Different effects of gellan gum and agar on change in root elongation in Arabidopsis thaliana by polyploidization: the key role of aluminum

Agar and gellan gum have been considered to have different effects on polyploidy-dependent growth in plants. We aim to demonstrate that agar and gellan gum differently affect the change in root elongation in Arabidopsis thaliana by polyploidization and examined the physico-chemical parameters in each gelling agent to elucidate key factors that caused the differences. Each polyploid strain was cultured vertically on agar and gellan gum solidified medium under fixed conditions. Root elongation rate was measured during 4–10 days after sowing. As a result, agar promoted root elongation of polyploids more than the gellan gum. Then water potential, gel hardness, and trace elements of each medium were quantified in each medium. Water potential and gel hardness of agar medium were significantly higher than those of gellan gum medium. The decrease in water potential and gel hardness in agar medium, however, did not affect the change in polyploidy-dependent growth. Elemental analysis showed that gellan gum contained more aluminum than agar. Subsequently, the polyploids were grown on agar media with additional aluminum, on which the root elongation in tetraploids and octoploids was significantly suppressed. These results revealed that agar and gellan gum affect the change in growth of root elongation in A. thaliana by polyploidization in different ways and the different effects on change in polyploidy-dependent growth is partially caused by aluminum in the gellan gum, which may be due to cell wall composition of polyploids.


Introduction
Gelling agents are among the key ingredients that determine the characteristics of plant growth media (Debergh 1983;Gruber et al. 2013;Kumar and Reddy 2011). Agar and gellan gum are the most popular gelling agents used in plant growth media, and several studies have shown that their effects on plant growth are different.
Agar is a complex polysaccharide derived from red algae (Usov 2011) and has traditionally been used in plant growth media. Gellan gum is a bacterial polysaccharide produced using Sphingomonas spp. (Jansson et al. 1983), which forms transparent gels in the presence of multivalent cations (Kirchmajer et al. 2014;Morris et al. 2012). Agar media are more suitable than gellan gum media for plant growth; agar improved the growth characteristics of Swartzia madagascariensis (Berger and Schaffner 1995), lettuce (Ichimura and Oda 1998), and Stevia (Fatima and Khan 2011), and promoted the induction of shoots in certain apple varieties (Mitić et al. 2012). However, there have also been reports of gellan gum being a more suitable ingredient for plant growth than agar, as in the case of seedling growth in cucumber (Cucumis sativus) (Ichi et al. 1986), shoot regeneration in Eucomis autumnalis (Masondo et al. 2015), shoot growth in several Australian tree species (Williams and Taji 1987), and micropropagation in Scrophularia yoshimurae (Tsay et al. 2006). The above reports show that the choice of a gelling agent depends upon the effect it has on growth of plant species and/or the tissue used for analysis. Both gelling agents have been widely used for growth analysis of the model plant Arabidopsis thaliana (L.) Heynh.
Chromosome polyploidization is very common in plants (angiosperms) and plays a significant role in their speciation (Leitch and Bennett 1997;Soltis et al. 2003). Polyploidization affects various aspects of plant growth. In general, the cell volume of plants increases with polyploidy, resulting in an increase in organ size (Corneillie et al. 2019;Levin 1983;Robinson et al. 2018). Tsukaya (2008) reported that the hexaploid and octoploid Arabidopsis spp. showed suppressed growth when compared to the diploid strain. Growth suppression due to high ploidy (i.e., hexaploidy and octoploidy) is termed "high-polyploidy syndrome" (Tsukaya 2008). In previous studies, we examined the effects of polyploidization on plant growth using a synthetic autopolyploid series of A. thaliana (Iwamoto et al. 2006(Iwamoto et al. , 2013Kikuchi and Iwamoto 2020). We performed a kinematic analysis of the cellular basis of root growth in diploid and autotetraploid A. thaliana and showed that polyploidization accelerated root growth in tetraploids due to increase in volume growth rate (Iwamoto et al. 2006(Iwamoto et al. , 2013. We also performed fluorescence in situ hybridization (FISH) analysis with a centromeric DNA probe on the leaf cells of diploid and synthetic autopolyploid series of A. thaliana to observe and quantify chromosome polytenization in polyploid plants (Kikuchi and Iwamoto 2020). We observed that the degree of chromosome polytenization increased with an increase in the degree of ploidy, suggesting that the suppression of cell proliferation in the root of polyploid plants may be related to chromosome polytenization.
Our preliminary observations in these previous studies revealed that the root length in 10-day-old tetraploid seedlings was longer than that of diploids on agar media, while the root length of tetraploids is shorter than that of diploids, and those of hexaploids and octoploids were more severely suppressed than those of the diploids on gellan gum medium (Fig. 1). These results clearly indicate that the gelling agent influences the change in growth caused by polyploidization.
The effects of agar and gellan gum on the root growth of polyploids has been attributed to two possible factors: differences with respect to physical properties, and differences in trace elements. The physical attributes of gelling agents that affect plant growth are gel hardness (Jacques et al. 2020), water potential, and water availability of media (Buah et al. 1999;Klimaszewska et al. 2000). Further, trace elements present in gelling agents also affect plant growth (Al-Mayahi and Ali 2021; Ichi et al. 1986;Ichimura and Oda 1998).
The objective of the present study was to determine the mechanisms by which the difference in the gelling agent affected change in growth caused by polyploidization. We analyzed the effects of the two gelling agents, agar and gellan gum, on root elongation in the polyploid series of A. thaliana. Identifying the differences between the two gelling Fig. 1 Different effects of gellan gum and agar on root growth of polyploid series in Arabidopsis thaliana. Ten-day-old seedlings grown vertically on MS medium solidified by gellan gum or agar. Scale bar = 2 cm agents in terms of their physical properties as well as trace element content further helped us understand the effects of gelling agents on plant growth under various physical conditions. In particular, we aimed to study the effects of Al as an important trace element in agar media to elucidate its effect on root growth of polyploids.

Production of synthetic autopolyploid series of A. thaliana
Synthetic autopolyploids of A. thaliana (tetraploid, hexaploid, and octoploid) were produced using colchicine treatment of diploid seedlings (Kikuchi and Iwamoto 2021). The ploidy level was determined by flow cytometry analysis. Polyploid strains for which the ploidy level was confirmed for at least three generations after colchicine treatment were used in the experiment.

Flow cytometry analysis
Leaf samples for flow cytometry analysis were collected from seedlings 30-40 days after sowing (DAS) to produce polyploids and 10-15 DAS for the measurement of root elongation rate. Ploidy of each strain was confirmed with a flow cytometer (CyFlow Ploidy Analyzer PA, Münster, Germany) using the chopping method (Johnston et al. 1999).
The agar media used to assess the effect of Al on root growth was prepared by adjusting the pH to 5.5 with succinic acid, which was equal to the pH of the 0.8% (w/v) gellan gum media (the pH of 1.5% (w/v) agar media used for all other analyses was adjusted to 6.0).

Measurement of root elongation rate
Starting from 4 DAS, the level reached by the growing root tips was marked on the plate every day to record the position of the root tip of each seedling. This procedure was repeated until 10 DAS, at which point the plate was scanned and its digital image was saved. The distances between successive marks along the roots were determined from digitized images using ImageJ ver. 1.51m9 (NIH Image, Bethesda, MD, USA). Average root elongation rate was calculated each day as the measured distance of root tip movement divided by the corresponding time interval between markings (approximately 24 h).

Water potential
The water potential of each medium was measured using a dewpoint potentiometer WP4C, Decagon Devices, Inc., Pullman, WA, USA). After sterilization of the media, 6 mL of each medium was dispensed into steel WP4C cups and allowed to solidify into agar and gellan gum at room temperature. These cups were used as the samples for water potential measurement. Each sample had six replicates.

Gel hardness
The hardness of each media was measured by rupture strength analysis using a creep meter (RE2-33005B, YAMA-DEN, Tokyo, Japan). After sterilization of the media, 6 mL of each medium was dispensed into petri dishes and allowed to solidify at room temperature to be used as samples for gel hardness measurements. Each sample had six replicates.

Elemental analysis
Agar (approximately 60 mg) and gellan gum samples (approximately 300 mg) were dried at 70 °C and digested using concentrated HNO 3 . Elemental analysis was performed using inductively coupled plasma optical emission spectroscopy (ICP-OES) (SPS 3500, SII NanoTechnology, Tokyo, Japan). Sample measurements were taken using five replicates each of agar and gellan gum. The agar samples were diluted 10-fold for measurement of Na, and the gellan gum samples were diluted 40-fold for measurement of Ca, Mg, and Na.

Statistical analysis
The Bartlett test was conducted with the measurement datasets of root elongation rate in polyploid series on each date (Figs. 2, 4b and 5). The analysis did not indicate homogeneity of variance in some measurement datasets, therefore we decided to conduct the Kruskal-Wallis test with all measurement datasets. The analysis showed significance in all measurement datasets (p < 0.01) and then the Steel-Dwass test was conducted.
The Bartlett test was also conducted with the measurement dataset of gel hardness in 0.8% gellan gum, 1.0% agar and 1.5% agar ( Table 1). The analysis did not indicate homogeneity of variance in this dataset, therefore we decided to conduct the Kruskal-Wallis test with these datasets. The analysis showed significance in this dataset (p < 0.01) and then the Steel-Dwass test was conducted.
The Welch's t-test was conducted with the measurement data of root elongation rate in 1.5% agar medium with 0 µM and 30 gellan gum of polyploid series on each date (Fig. 6), the measurement data of water potential in the 0.8% gellan gum and 1.5% agar media, the measurement data of water potential in the 1.5% agar with 0 mM mannitol and 1.5% agar with 20 mM mannitol (Table 2), and the concentration of each element in agar and gellan gum media (Table 3).
The Bartlett test and the Steel-Dwass test were conducted using the software R version 3.4.0 (R Development Core Team). The Kruskal-Wallis test and the Welch's t-test were conducted using the software KaleidaGraph version 5.0 (Synergy Software).

Root elongation of seedlings of polyploid series grown on gellan gum and agar media
We observed the seedlings of polyploid series of A. thaliana and measured their root elongation rates on both gellan gum and agar media, which aims to characterize the effect of difference in gelling agent on polyploidy-dependent growth. A comparison of the polyploid series of A. thaliana grown vertically on 0.8% gellan gum and 1.5% agar media at 10 DAS showed that gellan gum promoted the root growth of diploids whereas agar promoted the growth of polyploids (tetraploid, hexaploid, and octoploid) (Fig. 1).
We measured the root elongation rates of polyploid series grown on 0.8% gellan gum and 1.5% agar media from 4 to 10 DAS (Fig. 2a, b). When grown on 0.8% gellan gum media, we observed a gradual reduction in root length of seedlings with increasing ploidy levels; root elongation rate of diploids was the highest (Fig. 2a).
Contrasting results were obtained in the case of agar (Fig. 2b). The root elongation rates of polyploids (tetraploid, hexaploid, and octoploid) grown on 1.5% agar media were higher than those for polyploids grown on 0.8% gellan gum media, while the root elongation rate of diploids was suppressed. Consequently, the root elongation rate of diploids was lower than that of tetraploids when grown on 1.5% agar media.

The assessment of gel hardness of agar and gellan gum
We measured the gel hardness of several media changing the type and concentration of gelling agents. Then, we measured the root elongation rate of polyploid series grown on the agar medium with gel hardness adjusted to be softer than that of the 0.8% gellan gum medium, which aims to determine the effects of gel hardness on the root elongation of polyploid series.
We measured the gel hardness of both gellan gum and agar media at several concentrations (Fig. 3a). The gel hardness of both media increased proportionally with the concentration of the gelling agent. When compared at the same  concentration of the gelling agent, the gellan gum media was always harder than the agar media. The gel hardness of 0.8% gellan gum, 1.0% agar, and 1.5% agar media (Fig. 3a) are listed in Table 1. The gel hardness of 1.5% agar medium was significantly higher than that of 0.8% gellan gum medium, while that of 1.0% agar medium was significantly lower than that of 0.8% gellan gum medium (Table 1).
Subsequently, we measured the root elongation rate of polyploids grown on 1.0% (w/v) agar medium (Fig. 2c). The root elongation rates of diploid, tetraploid, and hexaploid grown on the softer (1.0%) agar medium at all measurement dates were more than those grown on the harder (1.5%) agar media, while the root elongation rate of octoploid was almost the same on both media (Fig. 2b, c).

The assessment of water potential of agar and gellan gum
We measured the water potentials of several media changing the type and concentration of gelling agents. Then, we measured the root elongation rate of the polyploid series grown on the agar medium with water potential adjusted to nearly the same water potential as in the 0.8% gellan gum medium, which aims to determine the effects of water potential on the root elongation of polyploid series.
We measured the water potentials of gellan gum and agar media at several concentrations (Fig. 3b). The lower the water potential, the lesser the water transferred from a medium to the roots. The water potential of the agar was almost constant at all concentrations, while that of the gellan gum tended to decrease at high concentrations. The water potential of the agar was higher than that of the gellan gum at all concentrations.
We added mannitol to the 1.5% agar medium in several concentrations to reduce the water potential (Fig. 4a). The water potential of the medium was almost constant at lower concentrations of mannitol (0, 5, 10 mM), but it decreased proportionally with the concentration of mannitol at higher concentrations (20, 50, 100 mM).
The water potentials of 0.8% gellan gum and 1.5% agar media (Fig. 3b), and those of 1.5% agar medium with 0 mM and 20 mM mannitol (Fig. 4a) are listed in Table 2. The addition of 20 mM mannitol to 1.5% agar medium significantly decreased water potential. Water potential of agar medium with 20 mM mannitol was almost the same as that of 0.8% gellan gum. Subsequently, we measured the root elongation rate of the polyploid series grown on agar media with 0 mM and 20 mM mannitol from 4 to 10 DAS (Fig. 4b).
The temporal profile of elongation rate on the agar medium with 0 mM mannitol was similar to that on the agar medium with 20 mM mannitol.

The assessment of trace elements of agar and gellan gum
We determined the amount of trace elements in each gelling agent to analyze their effects on the root elongation There was a significant difference in the root elongation rate between every pair grown on agar media for all days from 4-5 DAS to 9-10 DAS (Steel-Dwass test, p < 0.01), except between diploid and tetraploid pairs for 5-6 DAS, 7-8 DAS, and 9-10 DAS in 0 mM mannitol medium and 8-9 DAS and 9-10 DAS in 20 mM mannitol medium (N.S., Steel-Dwass test, 0.05 < p), diploid and tetraploid pairs for 6-7 DAS in 0 mM mannitol medium and 5-6 DAS to 7-8 DAS in 20 mM mannitol medium (*, Steel-Dwass test, 0.01 < p < 0.05), and hexaploid and octoploid pairs for 6-7 DAS and 8-9 DAS in 0 mM mannitol medium and 5-6 DAS in 20 mM mannitol medium (*, Steel-Dwass test, 0.01 < p < 0.05). Bars indicate standard errors of polyploid series of A. thaliana. Then we determined a candidate element responsible for the change in polyploidydependent growth and examined its effects on the root elongation of polyploid series. We quantified the concentrations of trace elements in 0.8% gellan gum media and 1.5% agar media using ICP-OES (Table 3). Elemental analysis showed that gellan gum contained significantly more magnesium (Mg), aluminum (Al), and calcium (Ca) than agar, whereas agar contained significantly more sodium (Na), iron (Fe), and cadmium (Cd) than gellan gum (Welch's t-test, p < 0.05). No significant differences were found in boron (B), titanium (Ti), and manganese (Mn) contents between the gelling agents. Chromium (Cr) and copper (Cu) were not detected in either gelling agent.
As Al has been widely considered to suppress the root elongation in A. thaliana (Koyama et al. 1994;Rounds and Larsen 2008;Sjogren et al. 2015;Sun et al. 2010;Yang et al. 2014), we focused on Al among all trace elements and added it to the agar medium and examined whether Al inhibits root elongation in polyploids. We measured the root elongation rate of the polyploid series grown on 1.5% agar media supplemented with the same concentration of Al (30 µM) in the 0.8% gellan gum media (Fig. 5). The root elongation rate of diploids did not change significantly between the 1.5% agar medium without additional Al and that with 30 µM Al (Fig. 6a), while the root elongation rate of tetraploids significantly decreased from 4 to 5 DAS to 7-8 DAS (Fig. 6b). The root elongation rate of octoploid significantly decreased from 6 to 7 DAS to 8-9 DAS (Fig. 6d), whereas the root elongation rate of hexaploids did not decrease but significantly increased at 6-7 DAS (Fig. 6c).

Gellan gum and agar differently affect the change in root elongation by polyploidization
The measurement of root elongation rate in polyploid series grown on gellan gum and agar media (Fig. 2) confirmed our preliminary observation that the change in growth due to polyploidization differs between the two gelling agents, as the components of each medium were identical except for the gelling agent. This is the first study to demonstrate that the type of gelling agent influences the polyploidization effect on plant growth. Nakagawa et al. (2007) revealed that the primary root of the wild type of A. thaliana could penetrate the harder layer of double-layer agar medium, whereas mca1-null mutant could not. As MCA1 is considered to be a Ca 2+ -permeable stretch-activated (SA) channel (Yoshimura et al. 2021

Fig. 5
Effect of aluminum on root elongation rate of polyploid series on agar media. a Cultured on 1.5% agar medium with 0 µM Al. n = 74 (diploid), 70 (tetraploid), 23 (hexaploid), 20 (octoploid). b Cultured on 1.5% agar medium with 30 µM Al medium. n = 74 (diploid), 54 (tetraploid), 27 (hexaploid), 16 (octoploid). There was a significant difference in the root elongation rate between every pair grown on agar media for all days from 4-5 DAS to 9-10 DAS (Steel-Dwass test, p < 0.01), except between diploid and tetraploid pairs for 5-6 DAS to 7-8 DAS (N.S., Steel-Dwass test, 0.05 < p) and 8-9 DAS (*, Steel-Dwass test, 0.01 < p < 0.05) in b, and tetraploid and hexaploid pairs for 4-5 DAS (N.S., Steel-Dwass test, 0.05 < p) in a and b. Bars indicate standard errors result of the experiment using double-layer agar medium suggests that the Ca 2+ -permeable SA channel could detect the gel hardness and affect the root elongation of A. thaliana. Therefore, we expected that each polyploid might detect the gel hardness of the gellan gum and agar media differently, leading to differences in change in growth due to polyploidization between the two gelling agents (Figs. 1 and 2). However, the results suggest that the gel hardness was not a critical factor in determining the difference of change in growth by polyploidization (Fig. 2). The root elongation rate in diploids, tetraploids, and hexaploids was promoted when they were grown on the softer agar medium (1.0% agar) with a gel hardness similar to that of the 0.8% gellan gum medium; however, the relationship among the temporal profiles of root elongation rate in polyploids was mostly unchanged compared with those grown on the normal agar medium (1.5% agar) (Fig. 2b, c). Schultz et al. (2016) analyzed the root growth grown on three kinds of gelling agents and revealed that differences in media type had more of an impact on root growth than hardness itself, which is consistent with the results of this study.
The control of water potential in media is related to the control of water availability, and the water availability is directly related to the plant growth. In the roots of A. thaliana, the lower the water potential of the medium was, the lower the elongation rate was (van der Weele et al. 2000). The results suggest that the water potential affects the root elongation rate and each polyploid may detect the water availability differently, leading to differences in change in growth due to polyploidization between gellan gum and agar. However, the results of this study suggest that the water potential is not a critical factor in determining the difference in change in growth due to polyploidization (Figs. 2 and 4). The temporal profiles of root elongation rate in polyploids in the normal 1.5% agar medium (0 mM mannitol) and in the 1.5% agar medium with the lower water potential (20 mM mannitol) are similar, which did not correspond to that of the 0.8% gellan gum medium. In a previous study, a difference of water potential between each pair of growth conditions was at least more than 0.1 MPa (van der Weele et al. 2000); however, the difference of water potential between the 1.5% agar medium + 0 mM mannitol and the 1.5% agar medium + 20 mM mannitol was 0.05 MPa in this study (Table 2), which may be too low to affect the change in growth due to polyploidization.

Aluminum in the gellan gum could partially explain the effects of agar and gellan gum on the root growth of polyploids
Quantitative analysis of trace elements in gellan gum and agar showed that the contents of certain elements differed significantly (Table 3). Agar contained significantly higher Na, Fe, and Cd than gellan gum. In contrast, gellan gum contains significantly higher Mg, Al, and Ca. The root elongation of A. thaliana was usually suppressed by more than 10 mM NaCl in previous studies (Fu et al. 2019;Jiang et al. 2016;West et al. 2004;Zhao et al. 2017), and the contents of Na in gellan gum and agar were significantly lower than 10 mM in this study (ca. 7.8 mM in gellan gum and 1.8 mM, Table 3), which could not affect the root elongation because they are too low. The deficiency of Fe, Mg, and Ca in media severely suppressed the root elongation of A. thaliana (Gruber et al. 2013), but the 1/2 MS medium, which is the base medium used in this study, contains a large amount of Fe, Mg, and Ca (1500 µM Ca, 750 µM Mg and 45 µM Fe, Murashige and Skoog 1962). Therefore, the contents of Fe, Mg, and Ca in gellan gum and agar could not affect the root elongation (Table 3). Cd is well known to suppress root elongation in plants (Godbold and Hüttermann 1985;Munzuroglu and Geckil 2002). However, less than 5 µM Cd did not severely suppressed the root elongation of A. thaliana (Van Belleghem et al. 2007;Wójcik and Tukiendorf 2004), which suggests that the contents of Cd in gellan gum and agar were too low to affect the root elongation (0.320 µM in agar and 0.028 µM in gellan gum, Table 3).
On the other hand, Al significantly suppressed the root elongation of A. thaliana at 20 µM (Sun et al. 2010), 50 µM (Zhu et al. 2012), 6 µM (Yang et al. 2014). The content of Al in gellan gum was 33.401 µM (Table 3), which is within the range of content that could suppress the root elongation of A. thaliana. Therefore, we focused on the effect of Al in gellan gum and conducted Al addition experiments in agar media for polyploids.
The results of Al addition experiments showed that the root elongation rate of tetraploids and octoploids grown on the 1.5% agar medium with 30 µM Al was significantly suppressed, while that of diploids remained unchanged (Fig. 6a, b, d). The root elongations of diploids and tetraploids were almost the same at 5-6, 6-7, and 7-8 DAS and the differences between the root elongations of diploids and tetraploids were relatively small at other measurement dates when grown on the 1.5% agar medium with 30 µM Al, whereas the root elongation rate of tetraploids was significantly higher than that of diploids at all measurement dates when grown on the 1.5% agar medium with no additional Al (0 µM Al) (Fig. 5a, b). In addition, the root elongation of octoploid grown on the 1.5% agar medium with 30 µM Al significantly decreased. These results suggest that the higher Al concentration in gellan gum could partially explain the differences in change in growth due to polyploidization between 0.8% gellan gum and 1.5% agar medium. As the root elongation rate of diploids was significantly higher than that of tetraploids grown on the 0.8% gellan gum medium, the addition of Al to the agar medium did not completely reproduce the relationship of temporal profiles in root elongation between diploids and tetraploids grown on the 0.8% gellan gum medium. The relationship between diploids and tetraploids grown on the agar medium with 30 µM Al, however, is closer to that of those grown on the 0.8% gellan gum medium than that of those grown on the agar medium with 0 µM Al, which suggests that the decrease in root elongation of tetraploids in the 0.8% gellan gum medium should be partially attributed to the Al in gellan gum (Figs. 2 and 5).
Al binds to cell wall components, especially hemicellulose, thereby reducing cell wall extensibility and inhibiting root elongation (Yang et al. 2011). Polyploids of A. thaliana have been shown to have an increased amount of hemicellulose compared with diploid plants (Corneillie et al. 2019), which may have contributed to the tetraploids and octoploids being more sensitive to Al than diploid in root elongation rate.
The root elongation rate of hexaploids grown on the 1.5% agar medium with 30 µM Al was similar to that of those grown on the 1.5% agar medium with 0 µM Al. However, further studies are needed to understand as to why the addition of Al did not have any effect on the root elongation of hexaploids, while it suppressed those of tetraploids and octoploids.

Conclusion
This study is the first report of the different effects of two major gelling agents, gellan gum and agar, on change in growth due to polyploidization. Our findings indicate that Al, which is more abundant in gellan gum, suppresses root elongation in tetraploids and octoploids, which could be partly attributed to the different effects of gellan gum and agar on the change in growth due to polyploidization. However, the physical properties of gellan gum and agar media, gel hardness, and water potential, may not be responsible for these effects.