Development of ammonium nitrate free nutrient media for aluminium toxicity tolerance screening of rice genotypes from North-Eastern India

Abstract North-Eastern India is blessed with a vast diversity of rice genotypes with varying yielding abilities. However, the predominant strong soil acidity induced nutrient stresses from the toxicity of aluminum (Al3+) ions often causes sub-optimal rice productivity. The lack of suitable nutrient media for the screening of aluminum (Al) toxicity tolerance of rice genotypes is one of the limiting factors in Al tolerant varieties' development. Modified Magnavaca's solution has been the most recommended nutrient solution for this purpose where ammonium nitrate is the primary nitrogen source. However, strict regulations related to the handling and storage of ammonium nitrate in India limit the preparation of Modified Magnavaca’s solution for laboratory use. Here, a modified ammonium nitrate-free formulation based upon the Magnavaca's solution has been proposed. The modified formulation was found to have 160.9 µM of active Al3+ concentration when 550 µM of aluminum chloride hexahydrate (AlCl3.6H2O) was added at pH 4.1 to the solution. Through a plant growth screening experiment using ten diverse rice genotypes a significant difference in the response of various genotypes to differential Al toxicity levels on root growth performance indicators was observed. Based on the response, we could able to categorize the genotypes into tolerant and sensitive in relative terms. Thus, the present experiment provided an important nutrient formulation suitable for screening rice genotypes under Al toxicity conditions. Moreover, the selected tolerant and sensitive genotypes can further pave the way for studying the molecular mechanism of Al toxicity response in rice and their use in the breeding program.


Introduction
Aluminum (Al) toxicity is the principal growth-limiting cause for plants in the acid soils (Foy 1992) and in soils with low base saturation, poor calcium (Ca), and magnesium (Mg). In Northeastern India, soils in nearly 80% of the total geographical area (TGA: 26.2 million hectares) are acidic (pH < 6.0) with varying degrees of severity (pH <4.5 to <6.0) (Sharma, Mukhopadhyay, and Sawhney 2006). Acid soil induces fertility stress through imbalances in essential nutrient elements because of the cumulative effects of mineral toxicities (Al and Mn) and deficiencies (P, Ca, Mg and Mo) most importantly, due to the Al toxicity in the uplands. When soil gets acidic, the silicon is leached out from the soil profile leaving Al in the solid forms of aluminum oxyhydroxides, such as boehmite and gibbsite (Abebe 2007). These solid forms release the phytotoxic Al species, Al 3þ also known as Al(H 2 O) 6 3þ into the soil solution (Abebe 2007;Miyasaka, Hue, and Dunn 2007). The trivalent Al 3þ is the most dominant species in the soil solution when the soil becomes strongly acidic (pH < 5.0) and along with monomeric Al-hydroxyls (AlOH 2 þ and Al(OH) 2 þ ), they become the most phytotoxic Al species (Miyasaka et al. 1991, Miyasaka, Hue, andDunn 2007). The root apex is the most sensitive area of the plant to Al, and one of the distinguished symptoms of Al toxicity is the inhibition of root elongation since the root apex is the site for cell division and growth. The most usual and immediate outcome of Al 3þ toxicity within a few minutes to a few hours after exposure to micro-molar concentrations of Al in plants is the inhibition of root development (Barcelo and Poschenrieder 2002). Primary and lateral root apexes could show root growth inhibition, and such roots turn out to be thick and develop brown coloration (Vitorello, Capaldi, and Stefanuto 2005;Wang et al. 2006). Protracted Al toxicity in plants causes reduced fine branching and inhibition in root hair development, which ultimately results in a loss of root biomass. Such long-term exposure can cause uneven and radial expansion of the cortex cells, leading to abnormal root morphology such as distinguished cracks in the root apex, and thickening of the roots (Vitorello, Capaldi, and Stefanuto 2005;Miyasaka, Hue, andDunn 2007, Awasthi et al. 2017). The loss of crop yield due to Al-toxicity differs depending upon factors like free Al species in soil, the crop species, and the crop's variety/genotype. The overall effect of Al toxicity is generally expressed in terms of reduction of crop biomass.
Hydroponics is the most commonly adopted platform to access the plant's tolerance under Al toxicity conditions, where root growth indices can be easily and regularly monitored (Piñeros and Kochian 2001; Magalhaes et al. 2004;Sasaki et al. 2004). In hydroponics-based evaluation, the nutrient solution developed by Yoshida (1976) is still a popular choice to assess the mineral nutrition, nutrient stresses like salt, Fe, and Al toxicity, and P deficiencies in rice (Nguyen et al. 2001;Lin et al. 2004;Shimizu et al. 2004;Dufey et al. 2009). Rice (Oryza sativa) can tolerate a significantly high Al under field conditions compared to other cereals crops (Foy 1988). It has also been observed that the Al concentration required for screening rice seedlings is substantially higher (1112-1482 mM) to observe a visible difference in root growth between tolerant and sensitive genotypes as compared to the Al concentrations used to screen other cereal crops like maize (222 mM), sorghum (148 mM), and wheat (100 mM) (Wu et al. 2000;Nguyen et al. 2001Nguyen et al. , 2002Nguyen et al. , 2003. Usually, P and Fe are available in nutrient solutions as PO 4 2and Fe 3þ , and it has been well established that varying P and Fe concentrations can contribute to changes in root growth and architecture (Lynch and Brown 2001;Williamson et al. 2001;L opez-Bucio, Cruz-Ramı rez, and Herrera-Estrella 2003;Ward et al. 2008). High Al concentration along with lower pH in the media could lead to precipitation of Al along with other essential elements like phosphorus (P), sulfur (S), iron (Fe), etc. When Yoshida solution is used to evaluate Al toxicity response in rice the high ionic strength and high concentrations of mineral ions that complex Al in Yoshida's solution, lead to precipitation, and nutrient imbalances. Consequently, the difference in the growth of plants being screened could not be attributed directly to Al toxicity but also to the deficiency of critical nutrients. In light of the above, Famoso et al. (2010) developed and optimized the Magnavaca's solution (Magnavaca, Gardner, and Clark 1987) to screen rice genotypes. Moreover, looking into the high Al requirement to screen rice seedlings, the solution was modified to have a higher concentration of active Al and lower ionic strength along with reproducible quantities of Al, as well as critical components like P, S, and Fe. Famoso et al. (2010) found that when 540 mM AlCl 3 was supplied into the Modified Magnavaca's solution, it resulted in free Al 3þ activity up to 160 mM. However, the presence of ammonium nitrate as an essential ingredient in the modified Magnavaca's solution proposed by Famoso et al. (2010) limits its application in Indian conditions due to the restrictions imposed on the use of ammonium nitrate. As a result, the use of modified Magnavaca's solution in such screening studies could not be explored. Keeping this in view, an ammonium nitrate-free version of Magnavaca's nutrient solutions for Al toxicity screening studies was developed in the present study while ensuring the availability of other nutrients similar to the modified Magnavaca's solution formulated by Famoso et al. (2010). Therefore, the objective of this study was the development and optimization of a suitable nutrient solution for the Al tolerance screening of rice genotypes from North-Eastern India.

Plant material
Ten rice (Oryza sativa) genotypes (Nagina 22, Sahsarang 1, Motodhan, Jaldhubi, Megha Rice 1, Aaha, Sundari, Leispah, Swarna, and Naveen) were utilized for screening in order to determine the effectiveness of nutrient formulation and the differential response of rice genotypes to Al toxicity. Rice seeds were acquired from the Division of Crop Sciences, ICAR RC NEHR, Umiam, Meghalaya. It also includes the genotypes and varieties which already have been screened in previous experiments, such as Naveen, which had been reported tolerant (Awasthi et al. 2017).

Nutrient solutions
The formulation used in the current experiment is based upon the 'modified Magnavaca's solution' (referred to here as MM1) proposed by Famoso et al. (2010). However, due to ammonium nitrate in the MM1, it was impossible to reproduce and use it for screening in India. Looking into its potential applicability in Al-toxicity-related screening in India, replacing ammonium nitrate with other nitrogen-supplying compounds is crucial. Therefore, we maintained the elemental composition and ionic strength the same as suggested by Famoso et al. (2010) while replacing the ammonium nitrate with other commonly available N-sources: NH 4 H 2 PO 4 , NH 4 Cl, and Ca(NO 3 ) 2 .4H 2 O (Table 1). An online tool named GeoChem-EZ (http://www. Plantmineralnutrition.net/software/geochem_ez/index.html) was used to ensure the compatibility of the components in the solution, estimate Al 3þ activities, and create solutions without significantly lowering available phosphate or sulfate. We named the new formulation as Modified Magnavaca's-2 (MM2). For chemical analysis and plant growth experiments, the treatment (þAl) MM2 was supplied with 550 mM AlCl 3 .6H 2 O, and the final pH was adjusted to 4.1 with 1 N NaOH. Elemental profiling of the nutrient solution The availability of essential nutrients (P, Fe, and Al) was determined by elemental profiling of the MM2 solution. 50 mL samples of MM2 each for control (without Al) and treatment (with Al; 550 mM AlCl 3 .6H 2 O) were collected in triplicate and stored for three days at room temperature under dark conditions to allow chemical equilibrium. To determine the proportion of the soluble and precipitated forms of elements, the solution was homogenized, centrifuged at 3,250 g for 15 min, and then used for elemental profiling. The available P was estimated by the Bray and Kurtz method (Jackson, 1958). The soluble Fe and Al contents of the samples were analyzed using atomic absorption spectrophotometer (Thermo Fisher; Model: iCE 3500).

Plant growth conditions
A sufficient number of viable rice seeds were kept in Petri plates with moistened germination paper and allowed to germinate at 26 C to 30 C for 3 to 5 days under dark conditions. After that, the healthy germinated seeds with an equal length of root (2-3 cm) and shoots were transferred into plastic utility trays (15 litres) containing MM2 solution. Specially designed Styrofoam floaters with plastic seedling holders were used for transplanting the seedlings at room temperature (Day 25 C and Night 16 C; Figure 1a). Up to eight uniformly germinated seedlings per genotype were transferred to the MM2 control solution (without AlCl 3 .6H 2 O and pH 4.1), and the same numbers of seedlings to MM2 Al toxic treatment solution (550 mM AlCl 3 .6H 2 O at pH 4.1).

Measurement of growth and biomass
Following seven days of growth, six randomly selected rice seedlings per genotype and treatment were harvested, carefully washed with tap water, and used for further measurement and data collection. Root related parameters such as Total Root Length (TL), Absolute Root Length (RL), and Dry root weight (DRW) were measured. TL denoted the total root system length, whereas RL indicated the end-to-end length of a root from the shoot-root junction to the tip of the longest root. A root scanner (HP Scanjet 8300 series at 600 dpi, 24-bit colour) was used to accurately take the black and white image of the root system of selected individuals of each genotype. Fresh root weight (per plant root weight, mg) was measured using a digital balance. The roots were then kept in a hot air oven for drying at 60 C for 72 hours or until a constant weight was obtained. Measurements for dry root weight (mg) were taken using a digital balance. The RoodReader2D software (https://www.quantitative-plant.org/software/rootreader2d) is used to measure the root length of the seedlings. Moreover, the means of root growth under control and treatment conditions were calculated for each genotype and Relative Root Growth (RRG: Treatment root growth/control root growth) was determined. In this paper, R-TL indicated RRG for TL while R-RL represented RRG for RL. For the analysis of above-ground parameters, Shoot height (SH) was measured from the tips of the longest leaf to the shoot-root junction of the seedling using a scale. While measurements for dry shoot weight (DSW) were performed similarly to the DRW.

Statistical analysis
All variables were in the replication of six, and Graph Pad Prism 7 tools were used to analyze the data (GraphPad Software Inc., California, USA). The bars represent the mean's standard error (n ¼ 6). The data were all depicted as mean ± SEM values of six replicates. With Tukey's multiple comparison tests, the mean values sharing the same character are not significantly distinct

Results and discussions
The development of a novel composition for the assessment of rice genotypes in Altoxic conditions Considering the suitability of the screening platform for the rice genotypes under the Al toxicity condition, it was decided to move ahead with Magnavaca's solution. The newly formulated ammonium nitrate-free nutrient solution (MM2), can act as an alternative to the modified Magnavaca's solution (MM1) formulated by Famoso et al. (2010) while maintaining most of the compositions unaltered. This updated formulation ensures that the quantity of Al needed to invoke substantial levels of impairment of growth in rice seedlings is reduced ( Table 1).
The emphasis was also given to minimize the chemical interactions in the nutrient solution among Al and other mineral species at the high concentrations of Al required for rice. Chemical analysis through atomic absorption spectrophotometer demonstrated that in the treatment solution (MM2-T), 93.92% of Al was in soluble form while the rest (6.08%) of the total Al was precipitated. In comparison, available P was only reduced by 26.35%, and available Fe was reduced by 22.08%, indicating the availability of a sufficient quantity of essential nutrients for rice seedling growth. Preliminary plant growth experiments showed that the optimum total Al concentration needed for Al tolerance screening in rice would be 550 lM AlCl 3 (10 lM higher than MM1), resulting in 160.9 lM Al 3þ activity in the MM2 solution (Table 2). Moreover, the 550 mM of AlCl 3 .6H 2 O in the MM2 gave significant differences between the previously reported tolerant Naveen (Awasthi et al. 2017) and susceptible genotypes like Aaha. Therefore, 550 mM of AlCl 3 .6H 2 O was selected for carrying further analysis. Through GEO CHEM-EZ speciation software, the free Al 3þ activity in the MM2 was predicted as 160.9 mM (4.34 ppm). In the MM2 formulation, the predicted free Al 3þ activity was comparable to that reported in MM1 by Famoso et al. (2010), and the soluble Al 3þ was almost identical to it.
Similarly, the available P was identical to the reference formulation (MM1), but a substantial decrease in Fe in MM2 was observed as compared to MM1. The previous study had already reported that the modified Magnavaca's solution (MM1) lowered P, Fe, and Al precipitation in the treatment solutions for Al (Famoso et al. 2010). The distinct behaviors of genotypes in terms Reference (Famoso et al. 2010) of various growth parameters between the two nutrient solutions (control and treatment) were consistent with the predictions of Geochem-EZ.

Plant growth experiment
A plant growth study was performed to examine the impact of nutrient formulation on root growth under each treatment. Rice seedling root systems are fibrous and can have numerous major, secondary and tertiary roots within a couple of days of planting. As far as rice's root architecture is concerned, there is a significant genetic variation between varieties, ecotypes, and/or subpopulations. Therefore, for preliminary screening using MM2, ten diverse genotypes of rice were taken. After the statistical analysis, a significant reduction in the performance of selected varieties was observed compared to control under all the selected parameters (Supplementary data; S1-S5). Here, the genotypic difference was also taken into consideration. The mean of control (TL-control) of the ten genotypes was recorded as 43.7 cm in the MM2 after seven days of growth in the control solution (S1). Whereas the mean of treated (TLtreated)) of the same ten genotypes was dropped to 19.8 cm when 550 lM AlCl 3 was supplied in the þ Al treatment solutions (S1). The average R-TL of all the genotypes under evaluation was calculated as 0.48 which was further considered a mean line to classify genotypes based on Al toxicity tolerance. A lower value of R-TL represents a higher level of Al toxicity and viceversa. Therefore, a nutrient solution that could give an average R-TL of around 0.5 while screening randomly selected genotypes could create an Al toxic condition well suited to differentiate between tolerant to moderately tolerant genotype or moderately sensitive to highly sensitive genotype. When the genotypes were considered individually, the TL reduction due to Al toxicity was found insignificant in three genotypes Swarna, Nagina22, and Naveen ( Figure 2a). However, based on R-TL, Sahsarang topped the chart with 0.66, followed by Naveen (0.65), Swarna (0.62), Megha Rice 1 (0.62), and Nagina22 (0.61). Based on the R-TL value these five genotypes were considered Al 3þ toxicity tolerant. The least R -TL was recorded for Motodhan (0.23) followed by Aaha (0.26) and Sundari (0.27), which indicated their sensitive response under Al toxic environment. When the RL was measured, the result was non-concomitant with the values of TL observed. In contrast to TL, the RL showed a significant reduction in both Nagina22 and Swarna. At the same time, the reduction of RL in Naveen (R-RL: 0.80) and Sahsarang (R-RL: 0.73) was non-significant. The DRW was another crucial physiological parameter that showed a different response than TL and RL. Except for Aaha and Leispah, none of the genotypes used showed any significant reduction in the DRW (Figure 2b). However, Swarna showed a 13.3% increase in RDW (R-RDW: 1.09) under treatment conditions. When above-ground parameters were assessed, SH showed a significant reduction in only Sundari (R-SH: 0.65) and Leispah (RSG: 0.76). Whereas the Swarna variety again showed a 12.73% increase in R-SH. Based on R-SH, the performance of Swarna was followed by Megha Rice 1 (R-SH: 0.96), Naveen (R-SH: 0.88), and Nagina 22 (R-SH: 0.87). Another parameter, DSW, does not show any significant change under treatment conditions except for Sahsarang (R-DSW: 0.73) and Sundari (R-DSW: 0.70) where a reduction in DSW was seen.
The upland rice variety, Nagina 22, is a widely recognized aus genotype typically used in drought and heat stress research. (Jagadish et al. 2010;Rang et al. 2011;Prakash et al. 2016;Balyan et al. 2017). Therefore, the comparatively better response of Nagina 22 under root inhibiting conditions might be derived from the drought-adapting nature of Nagina 22. However, being a drought-sensitive variety, the relatively Al tolerance performance of Swarna was unanticipated and needed more studies to reveal the molecular basis of this performance.

Importance of quantifying the whole root system in Al tolerance studies
Physiological mechanisms for Al tolerance and P acquisition have been intensively studied. It is well known that under acidic soil in uplands, P deficiency and Al toxicity go side by side. Interestingly, it has been observed that P deficit enhances rice (Oryza sativa) Al tolerance through altering the physicochemical properties of root plasma membranes and cell walls (Maejima et al. 2014). In the present experiment, Al toxicity's effect was non-significant on Nagina 22 in terms of TL, SH, DRW, and DSW. Although there was a significant reduction in RL, the finding of Famoso et al. (2010) indicates that the RL could not serve as the best indicator of rice tolerance for Al. Previous studies (Famoso et al. 2011;Awasthi et al. 2017) used the length of the total root system (here, TL) as an ideal parameter for assessing tolerance response. Therefore, considering the lack of consistency among the rest of the parameters, we infer that the complete root system's R-TL and TL are the much greater objective measures of Al tolerance for rice. Based on TL reduction and R-TL, five genotypes Sahsarang, Swarna, Naveen, Nagina 22, and Megha Rice 1 were found to be relatively tolerant to Al toxicity, while Motodhan, followed by Aaha and Sundari, showed sensitive responses. Additionally, other parameters like shoot height and dry root weight can also be used as other supporting parameters for the same cause. Among all the parameters used, TL and DRW showed the highest level of correlation (r¼ þ0.694), while TL and DSW showed the lowest level of correlation (r¼ þ0.438). Interestingly, despite a significant reduction in TL in most of the genotypes screened, the DRW was reduced significantly only in the case of Aaha and Leispah. Here, it is important to consider that even though R-TL was 0.26 Figure 2. Bar chart representing non-significant or significant differences under control (without Aluminum) and Treated (with Aluminum) in terms of (a) Total root length, and (b) Dry root weight of ten different rice genotypes used for the study. V1: Nagina22, V2: Sahsarang1, V3: Motodhan, V4: Jaldhubi, V5: Megha rice1, V6: Aaha, V7: Sundari, V8: Leispah, V9: Swarna, and V10: Naveen. and 0.48, the RDRW was 0.59 and 0.70 for Aaha and Leispah, respectively. Even under high Al toxicity, the lesser change in DRW may be contributed by increased callusing following the tissue damage. In many previous studies, callose formation in root and lignin deposition in cortical cells of roots which ultimately causes a change in dry root weight has been reported as an early marker of Al toxicity in crops like maize and wheat (Horst, P€ uschel, and Schmohl 1997;Eticha, Stass, and Horst 2005;Miyasaka, Hue, and Dunn 2007). In the present study, under the treatment condition, an increase in root branch number, and root curling was observed while the intensity of tertiary roots decreased as compared to control (Figure 1b). Root curling is another manifestation of root damage due to Al toxicity. In the previous reports, Al toxicity has been found to cause severe inhibition in the growth of lateral roots primarily in the sensitive genotypes across the different crops like maize, wheat, sorghum, soybean, sugarcane, and tobacco. (Hetherington, Asher, and Blamey 1988;Bushamuka and Zobel 1998;Silva et al. 2001;Brichkova et al. 2007). The Naveen variety has already been proposed as a tolerant variety by previous work by Awasthi et al. (2017). Henceforth, the tolerance performance of Naveen under our experimental setup indicated two crucial things. Firstly, it establishes the utility of the new Magnavaca's formulation for Al toxicity screening, and secondly, it gives additional experimental evidence about the tolerant nature of Naveen.

Summary and conclusions
In the present work, the modified Magnavaca's solution (MM2) was found to confirm its suitability for screening rice genotypes for Al toxicity tolerance-related studies. Some rice genotypes were more vigorous at Al toxicity exposure, while a few had an incredibly sensitive response. Total root length and fresh and dry weight of rice were reduced dramatically in sensitive genotypes. Our analysis disclosed that among the different genotypes used in the current study Sahsarang, Nagina 22, Swarna, Megha Rice1, and Naveen are comparatively tolerant varieties based on different criteria. These genotypes can further be utilized in the Al toxicity tolerance research and breeding program in the North-Eastern Indian region and examine the improved tolerance mechanisms for prospective use. Molecular understanding of tolerance and sensitive behavior of the identified genotypes could lead to the development of genetically modified varieties suitable to grow under strong soil acidity and Al toxicity conditions.