Rice subspecies O. sativa ssp. indica cultivar “Kasalath” and O. sativa ssp. japonica cultivar “Nipponbare” were used in the present study. Seeds were surface-sterilized using a 1% sodium hypochlorite solution prior to germination on filter paper soaked in sterilized water. After germination the seedlings were transferred to a CaCl₂ solution (0.5 mM, pH 4.5) for 3 days. The 3-day-old seedlings were then subjected to a 24 h treatment of the same CaCl₂ solution in either ambient CO₂ (400 μL·L⁻¹), termed “CK” (control check), or elevated CO₂ (600 μL·L⁻¹), termed “CO₂”. Two further treatments were generated by the inclusion of 50 μM Al to the CaCl₂ solution in either the CK conditions (termed “Al”) or to the elevated CO₂ condition (termed “Al + CO₂”). The rice seedlings were grown in an environmentally controlled glasshouse, with a photoperiod of 14 h light (26°C) and 10 h dark (23°C). The relative humidity was around 60%.

For experiments involving the application of the NO scavenger c-PTIO (2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide), six further treatments were used. The CK treatment with additional c-PTIO (10 μM) were denoted as “c-PTIO”, while an elevated CO₂ treatment using the CaCl₂ solution containing 10 μM c-PTIO was also performed “CO₂ + c-PTIO”. The Al treatments were also performed with the addition of 10 μM c-PTIO (“Al+c-PTIO”), and a final treatment combining Al + CO₂ with 10 μM c-PTIO was termed “Al + CO₂ + c-PTIO”. The final pH of all treatment solutions was 4.5.

The elongation of the root before and after various treatments (50 μM Al or 10 μM La under “CK” or “CO₂” condition) was measured using a ruler. Relative root growth was defined as the increase in root length as a percentage of the root for a treatment compared to the increase in the CK-treated plants.

The root apex of rice seedlings was cut to extract the cell wall as described by Yang et al. (2008). Briefly, the root apex (0–1 cm root tip) was ground in liquid nitrogen, after which 5 mL of 75% ethanol was added for a 20 min incubation. The homogenate was centrifuged for 8 min at 12,000 g, and the resulting pellet was washed with 5 mL acetone, then with 5 mL methanol:chloroform (1:1), and finally with 5 mL methanol. The remaining precipitate was defined as the cell wall, and was freeze-dried for further use.

The Al in the root apex and the root cell walls was extracted in 1 mL 2 M HCl with occasional shaking for 24 h, followed by centrifugation at 12,000 g for 5 min. The Al content in the supernatant was analyzed by inductively coupled plasma-atomic emission spectrometry (ICP-AES).

To extract the cell wall hemicellulose, the cell wall material was treated with sterilized water at 100°C three times, after which 24% KOH containing 0.1% NaBH₄ was added to the remaining pellet twice, for 12 h each time, before centrifugation at 12,000 g for 15 min. After each KOH extraction, the supernatant was collected and combined to yield the hemicellulose.

The content of total sugars in hemicellulose was used to quantify the hemicellulose content of the cell walls. First, 200 μL of the extracted hemicellulose, 10 μL 80% phenol (v/v), and 1 mL 98% H₂SO₄ were incubated together at 30°C for 15 min. This solution was subsequently boiled in 100°C water bath for another 15 min, after which the solution was chilled and its absorbance of 490 nm light was determined.

Exudates from the roots of seedlings subjected to various treatments were collected, and their roots were weighed. The citrate in the root exudates was isolated and measured as described by Zhu et al. (2015).

The accumulation of NO in Kas roots was visualized using a NO probe, DAF-FM DA (4-amino-5-methylamino-2,7-difluorofluorescein diacetate). First, after washing with 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-KOH (pH 7.4), the root apex was immersed in 200 μL DAF-FM DA (10 μM) for 30 min in dark. Excess fluorescence was removed by washing with HEPES-KOH (pH 7.4) three times, then visualized using an Eclipse 80i upright microscope with the following filters: EX 460-500, DM 505, and BA 510-560 (Nikon, Minato, Tokyo, Japan). The intensity of the fluorescence was calculated using Photoshop 7.0 (Adobe Systems Inc., San Jose, CA, USA).

RNA was extracted from the root apexes of the treated rice seedlings and reverse-transcribed according to Zhu et al. (2012). The total volume of the real-time PCR mixture was 10 μL, which comprised 0.2 μL each of the forward and reverse primers, 3.6 μL RNase-free water, 5 μL SYBR Premix ExTaq (Takara Bio, Inc., Kusatsu, Shiga, Japan) and 1 μL cDNA. The sequences for the gene-specific primers were as follows: OsNRAT1 (forward: 5′-GAGGCCGTCTGCAGGAGAGG-3′; reverse: 5′-GGAAGTATCTGCAAGCAGCTCTGATGC-3′); OsFRDL4 (forward: 5′-CGTCATCAGCACCATCCACAG-3′; reverse: 5′-TCATTTGCGAAGAAACTTCCACG-3′); OsSTAR1 (forward: 5′-TCGCATTGGCTCGCACCCT-3′; reverse: 5′-TCGTCTTCTTCAGCCGCACGAT-3′); and OsALS1 (forward: 5′-GGTCGTCAGTCTCTGCCTTCTC-3′; reverse: 5′-CCTCCCCATCATTTTCATTTGT-3′). Expression data were normalized with the expression level of OsHISTONE (forward: 5′-AGTTTGGTCGCTCTCGATTTCG-3′; reverse: 5′-TCAACAAGTTGACCACGTCACG-3′; Xia et al., 2010; Yokosho et al., 2011; Huang et al., 2012). Each cDNA sample was run in triplicate.

Each experiment was repeated at least three times, with one set of data shown in Results. The data were analyzed using a one-way analysis of variance (ANOVA) and a Duncan's multiple range test. Different letters on the histograms indicate statistically difference at P < 0.05.

Three-day-old Kas and Nip rice plants grown for a further 24 h in ambient CO₂ (CK) in a medium containing 50 μM Al showed a significant reduction in root length relative to CK, producing roots 69 and 52% shorter than the control, respectively (Figure 1; Supplemental Table 1). This inhibition of root elongation was significantly less severe in plants grown in elevated CO₂, with reductions of only 49% in Kas and 32% in Nip, indicating that elevated CO₂ can alleviate Al toxicity in rice. Furthermore, the elongation of the Kas and Nip roots in control conditions (without Al) showed almost no difference between ambient and elevated CO₂ treatments (Figure 1; Supplemental Table 1), showing that elevated CO₂ did not affect root growth directly. However, elevated CO₂ did not affect the root growth under other trivalent metals stress such as La (Supplemental Figure 1), indicating that elevated CO₂ only can alleviate Al toxicity. Furthermore, as elevated CO₂ had a similar effect on the root elongation of Nip and Kas in the presence of Al, and Kas is the more Al-sensitive cultivar (Figure 1; Yokosho et al., 2016), thus all of the following experiments were performed in Kas.

To uncover the potential mechanism behind the alleviation of Al toxicity by elevated CO₂, Al accumulation in root apex was determined. There was ~0.01 μg Al in the root apex in the absence of Al, which may be due to the impurity of the chemicals used (Figure 2A). Under Al stress, elevated CO₂ significantly decreased the root apex Al content when compared with the ambient CO₂ treatment, indicating that the alleviation of Al-induced root growth inhibition by elevated CO₂ might achieved by reducing Al accumulation. This conclusion was further confirmed by the reduced expression level of NRAMP ALUMINUM TRANSPORTER 1 (NRAT1) in the Al-stressed plants in elevated CO₂ compared to those in ambient CO₂ (Figure 2B), as NRAT1 is a plasma membrane-localized protein responsible for the uptake of the trivalent Al ion (Xia et al., 2010).

The amount of citrate in the root exudate was quantified to determine whether the reduced Al accumulation in roots under elevated CO₂ was due to an alteration of the Al-induced secretion of citrate (Yang et al., 2008). Although Al significantly induced the secretion of citrate (Figure 3A), there was no significant difference between the ambient CO₂ and elevated CO₂ treatments, indicating that the differences in root Al accumulation in the two conditions could not be attributed to the secretion of citrate. This conclusion was further confirmed by the similar levels of FERRIC REDUCTASE DEFECTIVE LIKE4 (FRDL4) expression between plants in the ambient and elevated CO₂ conditions (Figure 3B), which is known to be correlated well with the secretion of citrate (Yokosho et al., 2011, 2016).

Recently, more and more evidences indicate that cell wall plays pivotal roles when plant in response to Al toxicity (Horst et al., 2010; Zhu et al., 2013), thus we determined the influence of elevated CO₂ on cell wall Al accumulation in Kas root. As shown in Figure 4, less Al was accumulated in the root cell walls of Al-treated plants grown in elevated CO₂ than those in ambient conditions (Figure 4). One of the component cell wall polysaccharides, hemicellulose, was previously shown to significantly contribute to the Al-binding capacity of the cell wall (Yang et al., 2011); therefore, its involvement in root cell wall Al accumulation was investigated. As expected, the cell walls of Al-treated plants contained significantly more hemicellulose (indicated by the total sugar content in hemicellulose) than the controls; however, when Al-treated plants were grown in elevated CO₂, their cell walls contained significantly less hemicellulose than those grown in ambient CO₂ (Figure 5B). This correlated well with the reduced Al accumulation in the hemicellulose of the cell walls in plants grown in elevated CO₂ (Figure 5A), as the content of hemicellulose correlated well with the Al retention in cell wall and hemicellulose (Zhu et al., 2012), and indicated that less Al entered into the cells under elevated CO₂, making the plants more Al resistant. It is noteworthy that Al largely accumulates in the hemicellulose of the cell wall, although a 2-fold difference in hemicellulose content in the Al vs. Al + CO₂ treatments was found, there's only a 25% change in Al accumulation in the wall. One possible explanation is that there may be a threshold of the hemicellulose to bind Al, thus a 2-fold increment of hemicelulose under Al treatment does not lead to a 2-fold Al accumulation here, which needs further study.

Although, most Al is fixed in the root cell wall (Ma et al., 2004), a fraction of the metal can rapidly enter into the cells, initiating an internal detoxification mechanism. OsALS1, a tonoplast transporter localized in rice root tips, can reallocate Al from the cytoplasm to the vacuole (Huang et al., 2012), and is thus essential for the internal detoxification of Al. While the expression of OsALS1 was significantly increased in plants treated with Al, there was almost no difference in its expression between plants grown in ambient or elevated CO₂ (Figure 6), suggesting that the OsALS1-based internal detoxification mechanism is not involved in the elevated CO₂-enhanced Al resistance.

As elevated CO₂ can influence endogenous NO content under different stress, such as Fe deficiency in tomato (Jin et al., 2009) and P deficiency in Arabidopsis (Niu et al., 2013), thus we then examined root NO level. Although Al significantly induced the accumulation of NO in the rice roots, the elevated CO₂ treatment significantly reduced their NO content (Figure 7). To determine whether this reduction of NO content was involved in the enhancement of Al resistance by elevated CO₂, a NO scavenger c-PTIO was added. The c-PTIO treatment had no influence on root growth in the absence of Al, however, Al-treated plants grown in elevated CO₂ (Al + CO₂ + c-PTIO) condition had a similar level of root growth as those treated with c-PTIO in ambient CO₂ condition (Al + c-PTIO; Figure 8B; Supplemental Table 2), indicating that alleviation of Al toxicity by CO₂ depends on NO accumulation as eliminating NO abolishes this beneficial effect (Figure 8).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.