67 a	3015 ± 70.87 b	65
	shoot	483 ± 25.43 a	628 ± 35.25 b	77.2

All data are based on mean values ± standard deviation. Different letters indicate significant differences at 0.05 level by Duncan’s multiple range test for multiple comparisons. There are significant differences between numerical values with different letter suffixes in the same line.

).

Eight-week-old transgenic and WT plants irrigated with 30 mM Cd displayed divergent responses. WT leaves exhibited chlorosis, curling, and apical necrosis within one week, culminating in complete mortality by day 15. Transgenic lines sustained normal growth with only basal leaf senescence (Figure 7a). Cd content in WT shoots exceeded transgenic by 71.8–158%, with line AH781-1 showing highly significant differences (p < 0.01) and two other lines demonstrating significant differences (p < 0.05) (Figure 7b).

3.7. Phenotypic Divergence Between Transgenic and WT Plants Under Cd Stress Across the Full Life Cycle

Transgenic and WT rice seedlings exposed to Cd exhibited concentration-dependent survival and growth responses. Both genotypes demonstrated 100% survival at 0–5 mM Cd after one week, but transgenics retained 50% viability at 10 mM Cd vs. complete WT mortality. By week two, all plants had succumbed at ≥10 mM Cd (Figure S1). Delayed senescence under 2–5 mM Cd manifested as persistent green leaves and immature panicles during grain filling, contrasting with natural maturation in unstressed controls (Figure 8a). Biomass suppression under Cd stress showed genotypic divergence: at 2 mM Cd, transgenic shoots maintained 166.2 g fresh weight, 14.5% greater than WT (145.2 g, p < 0.001), while root biomass loss was 69 g vs. WT’s 73 g (p < 0.05). This tolerance persisted at 5 mM Cd, with transgenics retaining 137.8 g shoot biomass (37.7% over WT’s 100.09 g, p < 0.001) and 84 g root loss compared to WT’s 89.22 g (p < 0.001) (Figure 8b).

Chlorophyll content analysis revealed that at maturity, 0 mM Cd plants had chlorophyll relative values of 33–35, while 2 mM Cd groups retained higher values (46–48) with no genotypic differences. At 5 mM Cd, transgenic lines maintained a chlorophyll value of 47.2 vs. 42.5 for WT (p < 0.001), indicating enhanced tolerance (Figure 8c). Cd also suppressed tillering and panicle productivity: in 0 mM Cd soil. Both genotypes had similar tiller counts (55) and panicle numbers (52), but at 2 mM Cd, WT produced 26 tillers vs. 30 for transgenic (p < 0.05), and at 5 mM Cd, WT had 22.5 tillers vs. transgenic 26 (p < 0.05). Productive panicles in WT dropped to 22 (2 mM) and 20 (5 mM), while transgenics retained 29 and 28 panicles, respectively (p < 0.01) (Figure 8d). Plant height diverged under Cd stress, with transgenic plants significantly taller than WT at 2 mM Cd (p < 0.01), though no difference occurred at 5 mM Cd (Figure 8e).

Post-harvest tissue analysis revealed differential Cd accumulation patterns, with concentrations following the order roots > stems > grains.Cd was detected in rice from the untreated control group, which may be attributable to trace amounts of Cd naturally present in the paddy soil. No significant differences were observed in Cd concentrations between WT and transgenic rice in roots (WT: 36.38 mg/kg; transgenic: 36.15 mg/kg) or stems (WT: 4.86 mg/kg; transgenic: 4.50 mg/kg). However, Cd accumulation in grains was significantly lower in transgenic plants (0.13 mg/kg) compared to WT (0.27 mg/kg; p < 0.05).Under 5 mM Cd exposure, both transgenic and WT plants exhibited elevated Cd accumulation across tissues though the transgenic line displayed significantly reduced uptake. Compared to the WT control, the transgenic plants accumulated 64.2% less Cd in roots (216.58 vs. 604.97 mg/kg), 17% less in stems (18.48 vs. 22.27 mg/kg), and 23.2% less in grains (1.33 vs. 1.73 mg/kg), with statistically significant (p < 0.01) genotypic differences. The distribution pattern remained consistent with typical rice Cd translocation, showing roots > stems > grains in both lines (Figure 8f). Cd exposure also altered mineral nutrient profiles in stems and grains. Stem Ca content declined sharply in both WT (10.12→5.59 g/kg) and transgenic (10.45→5.78 g/kg) plants. However, transgenic stems retained higher Fe, P, and Mg (p < 0.05), but lower K than WT. In grains, transgenic lines exhibited elevated Fe (p < 0.01 under Cd) and Ca, but reduced K, P, and Mg (p < 0.01) compared to WT (Figure 8g).

4. Discussion

Phylogenetic analysis revealed that the Vvmrp1 gene share the highest homology with the Arabidopsis GenBank homologue NM_102777.3 (Figure 1). However, the three-dimensional (3D) protein structures of MRP1-class genes remain experimentally uncharacterized. Using AlphaFold, the predicted 3D structure of VvMRP1 aligns with the canonical features of the ABCC subfamily, marking it as the first MRP1-class protein in plants with a comprehensively predicted 3D structure (Figure 2). Given that tertiary structure largely dictates protein function, this prediction provides critical insights. Notably, the L0 peptide sequence within MRP1-class proteins exhibits low sequence similarity among ABCC family members, yet its secondary structure appears conserved across the family. Initially predicted to localize to the cytoplasm, the L0 domain was later experimentally demonstrated to associate with the membrane [48], suggesting its necessity for membrane localization of VvMRP1.

The transporter genes associated with Cd transport are commonly utilized in low-Cd rice genetic engineering and breeding. These genes are either overexpressed or knocked out in rice. In this study, the coordinated expression of three transgenes significantly reduced Cd accumulation in rice without compromising growth or development (Figure 6 and Figure 7). In the absence of Cd supplementation, WT and transgenic rice showed no significant differences in plant height, tiller number, or effective panicle number (Figure 8d,e), suggesting that the transgene did not adversely affect normal rice growth. In contrast, the knockout of various transporter genes, while effective in decreasing Cd accumulation in rice, often impairs normal growth and development. For example, knockout of endogenous transporter genes, while effective in decreasing Cd content, adversely affects plant development [49,50]. OsNramp5-knockout lines exhibited markedly lower Cd concentrations in roots and grains, but simultaneously impair Mn and Fe uptake, resulting in stunted growth of mutant seedlings and a substantial yield reduction to merely 11% of wild-type levels [49]. The knockout of certain transporter genes unexpectedly increases Cd accumulation in rice. For instance, disruption of OsPDR20 not only elevates Cd content but also impairs root elongation and biomass [50]. Similarly, knockout of OsABCC9 enhances rice sensitivity to Cd, leading to significantly higher Cd accumulation in both roots and shoots [51]. Co-transgenic rice expressingVvmrp1, Vvmt1, andVvmt2 exhibit lower Cd accumulation in roots (51.9–65.5% of WT) and stems (56.4–77.2% of WT) under low Cd stress (Table 1 and Figure 7). Under 5 mM Cd, mature transgenic plants accumulated 216.58 mg/kg Cd in roots (35.8% of WT), 18.48 mg/kg in stems (83% of WT), and 1.33 mg/kg in grains (76.8% of WT) (Figure 8f).

Under 5 mM Cd treatment in soil, the fresh weight of both stems and roots, as well as the chlorophyll content, of transgenic plants were significantly higher than those of wild-type (WT) plants (Figure 8b,c). This outcome is directly attributed to the Cd content within the plants. The Cd content in transgenic plants was significantly lower than in WT plants (Figure 8f). Photosynthesis is the most fundamental life activity of plants, serving as the essential basis for the survival of green plants and the source of starch in cereal grains. Research indicates that Cd stress significantly reduces the photosynthetic rate and transpiration rate of plants, alters stomatal function, and exhibits a significant negative correlation between photosynthetic efficiency and Cd stress concentration [52]. The inhibition of photosynthesis by Cd results in excessive reactive oxygen species (ROS) within the plant, which can severely damage the photosynthetic apparatus and impair photosynthetic function. For instance, excessive ROS can block protein synthesis and repair in photosystem II (PSII), inhibit photoactivation, directly or indirectly reduce D1 protein levels, and impair the functionality of the reaction center [53]. Under Cd stress, the electron transport rate in PSII decreases and stomatal closure occurs [54]. The detrimental effects of Cd on plant photosynthesis extend beyond these impacts, including reduced transcription of associated genes, inactivation of enzymes involved in CO₂ fixation, enhanced protein hydrolysis, and disruption of nitrogen and sulfur metabolism [55]. Beyond affecting rice germination and seedling growth, Cd stress also exerts significant impacts on the growth and development during the mid-stage of rice. The influence of Cd on rice growth, development, and yield varies with its concentration. At concentrations ≤ 0.3 mg/L or 0.1 μM/L, Cd can promote rice growth and development. However, as Cd concentration increases, it inhibits tillering, plant height, and panicle length. The number of productive panicles per plant and the 1000-grain weight also decline with rising Cd stress levels. High Cd concentrations further significantly impair reproductive growth in rice [56]. Consistently with this, rice plants grown in high-Cd soil exhibited reduced biomass, shortened and curled leaves, severe chlorosis in middle and lower leaves, and a delayed heading date by 3–4 days [57]. These plants also showed delayed maturity, increased late tillering, shorter panicles, and fewer grains per panicle. Notably, hybrid rice combinations exhibited more pronounced inhibition under Cd stress compared to conventional rice cultivars.

Multiple MRP1 transporters have been shown to transport metal/metalloid–ligand complexes, regulating Cd detoxification and accumulation within plants. AtABCC1, AtABCC2, and AtABCC3, have been shown to confer Cd tolerance in Arabidopsis by sequestering PC–Cd in vacuoles [32,58]. AtMRP1, AtMRP2 [26], AtMRP3, and AtMRP6 [27] can enhance plant tolerance to divalent Cd. Additionally, TaABCC3, TaABCC4, TaABCC11, and TaABCC14 exhibit upregulated expression under cadmium exposure [28,29]. Overexpression of these genes not only enhances Cd accumulation in plants but also improves their Cd tolerance. In this study, the coordinated expression of single genes such as VvMRP1 effectively reduced Cd content in rice roots, stems, and grains, which aligns with findings from some published studies. For instance, OsABCG36 (OsPDR9), localized to the root cell plasma membrane, enhances Cd tolerance in yeast by mediating Cd efflux [15]. Similarly, overexpression ofAtPDR8, a member of the ABC subfamily, increases Cd tolerance while reducing Cd accumulation in plants. Conversely, knockout of OsABCC9 elevates Cd levels in rice. Additionally, AtATM3 contributes to Cd resistance by exporting glutathione-Cd (II) complexes synthesized in mitochondria [59]. Thisindicates that Vvmrp1 participates in Cd detoxification by mediating membrane transport of glutathione-conjugated substrates.

Co-expression of MT genes, such as CcMT1 and PpMT2, enhances heavy metal tolerance by chelating free Cd via disulfide bonds, thereby mitigating cellular damage [60,61]. MT–Cd complexes are trafficked to metabolically inert compartments, reducing toxicity while promoting Cd uptake, as observed in ryegrass, tomato, and rice [34,62,63]. This observed phenomenon may be attributed to the transport function of VvMRP1s. As a transmembrane transporter, VvMRP1 may occupy binding sites on the plasma membrane, thereby reducing the availability of binding sites for other transport proteins. This competitive occupation ultimately decreases the number of functional transporters for phytochelatin (PC)–Cd²⁺ and MT–Cd²⁺ complexes, impairing cellular compartmentalization and sequestration of Cd. Consequently, these combined effects contribute to the observed reduction in intracellular Cd accumulation.

Mineral nutrients are essential components of plant functional substances and participate in numerous metabolic processes critical for growth and development. Cd primarily alters plant nutrient composition by affecting plasma membrane permeability [64]. Cd stress disrupts nutrient uptake, thereby impairing plant growth. Research by Ou et al. [65] demonstrated that Cd significantly reduces Mn and Zn levels in barley, while Liu et al. [66] reported Cd-induced interference in rice absorption of Fe, Zn, Mn, and Mg, with effects varying by element. In this study, Cd’s impact on five elements in stems was categorized as either negligible or inhibitory. The transgene exerted three effects on nutrient dynamics: promotion, inhibition, or no impact (Figure 8g). Transgenic stems exhibited significantly higher Fe, P, and Mg than WT, suggesting enhanced root-to-shoot translocation. This may stem from reduced Cd competition for transport proteins or the ability of VvMRP1 to transport ion complexes, aided by VvMT1 and VvMT2 metallothioneins, which bind and mediate metal transport [67,68]. Conversely, transgenic lines showed lower K, likely due to competition between overexpressed MRP1 and K⁺ channel proteins for membrane localization or ATP. No significant difference in Ca levels between genotypes (Figure 8g) indicated neutral transgene effects on Ca transport. In grains, the transgene increased Ca and Fe, but decreased K, P, and Mg. Notably, while stem Ca showed no genotypic disparity, grain Ca differed significantly (Figure 8g), highlighting the transgene’s role in tissue-specific nutrient allocation. These findings underscore the complexity of mineral absorption and distribution mechanisms, warranting further investigation to elucidate underlying pathways.

Despite extensive research on structure, substrates, and inhibitors of MRP1, its transport mechanism remains enigmatic due to substrate cooperation/competition and multifunctional tissue distribution [69]. Further studies are needed to elucidate: (1) the chemical form of Cd transported by VvMRP1, (2) Cd speciation in Vvmt1 andVvmt2 co-expressing plants, and (3) whether MT–Cd complexes serve as novel ABCC substrates. Resolving these mechanisms will advance low-Cd rice breeding strategies. Identification and utilization of exogenous superior genes through transgenic approaches enables the development of low-Cd rice varieties. The genes Vvmrp1, Vvmt1, and Vvmt2 demonstrate promising potential for practical application. This study provides valuable insights for exploring novel low-Cd genes and offers a new strategy for breeding low-Cd-accumulating rice.

5. Conclusions

Based on rice codon preference, three genes (Vvmrp1, Vvmt1, and Vvmt2) derived from Vitis vinifera were chemically synthesized. Vvmrp1 belongs to the MRP1 subfamily of ABC transporters, while Vvmt1 and Vvmt2 encode metallothioneins. These genes were introduced into rice cultivar Zhonghua 11 via Agrobacterium-mediated transformation. Results demonstrated that transgenic lines exhibited superior Cd tolerance throughout all growth stages compared to WT plants. Transgenic lines showed higher survival rates, increased tiller numbers, more effective panicles, and significantly greater fresh weight and plant height than WT. Crucially, Cd accumulation in roots, stems, and grains of transgenic lines was significantly reduced relative to WT. This indicates that the three genes not only substantially enhance Cd tolerance but also effectively decrease Cd accumulation in rice, offering significant practical value for low-Cd rice breeding and safer food production. Furthermore, this study provides insights for polygenic breeding strategies.