MIR408 is induced by Cd and involved in Cd uptake in rice

To examine the response of MIR408 to Cd, wild-type (WT) rice seedlings were first subjected to moderate Cd (2 μM) stress for 12–72 h. The expression of stem-loop miR408 precursor (pre-miR408) exhibited a significant increase, with the levels rising approximately 5-fold in roots after 12 h of treatment, while enhanced by over 25-fold after 3 days (Fig 1A). In a time-dependent experiment involving high-concentration Cd treatment (10 μM), an increased induction of pre-miR408 by Cd supply was observed (S1A Fig). In addition, GUS staining in both primary and adventitious roots from transgenic lines expressing the GUS reporter gene under the control of the MIR408 promoter implicated a significant enhancement of primary miR408 (pri-miR408) transcription after 4h of moderate Cd treatment (Fig 1B). Taken together, these results suggest that the MIR408 is strongly induced in Cd-exposed rice roots.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. MIR408 is involved in Cd uptake regulation in rice.

(A) Time course analysis of pre-miR408 expressions in 14 d-old Nipponbare (Nip) plants with 2 μM CdCl₂ treatment. Actin was used as an internal control for the normalization of the qRT-PCR results. Values are means ± SD (n = 3 biological replicates). (B) MIR408p-GUS staining in 10d-old rice plants with and without 4 h 2 μM CdCl₂ treatment. Bars = 1 cm. (C) The contents of Cd in roots and shoots of 14 d-old WT and MIR408-OE seedlings grown in 2 μM CdCl₂ conditions. (D) Translocation of Cd from roots to shoots of WT and MIR408-OE seedlings exposed in 2 μM CdCl₂ for 14 days. ns, not significant. (E) The contents of Cd in roots and shoots of 14 d-old WT and mir408 mutant seedlings grown in 2 μM CdCl₂ conditions. (F) Translocation of Cd from roots to shoots of WT and mir408 mutants exposed in 2 μM CdCl₂ for 14 days. Error bars in (C) to (F) indicate SD (Student’s t test, *P < 0.05; **P < 0.01; ***P < 0.001). (n = 5 biological replicates). The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003811.g001

To investigate the potential involvement of MIR408 in Cd regulation in rice, we measured Cd contents in roots and shoots of MIR408-overexpressing transgenic plants (MIR408-OE). In contrast with similar contents of Fe, Mn, Cu, and Zn (S1B and S1C Fig), the concentration of Cd in two-week-old transgenic plants was remarkably lower than that in WT plants after exposure to moderate Cd for 14 days, regardless of accumulation in roots or shoots (Fig 1C). Meanwhile, when we calculated the translocation ratio of Cd in shoots and roots, we found that it was comparable in WT and MIR408-OE plants (Fig 1D), indicating that ectopic expression of MIR408 in rice suppresses the intake of Cd from roots, but may not affect the translocation efficiency of Cd from underground to aboveground.

To validate the involvement of MIR408 in Cd absorption in rice, contents of Cd in roots and shoots were compared between WT and two independent mir408 mutant lines. The concentration of Cd in roots increased from 600 to 700 mg kg⁻¹ dry weight (DW), with a similar proportion of increase in shoots in mir408 mutants compared with that in WT plants (Fig 1E and 1F), confirming the negative effects of MIR408 on Cd uptake in rice. Moreover, we observed similar accumulations of Mn, Cu and Zn in roots and shoots, while only a slight decrease of Fe was observed in mir408 roots compared with that in WT (S1D and S1E Fig), implying the feasibility of reducing Cd uptake and accumulation via MIR408 manipulation in rice.

miR408-5p regulates HIPP19 via translation repression

MIR408 can produce two mature miRNAs, namely miR408-5p and miR408-3p, from different arms [6]. First, we determined the abundance of miR408-5p under moderate and high Cd treatments. Similar to the pri-miR408 and pre-miR408, we observed that miR408-5p was dramatically induced by environmental Cd in roots, particularly under the 3d long-term treatments (Figs 2A and S2A), implying a possible role of miR408-5p in Cd regulation in rice.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. miR408-5p regulates HIPP19 in a translation repression manner.

(A) Time course analysis of miR408-5p accumulations in Nip plants with 2 μM CdCl₂ treatment. Student’s t test, ***P < 0.001. Values are means ± SD (n = 3 biological replicates). (B) Gene structure of HIPP19 and alignments of miR408-5p with target sites in HIPP19 CDS and the indicated mutant CDS (m1CDS). (C) Relative mRNA levels of HIPP19 in 14 d-old Nip plants with different time of 2 μM CdCl₂ treatment. Values are means ± SD (n = 3 biological replicates). (D) Validation of HIPP19 as miR408-5p target through transient expression analysis in N. benthamiana leaves. Left: The constructs in A. tumefaciens transiently introduced in N. benthamiana leaves. Middle: Representative photograph of firefly luciferase fluorescence signals when the indicated construct combinations were introduced in N. benthamiana leaves. Right: Relative reporter activity in N. benthamiana leaves expressing the indicated construct combinations. Error bars indicate SD (Tukey’s honestly significant difference, P < 0.05) (n = 3 biological replicates). (E) Relative mRNA levels of HIPP19 in WT, MIR408-OE, STTM-5p, and mir408 plants. ns, not significant. Values are means ± SD (n = 3 biological replicates). (F) The constructs that were introduced into rice protoplast to determine the effect of miR408-5p on HIPP19. (G) The relative protein level of HIPP19 when HIPP19 CDS or mCDS cassettes were introduced into rice protoplast with co-overexpression of MIR408. MIR530 was used as a control. RT-P1 and immunoblot of GFP were used to show the efficiency of plasmid transiently introduced into the indicated protoplast. RT indicates transcript abundance determined by reverse transcription PCR. Anti denotes the protein level examined by western blot analysis using the indicated antibody (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 3 biological replicates). (H) The relative protein level of HIPP19 when HIPP19 CDS or mCDS cassettes were introduced into rice protoplasts isolated from WT, MIR408-OE, and mir408 plants. The different letters on top of each bar denote significant differences (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 3 biological replicates). (I) The relative protein level of HIPP19 when HIPP19 CDS or mCDS cassettes were introduced into rice protoplasts isolated from WT, STTM-5p, and STTM-3p plants. The experiments were performed three times and one of the representative results was shown below the columns (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 3 biological replicates). (J) The constructs were transformed into rice plants to determine the effect of miR408-5p on HIPP19. (K) RNA and protein abundance of HIPP19 in WT and transgenic plants with indicated constructs shown in (J). The number below the lane represents the relative amounts of transcripts and proteins. The data underlying this Figure can be found in S1 Data and S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003811.g002

Using PsRobot, a widely-used small RNA analysis toolbox [27], we predicted a metallochaperone family gene, HIPP19, as a novel target of miR408-5p in rice (Figs 2B and S2B). Given that the expression of HIPP19 similarly responds to Cd treatments (Figs 2C and S2C), we examined whether miR408-5p can regulate HIPP19 in rice. To this end, we first performed a luciferase (LUC)-based reporter assay, wherein HIPP19 was transiently expressed in Nicotiana benthamiana leaves under the control of 35S promoter with MIR408 overexpression or with that of MIR530 as a control (Figs 2D and S2D). We observed that the LUC signal generated by HIPP19 activity was remarkably reduced when miR408-5p was overproduced compared with miR530 (Figs 2D and S2D). However, when we introduced mutated forms of HIPP19 that were mismatched with miR408-5p target sites, the repressed LUC signal was restored (Figs 2D and S2D). These results suggested that HIPP19 could be targeted and suppressed by miR408-5p.

Next, we investigated the expression of HIPP19 in MIR408-OE transgenic plants and were surprised to observe that it was comparable to that in WT (Fig 2E). Furthermore, the HIPP19 mRNA level exhibited similarity in both WT and mir408 mutants (Fig 2E), in contrast with the significantly enhanced expression of UCL8 (S2E Fig), a known miR408-3p target [20], when MIR408 was knocked out. In addition, HIPP19 exhibited a comparable expression pattern in WT and Small Tandem Target Mimic-5p (STTM-5p) transgenic plants with or without Cd treatment (Figs 2E and S2F), where miR408-5p accumulation was specifically repressed and miR408-5p activity was blocked (S2G Fig) [6]. This implies that miR408-5p has minimal effects on HIPP19 transcription.

Since miR408-5p has been demonstrated to regulate targets via either mRNA cleavage or translation repression [6], we next determined whether miR408-5p inhibits HIPP19 translation. Initially, we co-expressed miR408 with intact or mutated HIPP19 in rice protoplasts and examined the transcripts and proteins of HIPP19 (Fig 2F). Although the GFP protein level was similar, which indicated comparable transformation efficiency, the abundance of HIPP19 protein shown by western blot using FLAG antibody was significantly lower when MIR408 was co-expressed with intact HIPP19 compared to that with a mutation at the target site or when MIR408 was replaced to MIR530 as a control (Fig 2G). Secondly, we transiently overexpressed HIPP19 in the protoplasts of miR408 gain- and loss-of-function plants. As shown in Fig 2H, the protein level of HIPP19 was diminished in MIR408-OE and enhanced in mir408 mutants, respectively. However, no altered HIPP19 protein level was observed when the mutated form of HIPP19 was introduced instead of the intact version (Fig 2H). Considering that the mRNA level of HIPP19 was similar in the protoplasts from all plants (Fig 2H), these data indicate that the downregulation of HIPP19 by MIR408 takes place at the protein level.

To demonstrate that HIPP19 regulation is specific to miR408-5p, we introduced the plasmid described above into STTM-5p and STTM-3p plants, where the activity of miR408-5p and miR408-3p was respectively blocked [6]. Interestingly, the HIPP19 protein was significantly increased in the protoplasts from STTM-5p plants, while it remained consistent in STTM-3p plants compared with that in WT plants (Fig 2I). In addition, when the target site was mutated, the increased HIPP19 protein levels in STTM-5p disappeared (Fig 2I). In addition, when we introduced intact or mutated HIPP19 at the miR408-5p target site with a GFP tag driven by the native HIPP19 promoter (named HIPP19 and HIPP19m, respectively) into rice (Fig 2J), we found that although the transcription of HIPP19 was comparable (Fig 2K), its protein level was obviously higher in HIPP19m than in HIPP19 transgenic plants (Fig 2K). A similar increase in HIPP19 protein accumulation in HIPP19m plants relative to HIPP19 plants was observed under Cd treatment. Notably, although HIPP19 abundance was elevated in HIPP19m plants compared to HIPP19 plants under Cd stress, this increase was less pronounced than that observed under normal growth conditions (S2H Fig). This discrepancy is likely attributable to the concurrent induction of both miR408-5p and HIPP19 expression upon Cd exposure (Fig 2A and 2C). These data collectively demonstrate that miR408-5p mediates HIPP19 regulation through translation inhibition.

Since both miR408-5p and HIPP19 are induced by Cd exposure (Fig 2A and 2C), we investigated whether Cd treatment alters the inhibitory effect of miR408-5p on HIPP19 protein accumulation. We isolated protoplasts from WT, STTM-5p, and STTM-3p plants with or without Cd exposure, then transiently overexpressed either intact HIPP19 or HIPP19 with mutated miR408-5p target sites in these protoplasts (Fig 2F). HIPP19 protein levels were significantly higher in STTM-5p protoplasts than in WT or STTM-3p protoplasts when overexpressing the intact HIPP19. However, Cd treatment attenuated this increase in STTM-5p plants compared to mock-treated controls (S2I Fig). Critically, when the miR408-5p target sites in HIPP19 were mutated, HIPP19 protein levels became comparable across WT, STTM-5p, and STTM-3p protoplasts regardless of Cd exposure (S2I Fig). Given that miR408-5p is induced by Cd and its activity is blocked in STTM-5p plants, these results demonstrate that Cd exposure enhances miR408-5p-mediated repression of HIPP19 protein, dependent on functional target sites.

Collectively, these data demonstrate that miR408-5p suppresses HIPP19 via translational repression, and this inhibitory effect is potentially enhanced under Cd treatment.

HIPP19 is involved in Cd uptake and accumulation regulation

To investigate the physiological significance of the miR408-5p-HIPP19 module in Cd regulation in rice, we generated HIPP19 knock-out mutants using the CRISPR/Cas9 approach and measured Cd contents in roots and shoots of hipp19 mutants grown in nutrient solution with 10-day 2 μM Cd supply (S3 Fig). Consistent with findings in MIR408-OE lines, we observed a remarkable decrease in Cd concentrations both in roots and shoots of hipp19 mutants at the vegetative growth stage (Fig 3A), but the distribution of Cd between shoots and roots showed no obvious differences (Fig 3B). Additionally, 30 min and 2-hour short-term Cd exposure experiments showed decreased Cd contents in both roots and shoots of hipp19 mutants compared to WT plants (S4 Fig), collectively demonstrating that HIPP19 mediates Cd uptake rather than long-distance translocation in rice.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. HIPP19 is important for Cd uptake and facilitates Cd accumulation in root exodermis and endodermis.

(A) The contents of Cd in roots and shoots of 14 d-old WT and hipp19 mutants grown in 2 μM CdCl₂ conditions. Student’s t test, *P < 0.05; **P < 0.01. Values are means ± SD (n = 5 biological replicates). (B) Translocation of Cd from roots to shoots of WT and hipp19 mutants exposed in 2 μM CdCl₂ for 14 days. ns, not significant. (C) The contents of Cd in roots and shoots of 14 d-old WT and STTM-5p plants grown in 2 μM CdCl₂ conditions (Student’s t test, **P < 0.01; ***P < 0.001). (D) Translocation of Cd from roots to shoots of WT and STTM-5p plants exposed in 2 μM CdCl₂ ns, not significant. Values are means ± SD (n = 5 biological replicates). (E) The contents of Cd in roots and shoots of 14 d-old WT, mir408, hipp19, and mir408/hipp19 mutants grown in 2 μM CdCl₂ conditions (Tukey’s honestly significant difference, P < 0.05). (F) The contents of Cd in roots and shoots of 14 d-old WT, STTM-5p, hipp19, and STTM-5p/hipp19 plants grown in 2 μM CdCl₂ conditions (Tukey’s honestly significant difference, P < 0.05). (G) Dilution-series spot assays of yeast strain SEY6210 growth expressing HIPP19, Cd1, or empty vector YES2 in a medium containing different concentrations of Cd. (H) Growth of yeast strains shown in (G) with or without different time of 30 μM CdCl₂ treatment. (I) Cd concentrations in yeast cells expressing HIPP19, Cd1, or empty vector YES2 after incubation in a liquid medium containing 30 μM Cd for 12 h (Student’s t test, ***P < 0.001). Values are means ± SD (n = 3 biological replicates). (J) In vitro Cd binding assay of HIPP19 to Cd. Full-length HIPP19 recombinant proteins were extracted from BL21 and then incubated with 10 µM Cd for 4 h, pH = 7.4. MBP represents the E. coli trigger factor protein that fused to the N-terminus of the indicated proteins. GFP protein was used as a control (Student’s t test, ***P < 0.001). Values are means ± SD (n = 3 biological replicates). (K) Subcellular location analysis of HIPP19-GFP and GFP-HIPP19 in rice protoplasts. PIP2a-mCherry and NLS-CFP represents the localization in plasma membrane and nucleus, respectively. Bars = 5 μm. (L) Immunostaining of HIPP19-GFP transgenic rice roots under the control of HIPP19 native promoter (HIPP19pro::HIPP19CDS-GFP). Immunostaining was performed using a GFP antibody. Red represents the signal. Bars = 100 μm. The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003811.g003

In mir408 mutants and STTM-5p plants, the Cd content was significantly elevated in both roots and shoots, and the distribution ratio was comparable to that of the WT (Fig 3C and 3D). However, when we knocked out HIPP19 in mir408 mutants or STTM-5p plants, their increased Cd accumulation decreased as that in hipp19 mutants (Figs 3E, 3F, and S5A), suggesting that HIPP19 is epistatic to MIR408 and miR408-5p in Cd uptake regulation in rice.

To perform functional characterization of HIPP19, full-length HIPP19 was cloned into the yeast expression vector pYES2 and transformed into the yeast SEY6210 strain [28]. Under Cd stress, the growth status of the HIPP19 strain was considerably worse than that of the control strain, similar to the strain with Cd1 introduction (Fig 3G and 3H), which has been identified as a facilitator for Cd uptake in rice [29]. In addition, the determination of Cd concentrations revealed that yeast strains expressing HIPP19 accumulated much higher Cd than the control strains (Fig 3I), suggesting that heterologous expression of HIPP19 suppressed Cd tolerance due to enhanced Cd uptake in yeast.

Given that HIPP19 contains a typical HMA domain, we investigated whether it can directly bind to Cd ions. In vitro metal binding assays revealed that while HIPP19 exhibits a capacity for Cd, which is more than 10-fold high than that of the control GFP proteins (Fig 3J). These results suggest that HIPP19 may interact with and form a stable complex with Cd in rice, similar to the observed association of phytochelatins with Cd and metal ions [30].

To determine the subcellular localization of HIPP19, we fused GFP to the N- or C-terminus of HIPP19 and transiently introduced them in rice protoplasts and N. benthamiana cells under the control of the 35S promoter. Both GFP-HIPP19 and HIPP19-GFP were observed in the plasma membrane, cytoplasm, and nucleus (Figs 3K and S5B), implying that HIPP19 may transport or/and chelate Cd within the cells. To further confirm the cellular localization of HIPP19 in the root mature zone, we performed immunostaining using an antibody against GFP in HIPP19 promoter-driven HIPP19-GFP transgenic rice plants [31,32]. As shown in Fig 3L, HIPP19 was notably observed in the outer regions of the exodermis and endodermis. Given that hipp19 mutants exhibited reduced Cd accumulation in both shoots and roots (Fig 3A), it is possible that HIPP19 binds Cd and facilitate Cd accumulation in exodermis and endodermis cells, thereby augmenting Cd uptake in roots and accumulation in shoots in rice.

Collectively, our findings demonstrate that miR408-5p suppresses HIPP19 translation, potentially reducing HIPP19 accumulation in root exodermis and endodermis cells. This mechanism may impede Cd uptake in roots and consequently diminish Cd accumulation in shoots.

miR408-3p modulates Cd uptake via targeting UCL7

In addition to miR408-5p, we observed the induction of miR408-3p by Cd treatment, although with a lower fold change compared to miR408-5p (Figs 4A and S6A). While miR408-3p has been documented to target a subset of UCL family genes in rice [20,33–35], our examination of potential UCL targets in response to Cd revealed that only UCL7 was downregulated in rice roots following 2 μM Cd treatment (Figs 4B, 4C, and S6B). Given the inverse expressions of miR408-3p and UCL7, our subsequent investigation focused on the regulation of UCL7 by miR408-3p involved in Cd accumulation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. miR408-3p modulates Cd uptake via targeting UCL7.

(A) Time course analysis of miR408-3p accumulations in Nip rice with 2 μM CdCl₂ treatment. Student’s t test, *P < 0.05; **P < 0.01. Values are means ± SD (n = 3 biological replicates). (B) Gene structure of UCL7 and alignments of miR408-3p with target sites in UCL7 CDS and the indicated UCL7 mutated CDS. (C) Time course analysis of UCL7 expressions in Nip plants with 2 μM CdCl₂ treatment. Values are means ± SD (n = 3 biological replicates). (D) Validation of UCL7 as miR408-3p target through transient expression analysis in N. benthamiana leaves. Left: The constructs in A. tumefaciens transiently introduced in N. benthamiana leaves. Middle: Representative photograph of firefly luciferase fluorescence signals when the indicated construct combinations were introduced in N. benthamiana leaves. Right: Relative reporter activity in N. benthamiana leaves expressing the indicated construct combinations. Error bars indicate SD (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 3 biological replicates). (E) Cleavage events of miR408-3p to UCL7 mRNA shown by degradome sequencing data obtained from MIR408-OE transgenic plants. (F) Relative expressions of UCL7 in WT, MIR408-OE, mir408, and STTM-3p plants. Values are means ± SD (n = 3 biological replicates). (G) The contents of Cd in roots and shoots of 14 d-old WT, STTM-3p, and ucl7 plants grown in 2 μM CdCl₂ conditions (Student’s t test, *P < 0.05; **P < 0.01; ***P < 0.001). Values are means ± SD (n = 5 biological replicates). (H) Translocation of Cd from roots to shoots of WT, STTM-3p, and ucl7 plants exposed in 2 μM CdCl_2. ns, not significant. Values are means ± SD (n = 5 biological replicates). The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003811.g004

Transient overexpression of UCL7 alongside MIR408 in N. benthamiana significantly suppressed the LUC signal derived from UCL7 protein, compared to co-expression with MIR530 or a mutated UCL7 construct (Fig 4B and 4D). This suggests that UCL7 is repressed by mature miRNAs processed from pre-miR408. Searching for our previous degradome datasets [6], we revealed that the mRNA of UCL7 undergoes cleavage by miR408-3p at a typical miRNA-target cleavage site (Fig 4E). Consequently, we assessed the expression of UCL7 in miR408-3p gain- and loss-of-function plants. Compared to WT, UCL7 transcripts were significantly decreased in MIR408-OE plants, whereas substantially enhanced in mir408 mutants as well as in STTM-3p plants with or without Cd treatment (Figs 4F and S6C), where miR408-3p accumulation was specifically decreased and miR408-3p activity was blocked (S6D Fig) [6]. These findings strongly suggest that miR408-3p regulates UCL7 in a mRNA digestion manner.

In line with the enhancement of Cd in mir408 mutants, we found that the Cd levels in STTM-3p transgenic plants were increased, both in shoots and roots (Figs 4G and S6E). In contrast, Cd contents were detected decreased in ucl7 mutants, regardless in roots or shoots (Figs 4G, S6F, and S6G). Nevertheless, the distribution ratio of Cd between shoots and roots remained identical in STTM-3p and ucl7 plants, resembling the ratio observed in WT plants (Fig 4H). This supports the idea that miR408-3p is involved in Cd uptake rather than regulating Cd transportation through post-transcriptional control of UCL7.

UCL7 regulates Cd intake and accumulation by affecting ROS metabolites

The similar regulatory roles of miR408-3p and miR408-5p in Cd uptake rather than in Cd translocation led us to investigate whether their respective targets, UCL7 and HIPP19, share similar molecular contributions to Cd absorption in rice.

To achieve this, we first assessed the sensitivity of UCL7-transgenic yeast to Cd stress (S7A Fig). However, the findings indicated that the growth rate of yeast expressing UCL7 remained unaltered in the Cd environment (S7A and S7B Fig). Furthermore, Cd accumulation in yeast expressing UCL7 was similar to that in WT yeast, considerably lower than in yeast expressing Nramp 5 (S7C Fig), a recognized Cd transporter [36], suggesting that UCL7 does not function as a Cd transporter.

Next, we purified the UCL7 protein in E.coli and performed in vitro metal binding assays to examine its metal binding capabilities. Notably, interactions between UCL7 and Cu were readily detected (S7D Fig), aligning with the established understanding that UCLs are small copper protein [37]. By contrast, associations of UCL7 with Cd ions could not be detected (S7D Fig). These findings collectively imply that the molecular mechanism underlying the reduction of Cd content in ucl7 roots and shoots differs from that in hipp19 mutants.

ROS signaling serves as a small-molecule secondary messengers in plant development and responses to environmental stresses [38]. Recent studies have highlighted the significance of the reactive oxygen species (ROS) signaling cascade in facilitating plant adaptation to toxic heavy metals in soil [39], including Cd and aluminum, therefore we sought to determine whether UCL7 regulates Cd uptake and accumulation via influencing ROS products. We assessed the hydrogen peroxide (H₂O₂) content, as H₂O₂ possesses the longest half-life among ROS [40]. As depicted in Fig 5A, although H₂O₂ accumulation was comparable without Cd treatment in WT and ucl7 mutants, a significantly higher accumulation was observed in the shoots of ucl7 compared to WT under Cd regime, suggesting that the regulation of ROS homeostasis in response to Cd sensing is abnormal when UCL7 is deficient in rice. Additionally, under Cd stress, a similar enhancement of H₂O₂ accumulation was observed in MIR408-OE transgenic plants through H₂O₂ quantification (Fig 5A). This implies that the overproduction of miR408-3p may influence UCL7 expression and control subsequent ROS signaling in rice in response to Cd toxicity.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. UCL7 regulates Cd intake by affecting SOD activity and H₂O₂ production.

(A) Quantification of H₂O₂ contents in the shoot of WT, ucl7, and MIR408-OE seedlings exposed to mock and 2 μM CdCl₂ conditions for 14 days (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 6 biological replicates). (B) Subcellular location of UCL7-GFP in protoplasts isolated from rice shoots. Bars = 5 μm. (C) Green fluorescent signal generated by ROSGreen dye in protoplasts isolated from WT, MIR408-OE, and ucl7 seedlings with or without 2 μM CdCl₂ treatment. Bars = 25 μm. (D) The relative intensity of the fluorescent signal shown in (C) (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 10 biological replicates). (E) Quantification of SOD activities in the shoots of WT, ucl7, and MIR408-OE seedlings exposed to mock and 2 μM CdCl₂ conditions for 14 days (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 6 biological replicates). (F) Subcellular location of UCL7-GFP in protoplasts isolated from rice root tips. Bars = 5 μm. (G) Subcellular location of UCL7-GFP in rice root cells. Bars = 100 μm. (H) Green fluorescent signal generated by ROSGreen dye in root cells of WT, MIR408-OE, and ucl7 seedlings with or without 2 μM CdCl₂ treatment. Bars = 100 μm. (I) The relative intensity of the fluorescent signal shown in (H) (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 8 biological replicates). (J) Quantification of H₂O₂ contents in the roots of WT, ucl7, and MIR408-OE plants exposed to mock and 2 μM CdCl₂ conditions for 14 days (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 8 biological replicates). (K) Quantification of SOD activities in the roots of WT, ucl7, and MIR408-OE roots exposed to mock and 2 μM CdCl₂ conditions for 14 days (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 6 biological replicates). The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003811.g005

UCL8, the closest ortholog of UCL7 in rice, has been demonstrated to subcellularly localize in the cytoplasm, modulating plastocyanin content and photosynthesis [20]. Similarly, the fluorescence signal in protoplasts isolated from rice shoots expressing UCL7-GFP indicates localization of UCL7 in the cytoplasm (Fig 5B). Considering that the cytoplasm is one of the locations that produce ROS [38,41,42], we assessed ROS accumulation in cell cytoplasm of rice shoots with or without Cd treatment. Fluorescent dyes, ROSGreen, have been used to visualize cellular H₂O₂ in plants [43,44]. Under Cd treatment, we noted a significant increase in green fluorescent signals (Fig 5C and 5D). This observation implies a substantial induction of H₂O₂ accumulation, representing a source of ROS, in response to Cd treatment in rice shoot cells. Strikingly, upon assessing the fluorescence intensity that reflected H₂O₂ levels in WT, MIR408-OE transgenic plants, and ucl7 mutants after Cd treatment, we observed a pronounced increase in those plants overexpressing miR408-3p or harboring ucl7 deletions (Fig 5C and 5D), indicating that UCL7 represses H₂O₂ generation in rice, especially under Cd conditions.

Superoxide dismutase (SODs) represent a family of metalloenzymes responsible for catalyzing the dismutation of superoxide radicals into H₂O₂ [38]. Given that the cytoplasm is one of primary sites for SODs distribution [45,46], we sought to investigate the potential impact of the miR408-3p-UCL7 module on SOD enzyme activity in rice. Interestingly, we observed a significant increase in the activity of SOD in MIR408-OE plants compared to WT when subjected to Cd stress (Fig 5E). However, this heightened activity was not evident under normal water culture growth conditions (Fig 5E). Similarly, SOD activity in ucl7 mutants demonstrated an increase compared with WT under Cd treatment, while remaining consistent in the absence of Cd exposure (Fig 5E). These findings imply that the induction of miR408-3p accumulation and the suppression of UCL7 expression by Cd lead to enhance SOD activity in rice.

Because the overexpression of MIR408 and deletion of UCL7 block Cd uptake in rice roots (Figs 1C and 4G), we next attempted to determine the localization of UCL7 in root cells in addition to shoot cells. Even though isolation of protoplasts from roots was much more difficult than that from shoots, we successfully obtained transgenic protoplasts from root tips, and observed that UCL7-GFP was localized in cytoplasm of root cells as that in leaves (Fig 5F and 5G). Similar to shoot cells, the cellular H₂O₂ was limited to be accumulated in untreated roots, irrespective of WT, MIR408-OE and ucl7 mutants, since fluorescent dyes of ROSGreen in them were difficult to be observed without Cd treatment (Fig 5H and 5I). In contrast, the green signal intensity of them was dramatically enhanced when rice plants were subject to 2 μM CdCl₂ regimes (Fig 5H and 5I). Moreover, the measured H₂O₂ contents in MIR408-OE and ucl7 roots displayed much higher than those in WT root cells (Fig 5J), and the cause of this observation was similar to that we found in shoots, because the measured SOD activity in roots was substantially increased in MIR408-OE transgenic plants and ucl7 mutants compared to WT rice when plants were grown in Cd environments (Fig 5K).

Taken together, these data demonstrate that the inhibition of UCL7 by miR408-3p enhances SOD enzyme activity and promotes H₂O₂ production, ultimately leading to impair Cd absorption and accumulation in rice.

MIR408 and targets enable the generation of low-Cd rice without yield penalty

The primary source of Cd in rice grains is the soil. Given the significant decrease in Cd uptake in roots through the overexpression of MIR408 and the deletion of HIPP19 or UCL7 in rice, we sought to investigate their impact on Cd accumulation in rice grains and their potential for generating low-Cd rice. To this end, experiments were conducted in large plastic pots with soil artificially contaminated with Cd (2 mg/kg), where both WT and transgenic or mutant rice were planted together. Inductively coupled plasma mass spectrometry (ICP-MS) determination of Cd contents revealed a significant decrease of Cd accumulation in brown rice in hipp19 and ucl7 mutants (Fig 6A and 6B). This indicates that the deficiency of HIPP19 and UCL7 interferes with Cd uptake, resulting in a block of Cd accumulation in grains. More interestingly, overexpression of MIR408, which leads to a simultaneous enhancement of miR408-5p and miR408-3p, accumulates much less Cd in brown rice than in hipp19 and ucl7 single mutants (Fig 6A–6C). This suggests that the dual effects of MIR408 overproduction, inhibiting the protein level of HIPP19 and the transcription level of UCL7 concurrently, can be explored more efficiently to create low-Cd germplasm to reduce grain Cd accumulation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Manipulation of MIR408 and targets enables the generation of low-Cd rice.

(A–C) The concentration of Cd in brown rice of hipp19 (A), ucl7 (B), and MIR408-OE (C) plants compared to WT when grown in soil containing 2 mg kg⁻¹ Cd (Student’s t test, **P < 0.01; ***P < 0.001). Values are means ± SD (n = 10 biological replicates). (D–F) The concentration of Mn, Fe, Cu, and Zn in brown rice of hipp19 (D), ucl7 (E), and MIR408-OE (F) plants compared to WT when they were grown in soil containing 2 mg kg⁻¹ Cd (Tukey’s honestly significant difference, P < 0.05). Values are means ± SD (n = 10 biological replicates). The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003811.g006

At the same time, the contents of essential minerals, including Mn, Fe, Cu, and Zn, were all comparable in the grains of MIR408 overexpression lines, hipp19 and ucl7 mutants, compared with the WT, implying a significant potential for their application in the generating low-Cd rice (Fig 6D–6F). In addition, under natural field conditions, the plant architecture, and key agricultural traits, including plant height, 1,000-grain weight, number of effective panicles per plant, and number of grains per panicle, were similar between the hipp19 and ucl7 mutants and the WT plants (S8A–S8C Fig). Together with the previous finding that overexpression of MIR408 remarkably enhances grain yield in rice and several other plants [20,21,47–49], the MIR408 overexpression lines, hipp19, and ucl7 mutants generated in this study enable the production of germplasm for low-Cd rice breeding without negative effects on grain yield and agronomical traits.

The potential mechanism of HIPP19 in regulating Cd uptake and accumulation in rice

As a large family of heavy metal-associated proteins, HIPPs have been implicated in plant growth, development, and stress responses by delivering metals to transporters for metal distribution or detoxification [52–55]. A previous study showed that rice HIPP9 does not mediate Cd uptake during the vegetative growth stage, but mediates Cd translocation and restricts Cd accumulation in grains at the reproductive stage [16]. In contrast, our study reveals that HIPP19, which is targeted by miR408-5p, is expressed in root exodermis and endodermis cells, and plays a role in Cd binding and Cd uptake in rice. Strikingly, knockout of HIPP19 lead to decreased, rather than increased, Cd accumulation in grains, suggesting that HIPP19 and HIPP9 play distinct roles in Cd regulation. Given that HIPPs constitute a large gene family in rice, and HIPP9 and HIPP19 belong to distinct subclades [53], we proposed that their distinct tissue expression patterns and amino acid compositions may determine their specialized functions in Cd uptake and translocation. Similarly, HMA3, a P_1B-type ATPase, can also sequester Cd, but it detains Cd in the vacuole, which contributes to Cd translocation and accumulation in rice [28,56]. This suggests that subcellular compartmentalization and cell-type specificity endow metallochaperones with distinct roles in Cd uptake and transport.

To date, Nramp5 has been identified as the major transporter mediating Cd uptake in rice roots [36,57]. Our results showed that while the tissue localization of HIPP19 partially overlapped with that of Nramp5, both are expressed in the exodermis and endodermis [36] (Fig 3L), the two proteins did not physically interact (S9 Fig). We observed that HIPP19 enhances Cd absorption in yeast cells and localizes to the plasma membrane in addition to the cytoplasm (Figs 3K and S5B). These findings imply that HIPP19 may either directly mediate Cd uptake into cells or facilitate this process indirectly by interacting with other proteins.

Additionally, given that HIPP19 localizes to the cytoplasm and can directly bind Cd²⁺ (Fig 3J), we also propose that a negative feedback mechanism of HIPP19 regulation may be involved within the cell. In WT rice, HIPP19 may bind to Cd²⁺ to form a HIPP19-Cd²⁺ complex, thereby reducing and buffering the levels of free Cd²⁺ in the cytosol. This buffering effect likely mitigates immediate Cd toxicity and modulates Cd-related signaling. In contrast, in the hipp19 mutant or miR408-5p gain-of-function transgenic plants, where HIPP19 activity is lost or attenuated, this Cd²⁺-binding capacity is impaired, possibly leading to elevated levels of unbound cytosolic Cd²⁺. We hypothesize that this increase in free Cd²⁺ may activate cytosolic heavy metal-sensing pathways, which in turn negatively regulate the expression or activity of metal influx transporters, ultimately reducing further Cd uptake.

Therefore, we propose two alternative mechanisms by which HIPP19 may modulate Cd uptake in rice (S10 Fig). First, the localization of HIPP19 to the plasma membrane likely facilitates Cd uptake into cells [58]. Second, cytoplasmic HIPP19 binds to Cd²⁺ and thus perhaps buffer the levels of free Cd²⁺ in the cytosol, which may ultimately trigger feedback regulation of heavy metal-sensing pathways. Further studies are required to elucidate the detailed mechanism by which HIPP19 mediates Cd uptake and accumulation in rice.

In addition, HIPP19 was also detected in xylem cells and the plasma membrane at the cellular and subcellular level (Fig 3L), suggesting that HIPP19 may also play roles in Cd xylem loading or xylem-to-phloem transfer, in addition to its involvement in Cd uptake and accumulation in root cells, although the distribution of Cd in roots and shoots of hipp19 mutants was not observed altered compared to WT plants (Fig 3B). We speculate that HIPP19 functions in Cd transfer may overlap with other HIPP family proteins, particularly HIPP20, given their high protein similarities (S11 Fig).

While Cd treatment simultaneously induces the expression of both HIPP19 and miR408-5p may appear counterintuitive, this regulatory configuration is biologically plausible. On one hand, HIPP19 is transcriptionally upregulated in response to Cd exposure. On the other hand, the concurrent accumulation of miR408-5p acts post-transcriptionally to repress HIPP19 translation. These two opposing regulatory layers likely function in concert to maintain HIPP19 protein within an optimal range, which is critical for Cd stress response in rice. In the absence of miR408-5p-mediated repression, the transcriptional activation of HIPP19 could become excessively pronounced, potentially leading to Cd hyperaccumulation and consequent detrimental effects on plant growth. Thus, while transcriptional induction of HIPP19 is present, it does not diminish the importance of its post-transcriptional fine-tuning by miR408-5p. This balanced regulatory circuit ensures that HIPP19 is maintained at a level conducive to effective Cd stress management.

The potential mechanism of UCL7 in regulating Cd uptake and accumulation in rice

UCL8, which is regulated by miR408-3p, has been shown localized to the plasma membrane, and affects photosynthesis and grain yield by regulating copper homeostasis [20]. Here, we found that UCL7, another target of miR408-3p, suppressed SOD activity, thereby decreasing H₂O₂ production in the cytoplasm. Cd induces MIR408 transcription and miR408-3p accumulation, which enhances the cleavage of UCL7 transcripts and thereby reduces the pool of UCL7 transcripts available for translation. Given that UCL7 protein binds Cu²⁺ (S7D Fig), we conjecture this, in turn, may increase intracellular copper availability, consequently boosting Cu/Zn SOD activity. These sequential events may ultimately promote localized H₂O₂ production. Since H₂O₂ is widely recognized as a key physiological signal in plants [38,40,59], the enhanced H₂O₂ generated in roots and shoots is likely to affect cation channels or modulate heavy metal-related protein modifications, ultimately resulting in the inhibition of Cd uptake (S12 Fig) [60–62].

Nevertheless, the detailed molecular evidence regarding how H₂O₂ and ROS signal directly mediates Cd and heavy-metal absorption awaits further exploration. Given that MnSOD is known to accumulate under Cu deficiency conditions, whether MnSOD in addition to Cu/Zn SOD is involved in ROS accumulation activation under Cd treatment requires to be determined. Because many copper protein genes are predicted as targets of miR408-3p [63], it is possible that other UCLs or phytocyanin, besides UCL7, are regulated by miR408-3p and play roles in ROS manufacture and Cd uptake in rice roots. Additionally, we detected that miR398, which targets Cu/Zn SOD-encoding genes, was induced by Cd treatment (S13 Fig). Thus, whether miR408-3p and miR398 synergistically regulate Cu/Zn SOD activity to influence Cd uptake and accumulation warrants future investigation. Finally, as HIPP19 possesses heavy metal-binding activity and may bind Cu, we cannot exclude the possibility that it similarly modulates Cu/Zn SOD activity and ROS accumulation under Cd exposure.

Development of low-Cd rice via harnessing MIR408-derived miRNAs and their targets

Over the past two decades, several genes involved in Cd uptake, translocation, and distribution have been characterized in rice; however, almost none of these genes have been directly utilized in breeding programs to date because they often play roles in the intake or transport of essential elements [36,56,64–69]. Therefore, simply knocking out these genes in rice often raises the dilemma of balancing between reducing Cd content and compromising yields due to the reduced accumulation of essential metals. For instance, Cd accumulation has been detected to be significantly suppressed in nramp5 mutants, but plant growth was somehow destroyed due to reduced Mn intake, although a few variations of NRAMP5 have been utilized in low-Cd rice generation [36,57]. Identifying genotypic variations and selecting superior alleles of genes involved in Cd accumulation may be an effective strategy for genetic improvement of rice; however, generating potential breeding materials for breeding rice varieties through marker-assisted backcrossing remains a time-consuming endeavor [29]. Therefore, another significant finding of our study is to provide an alternative approach to develop low-Cd rice. This includes knock-out mutants of HIPP19 and UCL7 and rice lines overexpressing MIR408, none of which show any alteration in essential metals or crop major traits (Fig 6). In particular, without a repulsive issue with transgenic plants, low-Cd rice varieties could be directly produced by knocking out UCL7 and HIPP19 in a variety of high-yield cultivars by gene-editing. Hence, we presume that our findings lay the groundwork for the development of crop varieties using gene-editing approaches to produce cultivars that are low in Cd content and without yield penalties, thereby addressing crucial global food security concerns.

Plant material and growth conditions

The seeds of the Japonica rice (Oryza sativa) variety Nipponbare (Nip) were subjected to a surface sterilization process and soaked in a 3% sodium hypochlorite (NaClO) solution for 30 min. Afterward, the seeds were rinsed three to four times with sterilized deionized water (ddH₂O). We then imbibed the seeds in distilled water for 48 h in dark at 37 °C to initiate germination. These germinated seeds were subsequently planted in a hydroponic box with 96 wells (manufactured by LabStar Company, Jiangsu), which were filled with Yoshida nutrient solution maintained at a pH of 5.8. After 4 days of growth, the seedlings were exposed to a 2 μM Cd solution at specified concentrations for 10 days. This exposure occurred in a controlled environment with a 14-h light (28 °C) and 10-h dark (22 °C) cycle. Alternatively, rice seedlings that reached the three-leaf stage were transplanted into pots filled with paddy soil containing 2 mg/kg of Cd. These plants were grown until the harvesting stage at the Agricultural Experiment Station of Zhejiang University.

Construction and genetic materials

The coding sequences of HIPP19 and UCL7 were amplified from rice cDNA via polymerase chain reaction (PCR) and constructed into a modified pCAMBIA1390 vector under the regulation of a maize ubiquitin (UBI) promoter for overexpression in rice. Transgenic MIR408 overexpression (MIR408-OE) lines were generated as before [6]. A ~500-bp genomic fragment, containing the miR408 precursor sequence along with 266 bp of upstream and 253 bp of downstream flanking sequences, was PCR-amplified from Nip genomic DNA and cloned into the pCAMBIA1390 vector, downstream of the UBI promoter (S14 Fig). STTM structures targeting miR408-3p and miR408-5p were constructed via overlapping PCR, and separately cloned into pCAMBIA1390 downstream of the UBI promoter. The MIR408-OE lines (for pre-miR408 overexpression) and STTM lines (STTM-5p and STTM-3p, which specifically impair miR408-5p and miR408-3p activity, respectively) had been characterized previously [6]. To generate the rice mutants hipp19, ucl7, and mir408, we employed the CRISPR-Cas9 system and designed the sgRNA and constructs in line with previously described methodologies [70]. To determine HIPP19 and UCL7 subcellular localization within rice, GFP-HIPP19 and GFP-UCL7 fragments were inserted into the pCAMBIA2300 vector to yield the 35S::GFP-HIPP19/pCAMBIA2300 and 35S::GFP-UCL7/pCAMBIA2300 constructs. In addition, the HIPP19 fragment was inserted into the pAN580 vector, resulting in the 35S::HIPP19-GFP/pAN580 construct. Further refinements included replacing the 35S promoter with the native HIPP19 and UCL7 promoters, leading to the creation of the pro-HIPP19::HIPP19-GFP/pCAMBIA1305 and pro-UCL7::UCL7-GFP/pCAMBIA1305 constructs. For transient transfection assays, we isolated the 500-bp fragment harboring the miR408 or miR530 precursor sequence from rice genomic DNA and inserted it into the pCAMBIA2300 vector under a CaMV35S promoter. Additionally, we engineered the 35S::HIPP19-LUC, 35S::mHIPP19-LUC, 35S::UCL7-LUC, and 35S::mUCL7-LUC constructs into the modified pGreen0800 II-LUC vector, incorporating the Renilla (REN) gene to quantify LUC activity relative to internal controls. We transferred all binary expression vectors into A. tumefaciens strain AGL1 for rice transformation and strain EHA105 for tobacco leaf infiltration assays. Detailed primer sequences for plasmid construction are listed in S1 Table.

Quantitative real-time PCR

Nip and the corresponding transgenic lines were exposed to both 0 and 2 μM CdCl₂ for 12 h, 24 h, and 3 days. Total root RNA was extracted using TRIzol reagent (Invitrogen), followed by DNase I treatment to remove any contaminating DNA. Subsequently, approximately 2 μg of RNA was reverse transcribed into cDNA, which served as the template for Quantitative PCR analysis. qPCR was conducted using SYBR Premix (Invitrogen) on a Real-Time PCR System (Thermo Fisher Scientific), with the Actin gene employed as an internal control for normalization. To assess the levels of mature miRNA, stem-loop qRT-PCR was used, as previously described [6], with U6 snRNA serving as the endogenous control. Relative expression levels were calculated using the ΔΔCt (cycle threshold) method. For each qRT-PCR analysis, three biological replicates were conducted. In each biological experiment, samples from three plants (both untreated and Cd-treated groups) were pooled, homogenized, and used for RNA isolation. qRT-PCR was subsequently performed with three technical repeats per sample. The entire procedure was independently replicated three times, and the results are presented as the mean ± standard error. Details of all primer sequences used in this study can be found in S1 Table.

GUS staining

Transgenic rice seedlings expressing pro-MIR408-GUS were subjected to GUS staining to visualize gene expression [6]. The GUS staining solution was composed of 2 mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid, buffered with 50 mM of NaH₂PO₄/Na₂HPO₄, 2 mM of potassium ferricyanide (K₃Fe (CN)₆) and potassium ferrocyanide (K₄Fe (CN)₆), 10 mM Na₂EDTA, and 0.1% Triton X-100. This mixture facilitated the staining process, which took place at an optimal temperature of 37 °C for 8–12 h to achieve clear results. After the GUS staining procedure, we employed 70% ethanol as a medium to efficiently remove chlorophyll, thereby enhancing the visibility of the staining outcomes.

Luciferase imaging assays

To investigate the influence of miR408-5p on HIPP19, we prepared uniform concentrations and volumes of A.tumefaciens strain EHA105 containing either wild-type (WT) or mutated target site versions of HIPP19 in conjunction with MIR408 or equivalent control vectors. These mixtures were then simultaneously infiltrated into the leaves of N. benthamiana. At least five leaves, each from a different N. benthamiana specimens underwent infiltration. Following this, the intensity of bioluminescence was quantified using the GloMax Luminometer System (Promega, USA). The results were analyzed based on the LUC/REN ratio, allowing for a quantitative assessment of miR408-5 p’s effects on HIPP19 expression.

Degradome dataset analysis

The degradome dataset generated in our study was analyzed as previously described [6]. Briefly, the degradome libraries were constructed from MIR408-OE transgenic plants. Degradome library construction, sequencing, and data analysis were essentially described as before [6]. Reads mapping to the predicted UCL7 target sites were used to determine the positions of the 5′transcript ends using a custom Perl script.

Subcellular localization and protoplast isolation

To elucidate the subcellular localization of HIPP19 and UCL7, 35S::GFP-HIPP19, 35S::HIPP19-GFP, and 35S::UCL7-GFP constructs were introduced into rice protoplasts using a polyethylene glycol-mediated transformation technique or transiently introduced in the N. benthamiana leaf cells, as described before [71,72]. The following transformation, the cells were maintained under dark conditions at room temperature for 15–18 h to facilitate expression. The expression and subcellular localization of the constructs were visualized and captured using a confocal laser scanning microscope (FV1000 MPE; Olympus), providing valuable insights into the cellular distribution and functionality of HIPP19 and UCL7.

Cell and tissue specificity expression

To investigate the cell and tissue-specific expression patterns of HIPP19 and UCL7, constructs containing the fusion genes pro-HIPP19::HIPP19-GFP and pro-UCL7::UCL7-GFP were assembled and introduced into A. tumefaciens strain AGL1, facilitating the A. tumefaciens-mediated transformation of Nip rice. Roots from 10-day-old seedlings of transgenic lines containing the pro-HIPP19::HIPP19-GFP construct were selected for detailed immunostaining analysis. The immunostaining procedure, utilizing an anti-GFP antibody according to the methodology described before [73], facilitated the observation of GFP fluorescence under a confocal laser scanning microscope (FV1000 MPE; Olympus). Similarly, transgenic lines harboring the pro-UCL7::UCL7-GFP construct were examined using the same confocal microscopy.

Yeast experiments

The entire open reading frame (ORF) of HIPP19 and UCL7 was successfully amplified from cDNA by RT-PCR, using specific primers for each gene. The resultant cDNA fragments were then cloned into the yeast expression vector pYES2, yielding the recombinant constructs pYES2-HIPP19 and pYES2-UCL7. To assess Cd tolerance, the constructs along with the empty vector pYES2 were introduced into the yeast strain SEY6210. Transformants harboring each plasmid were cultured on an SD-uracil solid medium containing 2% galactose and varying concentrations of CdCl₂ (0, 30, or 40 μM). Growth was documented after incubation at 30 °C for 3 d, and the resulting cultures were photographed. In a parallel set of experiments examining Cd uptake in the liquid culture of yeast (strain SEY6210), each transformant underwent pre-culture overnight in a liquid media formulation complemented with 2% glucose (Glc), 0.67% yeast nitrogen base minus amino acids, and 0.2% drop-out amino acid mix. Following overnight growth, yeast cells were collected by centrifugation and washed thoroughly with Milli-Q water (Millipore). The washed cell pellets were then resuspended in a galactose-containing liquid medium (2% Gal) supplemented with yeast nitrogen base, dropout amino acid mix, and 30 μM CdCl₂, to conduct the Cd uptake study. The growth of yeast cultures was monitored at distinct time intervals by measuring the optical density at 600 nm (OD600) using a NanoDrop One spectrophotometer (Thermo Scientific, USA). The uptake assay was conducted at 30 °C with horizontal shaking at 200 rpm for 4 h. At the end of the uptake period, the cells were washed three times with Milli-Q water. The dry weight of the yeast pellets was determined after a drying period of 48 h at 65 °C. Subsequently, dried yeast biomass was subjected to 65% HNO₃ digestion for the subsequent quantification of acquired Cd. Each experimental condition was independently replicated at least three times to ensure reproducibility of the results.

Protein purification and metal binding assay

To directly assess potential metal-binding properties, the full-length coding sequences of HIPP19 and UCL7 were amplified from cDNA using RT-PCR, employing specific primers, as detailed in S1 Table. These amplified fragments were then cloned into the vector pMAL-C2X. Subsequently, the constructs were introduced into the E. coli strain Rosette (DE3) for expression. After confirming the presence of the inserted sequences, a positive strain was cultured in Luria-Bertani (LB) liquid medium (5 mL) supplemented with 50 mg/L ampicillin and incubated overnight at 37 °C with agitation at 180 rpm. The overnight cultures (2.5 mL each) were then transferred into 250 mL of LB medium to facilitate further growth. Upon reaching an OD600 of 0.6, expression of the recombinant proteins was induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and maintained at 25 °C for 12 h before harvesting the cells. Protein purification was performed in accordance with the MBP protein purification protocol provided by Smart-lifesciences, with slight modifications. The cell cultures were lysed by sonication, and the resulting MBP-HIPP19 and MBP-UCL7 fusion proteins were then incubated with 10 μM Cd and Cu in 25 mL PBS buffer (comprising 0.02% KCl, 0.8% NaCl, 0.144% Na₂HPO₄, and 0.024% KH₂PO₄, at pH 7.0) for 4 h at 4 °C. Proteins were eluted with a solution containing 10 mM maltose, 20 mM Tris-HCl, and 1 mM EDTA. Before ICP-MS analysis, the purified proteins were digested with 65% HNO3 at 120 °C for 2 h. The concentrations of the purified proteins were quantified using a BCA protein assay kit supplied by Beyotime. To guarantee the reliability and repeatability of our results, each experiment was independently conducted at least three times, following the methodologies outlined [74].

H₂O₂ and SOD measurements

Hydrogen peroxide (H₂O₂) levels in the roots of 2-week-old rice seedlings were determined using a precise methodology. Initially, fresh rice roots were harvested and pulverized into a fine powder under liquid nitrogen. To this powder, lysate, and sodium phosphate buffer were added, followed by a 30-min incubation on ice. The resultant mixture was then centrifuged at 12,000g and 4 °C for 20 min. The concentration of H₂O₂ in the supernatant was accurately measured using a hydrogen peroxide assay kit (Beyotime) according to the instructions provided by the manufacturer. The Optical Density (O.D.) at 450 nm was read using a microtiter plate reader (BioTek Microplate Reader). This standard curve was employed to quantify the H₂O₂ concentration in an unknown sample. The activity of SOD was determined using the Plant Super Oxidase Dimutase ELISA Kit (MEIMIAN), ensuring a comprehensive analysis of oxidative stress responses in rice seedling roots. The specific ROS fluorescence H₂O₂ Probe (MKBio, China) were used to visualize H₂O₂ in protoplasts and root cells, respectively. This approach follows the methodologies described before [75,76]. The relative fluorescence intensity was quantified using ImageJ software.

Transfection of rice protoplasts and western blots

Generally, the plasmids 35S-Flag-HIPP19-CDS or 35S-Flag-HIPP19-mCDS were transformed into rice protoplasts according to previously described protocols [72]. Briefly, protoplasts were isolated from 10-day-old WT and the indicated transgenic rice seedlings, which were either untreated or treated with 2 μM CdCl₂ for 24 hours. Around 100 μL protoplast suspension (containing ~2 × 10⁵ protoplasts) was transfected with 2 μg of plasmid in a 120 μL PEG solution. The transformation mixture was incubated in the dark for 15 min at 28 °C, then diluted with 1 mL W5 solution (NaCl, 154 mM; CaCl₂, 125 mM; KCl, 5 mM; MES, 2 mM, pH 5.7), and centrifuged at 120g for 3 min. Protoplasts were suspended in W1 solution (Mannitol, 0.5 M; KCl, 20 mM; MES, 4 mM, pH 5.7), transferred into 2 mL centrifuge tubes, and incubated at 28 °C for 16 h. Total Proteins were extracted from half of the transfected protoplasts, and the remaining protoplasts were used for RNA extraction and RT-PCR.

For immunoblot analyses, total proteins were isolated from the corresponding protoplasts of WT and transgenic seedlings. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to polyvinylidene difluoride membranes, immunoblotted with corresponding commercial antibodies, and detected using High-sig ECL Western Blotting Substrate (Tanon). Quantification of immunoblots was performed according to the band intensities of Flag-HIPP19 and GFP as a control, which were measured using ImageJ software. Relative band intensities were then calculated using the ratio of HIPP19/GFP for each immunoblot. All immunoblot experiments were performed at least three times, with the same conclusions.

Determination of metal concentration and translocation

Rice samples were washed six times with EDTA solution to remove extracellularly bound Cd, dried at 65 °C, and then digested in 70% nitric acid at 120 °C for 4 hours, as previously described [77–79]. Samples were diluted with Millipore-filtered deionized water. The metal contents were quantified using inductively coupled plasma mass spectrometry (ICP-MS). The measurement of metal concentration was performed at least five biological replicates, with each replicate comprising a mixture of 3 plants. The rate of Cd translocation from the root to the shoot was estimated by calculating the percentage of Cd in the shoot relative to the entire plant [56].