Characterization of zinc uptake and translocation visualized with positron-emitting ⁶⁵Zn tracer and analysis of transport-related gene expression in two Lotus japonicus accessions

Abstract

Background and Aims

Zinc (Zn) is an essential element for humans and plants. However, Zn deficiency is widespread and 25 % of the world’s population is at risk of Zn deficiency. To overcome the deficiency of Zn intake, crops with high Zn content are required. However, most crop-producing areas have Zn-deficient soils, therefore crops with excellent Zn uptake/transport characteristics (i.e. high Zn efficiency) are needed. Our objective was to identify the crucial factors responsible for high Zn efficiency in the legume Lotus japonicus.

Methods

We evaluated Zn efficiency by static and real-time visualization of radioactive Zn (⁶⁵Zn) uptake/transport in two L. japonicus accessions, MG-20 and B-129, that differ in Zn efficiency. The combination of visualization methods verified the dynamics of Zn accumulation and transport within the plant. We compared gene expression under a normal Zn concentration (control) and Zn deficiency to evaluate genetic factors that may determine the differential Zn efficiency of the accessions.

Key Results

The accession B-129 accumulated almost twice the amount of Zn as MG-20. In the static ⁶⁵Zn images, ⁶⁵Zn accumulated in meristematic tissues, such as root tips and the shoot apex, in both accessions. The positron-emitting tracer imaging system (PETIS), which follows the transport process in real time, revealed that ⁶⁵Zn transport to the shoot was more rapid in B-129 than in MG-20. Many genes associated with Zn uptake and transport were more highly expressed in B-129 than in MG-20 under the control condition. These gene expression patterns under Zn deficiency differed from those under the control Zn condition.

Conclusions

PETIS confirmed that the real-time transport of ⁶⁵Zn to the shoot was faster in B-129 than in MG-20. The high Zn efficiency of B-129 may be due to the elevated expression of a suite of Zn uptake- and transport-related genes.

INTRODUCTION

It is estimated that world food production must increase more than 50 % from 2012 by 2050 to adequately feed the world population (FAO, 2018). In addition, given that adequate nutritional intake is necessary, the production of nutritious crops is also required. Among essential micronutrients, zinc (Zn) is vital for both animals and plants, and Zn deficiency adversely affects physiological and biochemical functions (Coleman, 1998; Zheng et al., 2021). It was estimated that the Zn nutrition of almost 2 billion people in developing countries was inadequate in 2012 (Wessells and Brown, 2012) and thus the need to improve their health through adequate Zn intake has been advocated (Pajarillo et al., 2021). However, 49 % of the world’s soils, especially in developing countries, are deficient in Zn and unsuitable for production of Zn-rich crops. To overcome this problem, the Zn uptake and transport (i.e. Zn efficiency) of a plant must be enhanced during crop production.

To improve the Zn efficiency in plants, identification of beneficial alleles in natural populations and exploitation of the mechanism of Zn accumulation in crop breeding has been proposed (Hacisalihoglu, 2020). According to previous research, Zn-efficient plants show sustainable growth and produce acceptable yields even in Zn-deficient soils, whereas soybean, an example of a Zn-inefficient plant (Hacisalihoglu, 2020), shows interveinal chlorosis under Zn deficiency (Noulas et al., 2018). By comparing phenotypic symptoms in 19 accessions of Arabidopsis thaliana grown under Zn deficiency, the different Zn requirements and crucial regulatory genes, AtZIP4 and AtIRT3, for tolerance of Zn deficiency were identified (Campos et al., 2017). This approach will enable improvement of Zn efficiency in plants. However, for the identification of more effective factors the use of Zn-efficient plants as experimental materials is preferable.

The most economically important crops, such as rice, maize and soybean, are considered to be Zn-inefficient plants (Hacisalihoglu, 2020). Deficiency in Zn causes an increase in root length to promote Zn uptake in rice (Zheng et al., 2019), whereas in maize the root length, chlorophyll content and photosynthesis rate are reduced (Mallikarjuna et al., 2020). Given that the average Zn concentration in soils is ~1 mm in areas where these crops are grown, it is expected to be difficult to produce sufficient yields of these crops in soils that are more severely deficient in Zn (Noulas et al., 2018). Several useful candidates have been identified as Zn-efficient plants and as potential sources of Zn transport related-genes. Alfalfa (lucerne; Medicago sativa), which is classified as a Zn-efficient plant by Hacisalihoglu (2020), grows favourably in hydroponic solution that contains only 0.5 µm Zn (Printz et al., 2016). Despite the low Zn availability, alfalfa plants accumulated sufficient Zn, suggesting that Zn-efficient plants may have a strongly activated Zn-uptake mechanism. With regard to Zn efficiency-related mechanisms, the zinc-regulated, iron-regulated transporter (IRT)-like protein (ZIP) family is a well-known transporter of Zn involved in uptake from the rhizosphere by root cells. Expression of several ZIP genes is induced by Zn deficiency in rice (OsZIP4, OsZIP5 and OsZIP8; Suzuki et al., 2012) and maize (ZmZIP1, ZmZIP4 and ZmIRT1; Khatun et al., 2018). Notably, these maize ZIP genes were upregulated only in a Zn deficiency-tolerant line and not in a Zn-deficiency-sensitive line, suggesting that induction of ZIP gene expression is involved in Zn efficiency. After Zn uptake by the root cell, heavy-metal ATPase (HMA), metal tolerance proteins (MTPs) and natural resistance-associated macrophage proteins (NRAMPs) regulate Zn accumulation. Analysis using Arabidopsis AtHMA2 and AtHMA4 mutants revealed that AtHMA4 transports Zn from the roots to the shoot (Hussain et al., 2004). Arabidopsis halleri HMA4 is localized in the pericycle cells in the root xylem and is indicated to function in Zn loading from the root cell to the xylem vessels (Hanikenne et al., 2008). The HMA protein family is expressed not only in the roots but also in the shoot organs, and OsHMA2 in the nodes contributes to proper Zn distribution in rice (Yamaji et al., 2013). Arabidopsis AtHMA3 is localized to the tonoplast membrane and likely participates in vacuolar sequestration of excess Zn from the cytoplasm (Morel et al., 2009). Expression analysis has shown that Arabidopsis MTP1 is localized to the tonoplast membrane and plays a role in Zn transport into the vacuole (Desbrosses-Fonrouge et al., 2005). MTP3, a member of the MTP family, is localized to the tonoplast membrane mainly in the root and is induced by a high Zn concentration to prevent excessive Zn transport to the shoot (Arrivault et al., 2016). Analysis of metal-hyperaccumulating plants shows that NRAMP3 and NRAMP4 also play roles in Zn transport into the vacuole to avoid excessive Zn accumulation (Oomen et al., 2009). As in the case of Zn deficiency-tolerant ZIP genes, the direct or indirect function of these transporters is likely responsible for the acquisition of Zn efficiency in plants.

In this study, we used Lotus japonicus, a Zn-efficient member of the legume family (Hacisalihoglu, 2020), to investigate Zn uptake and accumulation in two natural accessions, MG-20 and B-129. We employed the positron-emitting tracer imaging system (PETIS) to observe real-time changes in radioactive Zn distribution in both accessions. PETIS is a promising technique to visualize the distribution of positron-emitting radionuclides in living plants (Suzui et al., 2019). Whole-genome sequences of MG-20 and B-129 were assembled in 2008 and 2020, respectively (Sato et al., 2008; Kamal et al., 2020). Therefore, we analysed the expression levels of genes known to regulate Zn uptake and translocation to explore the association between differences in Zn transport and gene expression.

MATERIALS AND METHODS

Plant materials and growth conditions

Seeds of Lotus japonicus accessions MG-20 and B-129 were obtained from the National BioResource Project, Biological Resource Center in Lotus japonicus and Glycine max, Faculty of Regional Innovation, University of Miyazaki, Japan (https://www.legumebase.brc.miyazaki-u.ac.jp/). The seeds were scarified with a file, soaked in Milli-Q water (Millipore, Billerica, MA, USA) overnight, then transferred to a Petri dish lined with moist filter paper, and incubated in the dark for 5 d at 24 °C. Germinated seeds were placed on a net, which was floated on continuously aerated 10 %-strength Hoagland’s solution (0.5 mm KNO₃, 0.5 mm Ca(NO₃)₂, 0.2 mm MgSO₄ and 0.1 mm NH₄H₂PO₄). Micronutrients were adjusted to full-strength Hoagland’s solution (pH 5.7) in a 2-L plastic container. The solution was renewed every 3 d. The seedlings were grown under a 16-h/8-h (light/dark) photoperiod at 24 °C. In the Zn-deficiency treatment, seedlings were grown in 10 %-strength Hoagland’s solution modified to 1/10 Zn concentration for 1 month.

Determination of mineral concentration

Shoots and roots of MG-20 and B-129 seedlings grown for 1 month under the aforementioned conditions were harvested, rinsed with Milli-Q water, and dried at 70 °C for 2 d. After drying, the shoots and roots were weighed and then homogenized with a mortar and pestle. The ground samples (50 mg each) were pre-digested overnight in a solution of 40 % nitric acid and 10 % hydrogen peroxide. Subsequently, the samples were digested in concentrated nitric acid at 140 °C. To measure metal concentrations, the digested solutions were diluted with Milli-Q water and filtered through 0.45-μm membrane filters (Millipore). After dilution with 0.1 N HNO₃, the concentrations of sodium (Na), magnesium (Mg), phosphorus (P), potassium (K), calcium (Ca), manganese (Mn), iron (Fe), nickel (Ni), copper (Cu), and zinc (Zn) were determined by inductively coupled plasma–atomic emission spectroscopy (ICP-AES) (Optima 7300 DV; PerkinElmer, Boston, MA, USA) at the Chemical Analysis Center, University of Tsukuba. To calculate the concentrations of these elements, we obtained standard solutions from Wako Pure Chemical Industries (Osaka, Japan).

Static visualization of ⁶⁵Zn accumulation

Plants grown for 1 month under the aforementioned conditions were transferred to 50 mL of 10 %-strength Hoagland’s solution containing ⁶⁵ZnCl₂ (0.1 kBq mL⁻¹; RIKEN, Wako, Japan) for 2 d. Plant roots were rinsed with 10 %-strength Hoagland’s solution three times and then ⁶⁵Zn was detected with a BAS 1800-II Bio-Imaging Analyzer (Fujifilm, Tokyo, Japan). At least three independent experiments were performed.

Real-time visualization of ⁶⁵Zn behaviour using PETIS

Real-time visualization of Zn dynamics was performed using the method established by Suzui et al. (2017). One-month-old seedlings of the two accessions were transferred from the University of Tsukuba to the Takasaki Advanced Radiation Research Institute. An acrylic board was erected in the imaging field of view of the PETIS (a modified PPIS-4800 positron imaging system; Hamamatsu Photonics, Hamamatsu, Japan) and a 30-mL plastic disposable syringe (Termo Co., Tokyo, Japan) was fixed to the acrylic board. The roots of intact plants were inserted into the syringe and the shoots were fixed on the board with thread. The light and temperature were identical to those used in the hydroponic pre-culture. Plants were acclimatized to the PETIS experimental conditions over 1 d and then transferred to 30 mL of 10 %-strength Hoagland’s solution containing 200 kBq of ⁶⁵Zn (RIKEN, Wako, Japan). The uptake of ⁶⁵Zn from the solution by the root and its translocation to the shoot were monitored every 1 min for 24 h using the PETIS. The real-time imaging, integration of image data, and time–activity curves of the ⁶⁵Zn amount in the regions of interest in the images were analysed with NIH ImageJ software version 1.53e (https://imagej.nih.gov/ij/). The counts of every 240 frames of image data were integrated into the ⁶⁵Zn distribution data at 4 h per frame. The region of interest for the shoot area included the entire shoot enclosed in a square. To compare the three independent data, we first set the ⁶⁵Zn signal intensity in the whole plant of each accession as 100 %, and then the percentage relative intensity of ⁶⁵Zn in the shoot was calculated. To evaluate the increase in signal intensity of ⁶⁵Zn in the shoot between 8 and 16 h, a linear approximation of the time–activity curve was conducted for the hourly values and the inclination (i, difference in relative intensity per hour) was determined. Significant differences in inclination of the shoot between MG-20 and B-129 were evaluated using one-way ANOVA. In this experiment, three independent PETIS measurements were performed.

Acquisition of Zn transporter homologous genes in L. japonicus

Full-length coding sequences of homologous genes in the two accessions were obtained from miyakogusa.jp version 3.0 (http://www.kazusa.or.jp/lotus/summary3.0.html) and the L. japonicus Gifu genome database (http://viewer.shigen.info/lotus/index.php). For other plant species, we referred to sequence databases for Arabidopsis (TAIR; https://www.arabidopsis.org/) and rice (RAP-DB; http://rapdb.dna.affrc.go.jp/), and conducted a BLAST search of nucleotide sequences in the NCBI databases (https://blast.ncbi.nlm.nih.gov/). For each gene in L. japonicus, the gene with the highest similarity to the functionally identified gene in A. thaliana was selected by means of a BLAST protein search.

Gene expression analysis

The root or shoot samples from 1-month-old seedlings were flash-frozen in liquid nitrogen and ground with a mortar and pestle. Total RNA was extracted using the RNeasy Plant Mini Kit (Qiagen, Venlo, The Netherlands) and the RNA concentration was calculated with a Micro UV-Visible spectrophotometer (Tomy Digital Biology, Tokyo, Japan). Complementary DNA was synthesized with ReverTra Ace (Toyobo, Tokyo, Japan) with random primers. Quantitative PCR was performed using the PicoReal Real-time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) with the GoTaq qPCR Master Mix (Promega, Madison, WI, USA). Quantitative PCR reactions were performed on four biological replicates. Significant differences between gene expression levels were assessed with Student’s t-test. The sequences of qPCR primers, including the reference gene, are listed in Supplementary Data Table S1.

Phylogenetic tree reconstruction

The full-length coding sequences from L. japonicus and their orthologues in other plant species were converted to amino acid sequences and aligned with ClustalW. Phylogenetic trees were constructed using the maximum likelihood method with MEGA version 5.05. A bootstrap analysis with 2000 replications was performed to assess support for the tree topology.

RESULTS

Plant biomass and mineral concentration of L. japonicus accessions

The dry-weight biomass of the L. japonicus accession MG-20 was approximately three times higher than that of B-129, which has a dwarf phenotype, after growth for 1 month (Table 1). The concentrations of ten elements in the shoot and root were measured (Table 2). The concentrations of the macronutrients P, K and Ca were almost equal in the shoot and root in both accessions. The concentration profiles indicated that Na, Mn, Ni, Cu and Zn were more highly accumulated, both in the shoot and root, in B-129 compared with MG-20. Concentrations of these elements in B-129 were 1.5- to 3.4-fold of those in MG-20. The Fe concentration in the shoot was 1.6-fold higher in MG-20 than that in B-129; however, Fe accumulation in the root of B-129 was 1.7-fold higher than that in the root of MG-20. This profile indicated high uptake activity of Na, Mn, Fe, Ni, Cu and Zn in the root, and low Fe translocation to the shoot, in B-129. Among these highly absorbed minerals in B-129, the Zn efficiency of L. japonicus was the focus of the present research.

Table 1.

Plant biomass of Lotus japonicus accessions MG-20 and B-129

Accession	Dry weight (mg)
	Shoot	Root
MG-20	54.4 ± 10.9	18.7 ± 3.5
B-129	16.8 ± 6.2**	5.6 ± 2.1**

Data are for the root and shoot of MG-20 and B-129 seedlings grown under hydroponic culture for 1 month.

Values are mean ± standard deviation (n = 5).

Asterisks for B-129 indicate significant differences of B-129 values relative to MG-20 (one-way ANOVA). **P < 0.01.

Table 2.

Mineral concentrations in Lotus japonicus accessions MG-20 and B-129

Element	Accession	Concentration (μmol g−1 DM)
		Shoot	Root
Na	MG-20	5.7 ± 1.3	17.2 ± 2.9
	B-129	18.7 ± 8.3**	54.3 ± 24.0**
Mg	MG-20	129.6 ± 6.5	118.0 ± 10.2
	B-129	121.1 ± 10.2	114.3 ± 19.4
P	MG-20	242.0 ± 10.8	521.5 ± 30.1
	B-129	295.3 ± 16.0**	438.0 ± 42.0**
K	MG-20	1128.4 ± 69.7	1949.4 ± 94.8
	B-129	1041.4 ± 95.7	1586.4 ± 152.4**
Ca	MG-20	353.0 ± 15.4	69.3 ± 2.3
	B-129	302.7 ± 45.4*	70.2 ± 6.1
Element	Accession	Concentration (nmol g−1 DM)
		Shoot	Root
Mn	MG-20	514.6 ± 49.1	618.9 ± 77.1
	B-129	786.6 ± 72.6**	2113.4 ± 739.6**
Fe	MG-20	1293.5 ± 122.2	1905.9 ± 247.1
	B-129	810.8 ± 80.6**	3254.2 ± 473.6**
Ni	MG-20	23.0 ± 10.5	30.8 ± 7.9
	B-129	45.1 ± 14.6*	79.8 ± 20.4**
Cu	MG-20	50.1 ± 7.1	63.2 ± 5.9
	B-129	96.5 ± 9.5**	135.1 ± 28.3**
Zn	MG-20	594.2 ± 79.6	529.5 ± 36.8
	B-129	1215.4 ± 164.9**	936.3 ± 288.2*

Data are for the root and shoot of MG-20 and B-129 seedlings grown under hydroponic culture for 1 month.

Mineral concentrations were determined by ICP-AES.

Values are mean ± standard deviation (n = 5).

Asterisks for B-129 indicate significant differences in B-129 values relative to MG-20 (one-way ANOVA). *P < 0.05, **P < 0.01.

Observation of ⁶⁵Zn distribution

Given the high Zn accumulation of B-129, Zn localization was investigated by means of static radiological imaging. Compared with MG-20, B-129 accumulated high amounts of ⁶⁵Zn in the root and shoot, but especially in the shoot stem (Fig. 1). In both accessions, ⁶⁵Zn accumulated in the nodes, root tips, and newly developing leaves (Fig. 1; arrowheads). The nodes were estimated to be slightly thicker than the internodal stem; therefore, the high ⁶⁵Zn density might be the result of the thickness of the tissue. However, the root tips and developing leaves were not thicker than other root tissues and fully developed leaves, respectively. These sites of ⁶⁵Zn accumulation indicated that meristematic tissues require Zn nutrition. Furthermore, no difference in the organs and tissues that accumulated Zn was observed between the two accessions.

Fig. 1.

Distribution of ⁶⁵Zn radioactivity after application for 2 d in L. japonicus accessions MG-20 and B-129 visualized by static radio-imaging. Plants were grown hydroponically for 1 month in 10 %-strength Hoagland’s solution and subjected to ⁶⁵Zn application. The ⁶⁵Zn radioactivity in the plants was detected with a BAS 1800-II Bio-Imaging Analyzer. Arrowheads indicate nodes, root tips and developing leaves. Scale bar = 1 cm.

Observation of ⁶⁵Zn uptake and translocation by PETIS

A montage of PETIS images is shown in Fig. 2A. The real-time images of ⁶⁵Zn distribution obtained by PETIS indicated the rapid translocation of ⁶⁵Zn from the root to the shoot in B-129. Translocation of ⁶⁵Zn to the shoot was apparent after 4 h (second frame) in both accessions, and the relative radioactivity indicated by the pseudo-colour revealed the higher intensity of ⁶⁵Zn signals in the lower part of the stem in B-129. From 8 to 12 h, ⁶⁵Zn signal was first observed in the leaves of B-129, whereas MG-20 showed radiation in the stem but mostly still in the root. After treatment for 16 h, a high-intensity ⁶⁵Zn signal was observed in the shoot stem of B-129, whereas in MG-20 absorbed ⁶⁵Zn still remained in the root. Radiation of ⁶⁵Zn was barely detected in the root of B-129 after treatment for 24 h, but was still clearly detectable in the root of MG-20.

Fig. 2.

Observation of ⁶⁵Zn uptake and translocation in L. japonicus accessions MG-20 and B-129 by PETIS. (A) Real-time images of ⁶⁵Zn distribution in MG-20 and B-129 obtained by PETIS. Radiation intensity is represented as blue when weak and red when strong, as shown in the colour bar. Black line at bottom left in the first panel of (A) is Scale bars = 5 cm. The original images acquired at 1-min intervals were integrated into composite images for 4-h intervals. Red and blue boxes indicate the region of interest for the time–activity curve in (B). (B) Time–activity curve of ⁶⁵Zn relative intensity in the shoot of the two accessions. PETIS was performed with three independent biological replicates. The time–activity curve shows the mean values of three replicates; error bars indicate the standard error. The dashed red lines marking the period from 8 to 16 h indicate the interval for which the inclination of each ⁶⁵Zn translocation (i, the difference in relative intensity per hour) was calculated. The values of i are the inclination of each curve.

The rate of ⁶⁵Zn accumulation was represented by a time–activity curve, especially by the degree of inclination (Fig. 2B). We calculated the inclination (i) from 8 to 16 h when the translocation activity of ⁶⁵Zn began to differ visibly between the two accessions. The inclination of the increase in ⁶⁵Zn accumulation in the shoot from 8 to 16 h was calculated to be i = 0.023 ± 0.006 for MG-20 and i = 0.039 ± 0.014 for B-129. A one-way ANOVA revealed that the inclination of the accumulation rate in the two accessions did not differ significantly (P = 0.15), although B-129 clearly showed a tendency for higher translocation activity than MG-20. These results indicated that ⁶⁵Zn was more rapidly accumulated in the shoot in B-129 than in MG-20. The ⁶⁵Zn radioactivity in the shoot of both accessions began to attain equilibrium after ~24 h.

Expression of Zn transport-related genes

Expression analysis of Zn transport-related genes was conducted to explore the genetic basis for the difference in Zn efficiency in the two accessions. The elemental dynamics from the rhizosphere to the shoot involved several steps in the transport, namely from the rhizosphere to the root cells, transport in the root cell cytoplasm, xylem loading from the root cells to the xylem vessels, and transport within and from the xylem vessels. With reference to previous research, we investigated gene expression levels in the Zn influx transporter ZIP family, the intracellular Zn transport-related genes, NAS and Zn vacuole sequestration transporters, and the Zn efflux transporter HMA. Among these Zn transport-related genes, we chose LjIRT1, LjIRT3, LjZIP1, LjZIP4, LjZIP5, LjZIP6, LjZIP7, LjNAS1, LjHMA2, LjHMA4, LjYSL1 and LjZIP2 for analysis. For each gene, the corresponding gene of A. thaliana whose function was already known was first obtained from the TAIR database. The L. japonicus gene with the highest similarity based on its amino acid sequence was selected as the orthologous gene (Supplementary Data Fig. S1).

First, we examined the genes that regulate Zn uptake from the rhizosphere to the root cells (Fig. 3A). Most ZIP family transporters play a role in Zn uptake for influx from the rhizosphere into the root cells; however, AtZIP1 is a transporter localized to the tonoplast membrane that discharges Zn from the vacuole to the cytoplasm (Milner et al., 2013). A phylogenetic tree indicated that the LjZIP1 proteins of MG-20 and B-129 were genetically similar to AtZIP1 (Supplementary Data Fig. S1A). However, both LjZIP1 proteins were distinguished by at least two amino acids from AtZIP1 and were estimated to be similar to MsZIP1 from M. sativa (Supplementary Data Fig. S1B). Given that MsZIP1 plays a role in Zn uptake into root cells (Cardini et al., 2021), the present research suggested LjZIP1 is categorized as a Zn uptake transporter. Under the normal Zn concentration, LjIRT1, LjIRT3, LjZIP1, LjZIP4 and LjZIP7 showed significantly higher gene expression levels in B-129 than in MG-20, except for LjZIP5 and LjZIP6. The difference in expression levels of these genes ranged from 2.07- to 3.20-fold higher in B-129 compared with MG-20, with LjIRT3 being particularly prominent. It was noteworthy that none of these genes showed a distinctly low expression level in B-129, suggesting that Zn uptake activity was strongly activated in B-129. Among these genes, LjIRT3, LjZIP1, LjZIP4 and LjZIP5 expression was significantly induced by Zn deficiency in both accessions. MG-20 and B-129 showed induction of LjIRT3 expression by 19.7- and 5.9-fold, LjZIP1 by 7.6- and 6.7-fold, LjZIP4 by 7.4- and 3.4-fold, and LjZIP5 by 2.6- and 1.7-fold, respectively. These results suggested that the gene expression level was innately higher in B-129, whereas the induction ratio under Zn deficiency was higher in MG-20.

Fig. 3.

Transcriptional expression of Zn transport-related genes in the root of L. japonicus accessions MG-20 and B-129 under normal and Zn-deficiency conditions. Total RNA was isolated from the roots of 1-month-old seedlings and the difference in transcriptional expression between the two accessions was detected by qPCR. The genes analysed were involved in (A) Zn uptake from the rhizosphere into root cells, (B) intercellular transport up to loading of xylem vessels, and (C) vacuolar sequestration. ACTIN7 was used as an internal reference gene. N, normal Zn concentration; L, Zn deficiency. Asterisks and daggers above the bars indicate significant differences in gene expression in B-129 compared with MG-20 under normal and Zn-deficiency conditions, respectively. *P < 0.05, **P < 0.01, ^†P < 0.05, ^††P < 0.01. Error bars indicate the standard deviation. Analyses were performed with four biological replicates.

After its uptake by the root cells, Zn forms chelate complexes with nicotianamine in the cytoplasm and is transported in the cytoplasm to the xylem parenchyma cells. Nicotianamine is synthesized by nicotianamine synthase (NAS) and NAS expression is affected by the amount of Zn or other minerals absorbed in the plant (Higuchi et al., 1999; Johnson et al., 2011). We examined the expression of LjNAS1, an orthologue of AtNAS1, and observed that LjNAS1 expression in B-129 was 3.11-fold higher than that in MG-20 under the normal Zn concentration (Fig. 3B). Similar to the Zn uptake-related genes shown in Fig. 3A, LjNAS1 expression was significantly upregulated by Zn deficiency, with 10.9- and 3.5-fold increases in expression level observed in MG-20 and B-129, respectively.

After its transport to the xylem parenchyma cells, Zn is released from nicotianamine and is loaded into xylem vessels as an ionic form via the HMA family (Cornu et al., 2015). Among HMA family members, HMA2 and HMA4 are efflux transporters known to play roles in Zn xylem loading and regulate long-distance Zn transport (Takahashi et al., 2012; Hanikenne et al., 2008). Lotus japonicus HMA2 and HMA4 were resolved in the same clade as Arabidopsis HMA2 and HMA4, suggesting that they may have similar Zn efflux functions in the xylem (Supplementary Data Fig. S1C). In addition, we selected YSL1 and ZIP2, which participate in loading into and reuptake from xylem vessels, based on studies of the Zn-efficient plant M. sativa (Milner et al., 2013; Cardini et al., 2021; Kaur and Garg, 2021). The genes LjHMA2, LjHMA4 and LjYSL1 showed 1.82-, 2.12- and 1.64-fold higher expression levels, respectively, in B-129 than in MG-20 under the normal Zn concentration. The LjYSL1 expression level in B-129 was higher than that in MG-20, but in both accessions the expression levels were low. LjZIP2 showed similar expression levels in both accessions, suggesting that the ability for reuptake from the xylem was not strongly divergent in the two accessions. The Zn-deficiency treatment drastically induced increased expression of LjHMA4 and LjYSL1 in both accessions. However, LjZIP2 was significantly suppressed with 0.05- and 0.09-fold higher gene expression in MG-20 and B-129, respectively, compared with that under the normal Zn concentration.

We also considered the possibility that Zn sequestration activity in vacuoles in the root were lower in B-129 because of rapid and efficient Zn transport to the shoot. We examined the expression levels of LjHMA3, LjNRAMP3, LjNRAMP4, LjMTP1 and LjMTP3, which are homologues of Zn vacuole sequestration-related genes (Fig. 3C). For each gene in L. japonicus, the gene that showed the highest homology with Arabidopsis was selected (Supplementary Data Fig. S1C–E). The expression level of vacuolar sequestration-related genes was almost identical between MG-20 and B-129, although the expression levels of LjHMA3 and LjMTP1 were rather higher in B-129 than those in MG-20. Under Zn deficiency, the expression of LjHMA3, LjNRAMP4 and LjMTP3 in MG-20 was slightly suppressed, whereas the expression levels of these genes in B-129 were unchanged.

In summary, the expression levels of the Zn transporter genes LjIRT1, LjIRT3, LjZIP1, LjZIP4, LjZIP7, LjNAS1, LjHMA2, LjHMA4, LjYSL1, LjHMA3 and LjMTP1 were higher in B-129 than in MG-20 under the normal Zn concentration, which suggested that Zn uptake, Zn intracellular mobility and Zn loading into the xylem were higher in B-129. The similar gene expression profiles of Zn uptake- and intracellular transport-related genes in MG-20 and B-129 suggested the existence of upstream transcription factors or other factors that induce the expression of these genes, and the regulation by these factors might show higher activity in B-129. Under Zn deficiency, expression of LjIRT1, LjZIP1, LjZIP6, LjZIP7, LjHMA2 and LjZIP2 remained higher in B-129, but the expression levels of LjIRT3, LjZIP4 and LjHMA4 were similar in both accessions because of the high induction ratio in MG-20. Therefore, the induction pattern under Zn deficiency differed from that under the normal Zn concentration.

Gene expression analysis of candidate factors inducing Zn transport-related genes

To explore the Zn transport regulatory mechanisms in L. japonicus, we compared the gene expression levels of the candidate regulators that induce Zn transport-related genes. The elemental concentration analysis showed that B-129 accumulated less Fe in the shoot compared with the root. Iron deficiency in soybean positively induces the expression of genes involved in Fe and Zn uptake, and maintains its homeostasis (Moran Lauter et al., 2014). It is currently unknown whether the B-129 shoot is in a Fe-deficient state, but Fe homeostasis may differ between B-129 and MG-20. In the maintenance of nutrient homeostasis in the plant, PHR1 is a master regulator of gene expression of various macro- and micronutrient transporters, particularly contributing to Zn, Fe, P and sulphur homeostasis (Kumar et al., 2021). Therefore, we selected L. japonicus PHR1 (LjPHR1) as a candidate regulator of Zn efficiency in both accessions. In addition, IRON MAN 1 (IMA1), which shares a functional consensus motif at the C-terminus in various plant species, encodes peptides that enhance Fe uptake, and overexpression of IMA1 results in increased Fe and Zn accumulation in Arabidopsis (Grillet et al., 2018). Therefore, its homologous gene, L. japonicus IMA1 (LjIMA1), identified by Grillet et al. (2018), was also selected as a candidate genetic factor that contributes to Zn efficiency. We compared the sequences of AtIMA1 and LjIMA1 (from MG-20) and observed 42 % homology in the entire sequence and 50 % in the consensus motif (Fig. 4A). Expression of LjPHR1 in the root was similar in both accessions and its expression in the shoot of MG-20 was similar to that in the root, whereas its expression in the shoot of B-129 was significantly lower than in the root (Fig. 4B). IMA1 is presumed to act as a mobile signal peptide from the shoot to the root; therefore, its expression in the shoot may induce root Zn transport-related genes. However, significantly higher expression of LjIMA1 in the shoot of MG-20 was detected (Fig. 4C). These results indicated that these transcription factors and peptides were not responsible for the greater Zn efficiency of B-129, suggesting that other factors were active.

Fig. 4.

Transcriptional expression of putative regulators for Zn efficiency in L. japonicus accessions MG-20 and B-129. (A) Amino acid alignment of A. thaliana AtIMA1 and L. japonicus MG-20 LjIMA1. The red box indicates the consensus motif of plant IMA1. (B) Expression analysis of the putative regulator of Zn efficiency LjPHR1 in the root and the shoot. (C) Expression of LjIMA1 in the root and the shoot. The inset shows a magnified graph of gene expression in the root. Error bars indicate the standard deviation. Analyses were performed with four biological replicates. **P < 0.01.

DISCUSSION

In this study, we used radiological imaging to visually compare Zn dynamics between two L. japonicus accessions with different Zn efficiencies, and assumed that Zn uptake and transport related-genes were activated in the more efficient accession, B-129. Importantly, the goal of this study was not to identify just a single transporter gene that determines Zn transport, but to understand the Zn dynamics in the Zn-efficient plant and to identify a regulatory gene that influences the Zn efficiency.

We first confirmed that B-129, the genome of which was sequenced in 2020, accumulates almost twice the amount of Zn as MG-20 (Table 2; Fig. 1). The PETIS analysis showed that the high accumulation of Zn in B-129 is due to its rapid transport from the root to the shoot (Fig. 2). If Zn is predominantly transported as a result of transpiration activity, MG-20, which produces greater biomass, should show a high Zn translocation rate. However, despite the higher shoot biomass of MG-20, Zn accumulation was higher in B-129 (Tables 1 and 2; Fig. 1). This indicated that enhanced activity of a Zn translocation mechanism, rather than its transpiration capacity, contributed to Zn accumulation in B-129. Expression analysis showed that the expression level of each Zn uptake- and translocation-related gene was distinctly higher in B-129 than in MG-20, suggesting that these genes were responsible for the greater Zn efficiency in B-129 (Fig. 3A, B). The biological significance of the induction of these genes can be explained by their functions revealed in previous studies of Arabidopsis and rice, for example; therefore, it suggests that this Zn transport pathway may be widely conserved among plant species. This observation is important for breeding, as knowledge of the regulation of these genes is crucial for breeding crops with increased Zn efficiency.

To produce a highly Zn-efficient crop, further studies are needed to determine whether enhanced Zn uptake induces Zn accumulation in the edible parts. In an experiment using cassava (Manihot esculenta), heterologous overexpression of Arabidopsis ZIP1, which is involved in Zn uptake in the root, resulted in increased root Zn content but induced low leaf Zn content and leaf necrosis (Gaitán-Solís et al., 2015). It was suggested that regulation of a single transporter is not sufficient to produce high Zn-accumulating crops and that proper regulation of the functions of a suite of Zn transport-related genes is required. In the present static imaging of ⁶⁵Zn, both MG-20 and B-129 accumulated ⁶⁵Zn in mitotic tissue (Fig. 1). Thus, although B-129 exhibited high Zn uptake and translocation activities (Fig. 2B), Zn transport from the root to the meristem is expected to be properly regulated in both accessions. Comparing gene expression levels in MG-20 and B-129, we observed that LjIRT1, LjIRT3, LjZIP1, LjZIP4, LjZIP7, LjNAS1, LjHMA2 and LjHMA4 expression in B-129 was consistently two to three times higher than that in MG-20, whereas no genes were expressed at particularly high levels in B-129 (Fig. 3A, B). Therefore, it is suggested that the pattern of gene expression involved in multiple transport processes is optimized to ensure the efficient transport of Zn from its uptake in the root to its accumulation in the meristem.

The genes involved in vacuolar sequestration showed similar expression levels in both accessions under normal Zn concentration or Zn deficiency, except for LjHMA3 and LjMTP1 (Fig. 3C). This finding suggested that the ratio of the Zn amount sequestered into vacuoles to the amount of Zn absorbed from the rhizosphere might be higher in MG-20 than in B-129. Expression of MTP1 in Arabidopsis and MTP3 in Brassica napus promotes vacuolar sequestration of Zn and increases root Zn content (Desbrosses-Fonrouge et al., 2005; Gu et al., 2021). These genes involved in vacuolar sequestration of Zn might be responsible for the high Zn content in the root of MG-20 after treatment for 24 h (Fig. 2A). In the regulation of Zn distribution between the root and the shoot, LjZIP2 might be involved in the high Zn content in the B-129 shoot. MtZIP2 was previously reported to be a transporter for reuptake of Zn from the xylem sap to the root cell in Medicago truncatula (Burleigh et al., 2003). The expression level of LjZIP2 was almost identical in the two accessions of L. japonicus (Fig. 3B), suggesting that the ratio of the amount of Zn reabsorbed from the xylem sap to the amount of Zn loaded into the xylem sap might be lower in B-129 than in MG-20. Under Zn deficiency, the expression of LjZIP2 was significantly suppressed in both accessions (Fig. 3B). This suppression might be due to maintenance of the shoot Zn concentration by suppressing Zn reuptake from the xylem sap to the root.

The constant expression ratios of several Zn uptake- and transport-related genes in MG-20 and B-129 under the normal Zn concentration suggested the existence of factors that integrally regulate the expression levels of these genes. Previous reports have shown that PHR1 and IMA1 contribute to the enhancement of Zn accumulation, and are known to be master regulators that also control the accumulation of other trace elements in addition to Fe and Zn (Grillet et al., 2018; Kumar et al., 2021). The concentrations of Na, Mn, Ni, Cu and Zn in the root and shoot of B-129 were higher than those in MG-20, whereas the shoot Fe concentration in B-129 was lower than that in MG-20 (Table 2). Given that the Fe balance was drastically different in the two accessions, we first hypothesized that this Fe imbalance may be the cause of the difference in Zn accumulation. In A. thaliana, the uptake of several metal elements other than Fe, such as Zn, Mn and Cu, is increased in the FRD3 mutants with suppressed Fe translocation activity from the root to the shoot (Delhaize, 1996; Durrett et al., 2007). It is thought that the shoot Fe deficiency activates the uptake mechanism of Fe in the root and, as a secondary effect, the uptake of several elements other than Fe is also increased. However, in the present experiment, despite the relatively low shoot Fe concentration in B-129, the expression level of LjPHR1 in the shoot was similar in MG-20 and B-129, and that of LjIMA1 was low in B-129 (Fig. 4B, C). These results suggested that the high expression levels of Zn transport-related genes in the B-129 root were unlikely to be regulated by the expression of LjPHR1 and LjIMA1, and that a Zn uptake and translocation mechanism may be independent of the LjPHR1 and LjIMA1 signalling pathways. We initially focused on the reported Zn accumulation regulators, PHR1 and IMA1. However, concentrations of minerals other than Zn also differed between the two accessions (Table 2). Therefore, it is possible that mineral homeostasis regulators, whose relevance to Zn has not yet been determined, are involved in promoting Zn uptake and translocation. Indeed, Zn uptake is considered to be regulated by crosstalk with multiple elements (Kumar et al., 2021). Recently, it was reported that Zn accumulation was affected by nitrogen application (Xue et al., 2021). If a novel factor that induces Zn uptake and transport genes is identified, the production of high Zn-accumulating crops may be feasible. Furthermore, the gene expression profile under Zn deficiency suggested that other mechanisms different from the innate regulation in B-129 functioned to enhance Zn uptake as an adaption to Zn deficiency. Zinc deficiency induced expression of LjIRT3, LjZIP1, LjZIP4 and LjZIP5 in both accessions (Fig. 3A). Importantly, the induction intensities of these genes were not uniform and were particularly strong for LjIRT3 and LjZIP4, suggesting that a regulator other than the innate regulator of B-129 may contribute to Zn uptake under Zn deficiency. In Arabidopsis, suppression of IRT3 or ZIP4 alone has no effect on Zn accumulation, and accumulation of Zn is reduced in mutants in which multiple ZIP genes are suppressed (Lee et al., 2021). As discussed herein, these findings also indicate that it is difficult to produce crops that show high Zn accumulation by controlling these genes alone and that it is important to identify upstream factors.

Assuming an integrative gene expression mechanism, the promoter region of ZIP families might contain a motif to which certain other transcription factors could bind to regulate Zn efficiency. In A. thaliana, Zn deficiency-responsive ZIPs are regulated by bZIP19 and bZIP23 transcription factors, and they can bind to 10-bp Zn deficiency-responsive elements (ZDRE; RTGTCGACAY) in the promoter region (Assunção et al., 2010). However, no ZDRE sequences were detected in the 3000-bp promoter regions of the LjZIP family members (LjIRT1, LjIRT3, LjZIP1, LjZIP4 and LjZIP7) in both accessions. Although we cannot exclude the possibility that a ZDRE may be present in a more upstream region, its presence within 2000 bp of the promoter region in Arabidopsis suggests that different transcription factors are likely to be involved in L. japonicus. The New PLACE database (https://www.dna.affrc.go.jp/PLACE/?action=newplace; Higo et al., 1999) predicted several binding sites for biotic and abiotic stress-related transcription factors in the promoter region of LjZIP genes, but no binding sites for Zn response-related transcription factors were detected in the 3000-bp promoter region. In addition, the phytohormone balance in B-129 might differ from that in MG-20, because Zn status also influences hormonal balance (Kaur and Garg, 2021). Zinc deficiency decreases the synthesis of indole-3-acetic acid (auxin), cytokinins and gibberellins, resulting in lower biomass yields (Navarro-León et al., 2016). Although the phytohormone contents in MG-20 and B-129 are poorly investigated, a difference in germination response to abscisic acid between the two accessions has been reported and the hormonal regulation of germination is suggested to differ (Bandana et al., 2009). In the future, it might be necessary to focus on the phytohormone contents to identify the factors involved in the regulation of Zn uptake and transport in L. japonicus.

Conclusions

The aim of this study was to provide visual evidence that Zn efficiency in L. japonicus reflects rapid Zn uptake and translocation. Importantly, it was revealed that a series of Zn uptake- and transport-related genes, rather than a single one, are highly expressed in the high Zn-efficiency accession B-129. This result suggests that high Zn-accumulating plants cannot be generated by editing only a few transporter genes, and that identification of regulators that properly control the expression ratio of relevant genes is important. Furthermore, gene expression analysis under Zn deficiency suggests that the upstream regulators of Zn uptake might differ under normal and Zn-deficient conditions. This study shows that further understanding of the complexity of the mechanisms of Zn efficiency is essential for breeding crops with high Zn accumulation.

SUPPLEMENTARY DATA

Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Table S1: specific primers of Zn transport-related genes in MG-20 and B-129. Figure S1: phylogenetic tree for Zn transporters of L. japonicus and A. thaliana.

FUNDING

This work was supported by JSPS KAKENHI grants JP24780054 and JP21K21338.