The Pyla-1 Natural Accession of Arabidopsis thaliana Shows Little Nitrate-Induced Plasticity of Root Development

: Optimizing root system architecture is a strategy for coping with soil fertility, such as low nitrogen input. An ample number of Arabidopsis thaliana natural accessions have set the foundation for studies on mechanisms that regulate root morphology. This report compares the Columbia-0 (Col-0) reference and Pyla-1 (Pyl-1) from a coastal zone in France, known for having the tallest sand dune in Europe. Seedlings were grown on vertical agar plates with different nitrate concentrations. The lateral root outgrowth of Col-0 was stimulated under mild depletion and repressed under nitrate enrichment. The Pyl-1 produced a long primary root and any or very few visible lateral roots across the nitrate supplies. This could reﬂect an adaptation to sandy soil conditions, where the primary root grows downwards to the lower strata to take up water and mobile soil resources without elongating the lateral roots. Microscopic observations revealed similar densities of lateral root primordia in both accessions. The Pyl-1 maintained the ability to initiate lateral root primordia. However, the post-initiation events seemed to be critical in modulating the lateral-root-less phenotype. In Pyl-1, the emergence of primordia through the primary root tissues was slowed, and newly formed lateral roots stayed stunted. In brief, Pyl-1 is a fascinating genotype for studying the nutritional inﬂuences on lateral root development.


Introduction
Nitrogen (N) mineral fertilization is used to sustain crop production, but excessive concentrations of nitrate have harmful effects on the environment and human health by causing soil nitrate leaching, groundwater pollution and greenhouse gas emissions [1,2]. Improving decision tools for agricultural management and breeding crops for better nitrogen use efficiency (NUE) are implemented to reduce N fertilizer input [3][4][5]. The optimization of the root system architecture of crops for a more efficient N capture is a key determinant to reaching that goal. The root organ has plasticity to adapt to the nitrate supply: local nitrate patches stimulate lateral root outgrowth, whereas globally high external nitrate concentrations have a systemic inhibitory effect [6,7]. Increasing lateral root branching of a crop species would serve to explore a larger soil volume and to enhance nitrate capture [8][9][10].
In plant biology research, Arabidopsis thaliana has a long history of being used as a model for studying root organogenesis. The formation of a lateral root is a post-embryonic event that originates from the asymmetric division of the pericycle cell pair-a step referred to as the lateral root initiation [11][12][13]. The dividing pericycle cells develop into a lateral root primordium that emerges through the overlying tissues (subsequently, the endodermis, cortex and epidermis) of the parent root [14][15][16]. Next, the apical meristem is activated, and the newly formed lateral root elongates. Genetic screens have isolated a plethora of Arabidopsis mutants with lateral root developmental defects [17]. Additionally, the Arabidopsis model represents a step forward in understanding the nitrate influences on root growth [18]. The variability of root morphologies, offered by abundant natural accessions, is exploited to mine for the genes and alleles that regulate root traits in response to the nitrate supply [19]. This knowledge may provide new targets to breed genetically related crops with larger genomes (e.g., cultivated Brassica), which could be more efficient at capturing N in the soil.
This study provides a detailed root phenotype characterization of Pyla-1 (Pyl-1), one natural Arabidopsis accession that is part of a core collection that maximizes the genetic diversity of the species [20]. The Pyl-1 originates from the coastal zone of Arcachon Bay in France, known for having the tallest sand dune in Europe. Following previous phenotyping screens conducted in vitro [21,22], the accession shows any or very few visible lateral roots at a young developmental stage. This report provides (i) a comparison of a two-dimensional root morphology between the Columbia-0 (Col-0) reference and Pyl-1 in response to a nitrate supply, and together with four accessions collected from the Arcachon urban area, (ii) a quantification of the total number and developmental stages of lateral root primordia, and (iii) a temporal sequence of the lateral root formation after a gravitropic curvature. These findings underline Pyl-1 as a prevailing genotype for studying the mechanisms of repression exerted by nitrate on lateral root growth.

Plant Material
The Arabidopsis thaliana Columbia-0 (Col-0) accession was obtained from the Nottingham Arabidopsis Stock Centre, and Pyla-1 (Pyl-1) was obtained from the INRAE Versailles Genomic Resource Center. The four Arcachon (Arc-1, Arc-2, Arc-3, Arc-4) accessions were collected in the Arcachon-Pyla area, France. The geographical map of the harvesting sites and coordinates is shown in Figure S1.

In Vitro Culture
The in vitro culture procedure is described in [22]. Briefly, sterilized seeds were plated on a 1× Murashige and Skoog medium, modified with nitrate as the only N source, 1% (w/v) sucrose and 0.8% (w/v) plant agar (Duschefa Biochemie, Haarlem, the Netherlands). The pH was adjusted to 5.7. Five media with different nitrate concentrations were prepared: 0.1 mM KNO 3 + 9.9 mM KCl (added to prevent potassium depletion), 1 mM KNO 3 + 9 mM KCl, 10 mM KNO 3 , 25 mM KNO 3 or 50 mM KNO 3 . A volume of 50 mL of medium was poured into one square Petri plate (12 cm × 12 cm). Five seeds of one genotype were sown on a Petri plate and were stratified for two days at 4 • C in the dark. The plates were vertically incubated in a culture chamber with a constant temperature of 20 • C and a photoperiod of 16 h light (45 µmol photons m −2 s −1 ) /8 h darkness.

Two-Dimensional Root Morphology Analysis
Eleven days after germination, the root systems were scanned with an EPSON Scan Perfection V30 at a resolution of 400 dpi. The scans were annotated using the RootNav image analysis software [23]. The length of the primary root (L PR ), the number of lateral roots that visibly emerged from the primary root longer than 1 mm (N LR ) and the sum of the length of lateral roots (ΣL LR ) were quantified [22].

Quantification of Lateral Root Primordium Developmental Stages
The roots were cleared following the protocol detailed in [24]. The developmental stages of lateral root primordia were observed with an Optika B-350 microscope. Nine categories were established: I (initiating primordium), II-III (up to three cell layers structure), IV-VII (primordium dome formation and emergence through parental tissues), E (lateral root emerged from the epidermis and of less than 1 mm) and LR (lateral root strictly longer than 1 mm). The total number of lateral root primordia (N LRP ) and the density of lateral root primordia per primary root length (D LRP ) were measured 11 days after germination.

Lateral Root Bending
The seeds were plated on a medium containing 10 mM KNO 3 . The seedlings were grown for three days under continuous light (45 µmol photons m −2 s −1 ) conditions. The plates positioned vertically were turned 90 • , as described in [25]. The roots were collected at eight different time points (18 h, 24 h, 30 h, 36 h, 42 h, 48 h, 54 h and 60 h) after rotation. The root bends were cleared (see Section 2.4), and the development stages of the lateral root primordia (from initiation to emergence) were scored.

Statistical Treatment
All statistical analyses were done with R software [26] and XLSTAT [27]. An analysis of variance was performed for the variables L PR , N LR , ΣL LR , N LRP and D LRP , using the lme4 R package [28]. The phenotypic models were fitted on the Col-0 and Pyl-1 genotypes using the REML method, according to the equation where P ijk is the phenotypic value, µ is the general mean, G i is the genotype i, N j is the nitrate supply j, and e ijk is the residual. All effects were declared as random. Moreover, multiple t-tests were performed using the tidyverse and rstatix R packages [29,30] to compare the mean values of the root morphological traits under each genotype x nitrate combination.

The Pyla-1 Accession Exhibited Little Macroscopic Variation of Root Morphology in Response to Nitrate Supply
The seedlings of Col-0 and Pyl-1 grew vertically on media supplemented with 0.1, 1, 10, 25 or 50 mM nitrate. The morphological variation of the whole root organs is shown in Figure 1. The length of the primary root (L PR ), the number of visible lateral roots > 1 mm (N LR ) and the sum of the length of lateral roots (ΣL LR ) reached the greatest values in the Col-0 seedlings supplied with 1 mM nitrate ( Figure 2). The modulation of the root morphology by nitrate availability followed the previous observations in the reference accession [31,32]. A low but moderate nitrate supply (1 mM) promoted the growth of both the primary and lateral roots, while a severe nitrate shortage (0.1 mM) and elevated nitrate supply (>25 mM) repressed the number and the length of lateral roots in Col-0 ( Figure 2). By contrast, Pyl-1 was slightly responsive to the nitrate supply and produced very few short lateral roots, solely at 0.1 mM nitrate. For every nitrate concentration tested, the L PR values were significantly (p < 0.05) greater in Pyl-1 than in Col-0, while N LR and ΣL LR were lower ( Figure 2). We assessed the extent to which a variation in the root morphological traits was due to the genetic differences between accessions (G), nitrate treatment differences (N), genetic differences in the level of response to nitrate treatment (G × N) and unexplained experimental errors (residuals) ( Figure S2). Most of the variations for L PR and N LR were due to G (35% and 65%, respectively). The ΣL LR trait showed a nearly equal dependence on G (47%) and G × N (44%). This highlighted a larger genotype-dependent plasticity to the nitrate treatment for ΣL LR than for L PR or N LR .
All four Arcachon accessions showed similar root morphologies at 10 mM nitrate ( Figure 1). The L PR , N LR and ΣL LR values were significantly (p < 0.05) lower compared to Col-0 ( Figure 2). Only L PR was identical between Arc-4 and Col-0. Contrastingly, these accessions did not mimic the Pyl-1 root phenotype; rather, they presented shorter L PR and more important N LR and ΣL LR values than Pyl-1.

The Initiation of Lateral Root Primordium Was Not Impaired in Pyla-1
One possible explanation for the lateral-root-less phenotype harbored by Pyl-1 (Figure 1) could be an impairment of lateral root initiation. Hence, we examined the lateral root development at the microscopic level ( Figure 3). In both genotypes, the number of lateral root primordia (N LRP ) globally decreased with the nitrate concentration increasing. The N LRP was significantly (p < 0.05) more important in Pyl-1 compared to Col-0 at 0.1, 1 and 50 mM nitrate (Figure 3a). No significant difference in the density of lateral root primordia (D LRP ) was observed between the genotypes across all nitrate concentrations; however, D LRP was greater in Pyl-1 than Col-0 at 50 mM nitrate (Figure 3b). By contrast, with the macroscopic root morphological traits (Figure 2), N LRP and D LRP showed a large N treatment effect (44% and 71%, respectively) ( Figure S2).
The lateral root formation follows a precise developmental program during which successive phases, from the initiation to the emergence, arise to penetrate the parent root tissue layers [33] (Figure 3c). At 1 mM nitrate, almost three-quarters of the lateral root primordia emerged from the primary root in Col-0 ( Figure 3d). That proportion decreased when the seedlings were subjected to lower or greater nitrate concentrations. The percentages of the emerged (<1 mm) and visible (>1 mm) lateral roots were 54 % and 13%, respectively, at 0.1 mM nitrate, while these values reached 28% and 33% at 50 mM nitrate. By comparison, three-thirds of the Pyl-1 primordia had not emerged from the primary root at 1 mM nitrate. The percentages of the emerged and visible lateral roots reached 23% and 19%, respectively, at 0.1 mM nitrate, while only 7% emerged at 50 mM nitrate. Across all nitrate concentrations, Pyl-1 had greater cumulated percentages at stages III, IV and V compared to Col-0 (Figure 3d). At 50 mM nitrate, more than one-third of the primordia stayed at stage IV. This indicates that the intermediate stages of progression through the cortex tissue were affected rather than at the initiation.  Table S1.

Lateral Root Primordia Exhibit Slower Organ Emergence in Pyla-1
A root bending assay was conducted to follow the temporal sequence of lateral root formation. By applying a 90 • gravitropic stimulus, the lateral root initiation was synchronically induced at the outer edges of the bending roots. Then, the developmental stages of the lateral root primordia at the root bends were scored every 6 h between 18 h and until their emergence from the primary root. In Col-0, the lateral root initiation (stage I) was detected 18 h post gravity induction, and the emergence (stage E) happened after 42 h (Figure 4a). These observations matched with previous reports on the reference accession [25]. In Pyl-1, the lateral root initiation happened at the same time compared to Col-0. However, stages I and II accumulated during the first hours, and the progression to stages III and IV was drastically delayed (Figure 4b). After 42 h, most Pyl-1 primordia were still blocked around these stages. Eventually, the primordia progressed to the later stages, VI and VII, with the majority of them emerging from the parent root after 60 h. The newly emerged lateral roots did not elongate after that time (data not shown). These data indicated that Pyl-1 had a slower lateral root developmental program with a bottleneck around stages III-V. The accumulation of these specific stages pointed to a defect in the transition from a flat to dome-shaped lateral root primordium [15,34]. This prompted us to examine the shapes of primordia (length, height and area) at stages II, III and IV in the two genotypes ( Figure S2). The length and/or the area of the primordia were significantly (p < 0.05) increased by 5-15% in Pyl-1 compared to Col-0. This is documenting that the Pyl-1 lateral root primordium is misshaped.

Discussion
The Arabidopsis thaliana species has a vast geographical distribution. Natural populations growing in a wide range of soil conditions around the world are showing a large variation in the expression of numerous phenotypic traits and also at the root organ level [35,36]. These accessions provide a rich and diverse genetic resource for studying NUE processes [37][38][39][40] and the N-related adaptive differences in root morphology [41,42].
The Pyl-1 accession shows a fascinating root phenotype that is slightly responsive to nitrate availability (Figure 1).
The developmental and morphological defects are largely documented during the lateral root organogenesis of the Arabidopsis model [17,43]. For instance, the alf4-1, arf7arf19, gata23, and slr-1 mutants [44][45][46][47] are unable to trigger the initiation cascade, resulting in a lateral-root-less phenotype. In terms of the macroscopic root architecture, Pyl-1 looks very similar to these mutants. However, this natural accession has identical or even greater densities of lateral root primordia compared to the Col-0 wild-type (Figure 3a). This clearly indicates that Pyl-1 maintains the ability to initiate lateral root primordia. Thus, the postinitiation events seem to be critical in modulating the Pyl-1 macroscopic root morphology. A biomechanical constraint may be exerted on the primordia breaking through overlying tissues. Such a mechanical constraint is documented to cause a change in primordium shape [14,17]. The contours of some Pyl-1 primordia are altered ( Figure S2), further pointing to such mechanical properties of the cortex. Further studies on the tensile properties of tissues need to be carried out.
During the lateral root formation sequence (Figure 3d), the progression occurring from early primordium development is slowed down, with a bottleneck arising around stages III-V, followed by a later constraint on the elongation out of the parent root. These defects likely involve the phytohormone auxin, which is a key regulator of lateral root formation [48]. The auxin accumulation in the cells overlying the primordium depends on transcription factor LATERAL ORGAN BOUNDARIES-DOMAIN 29 (LBD29) regulating the auxin influx carrier LIKE-AUX1 3 (LAX3) [15,49]. The loss-of-function of these two genes results in a slowdown from stages III to V [15,49], as observed in Pyl-1. Moreover, this hormone alters cell wall properties by inducing wall loosening [50], and auxin-dependent cell separation promotes lateral organ emergence [15]. Furthermore, auxin regulates the root tissue hydraulic properties and aquaporin function to facilitate emergence [25]. Altering PLASMA MEMBRANE INTRINSIC PROTEIN 2;1 (PIP2;1) causes a flattened dome-shape primordia and delays emergence [25]. Another case was reported on the overexpression of MYB DOMAIN PROTEIN 93 (MYB93) negatively impacting the primordium progression [51,52]. Likewise, the loss of the MYB36 function results in the accumulation of stages IV-V [34]. A thorough analysis of the genetic and developmental background of the Pyl-1 root phenotype will be essential in the future.
The natural Arabidopsis populations undergo various selective environmental pressures. Developmental and physiological adaptations are likely to occur in order to optimize plant performance in different habitats [53,54]. The Pyl-1 accession was initially collected in the Pyla area, Arcachon Bay, France. There, the tallest sand dune in Europe is found. The Pyl-1 root phenotype observed in vitro ( Figure 1) may reflect an adaptation to sandy soil conditions, where the primary root grows downwards to the lower strata to take up water without elongating the lateral roots. However, the lack of data on local soil conditions makes it difficult to support this hypothesis further. Therefore, we launched a prospection campaign in the Arcachon Bay area. No accession was found in the Dune of Pilat, but four of them were collected in the Arcachon urban environment ( Figure S1). These accessions were growing on sidewalks in the city, and soil sampling was impossible. Thus, the hypothesis of an adaptation to a sand habitat is still lacking to support field evidence at this stage. Additional observations of root morphology upon growth conditions in rhizotrons filled with sand could provide more indications of the Pyl-1 phenotype.

Conclusions
Pyl-1 is a fascinating plant material for studying root organogenesis. The next step will be to identify the alleles(s) responsible for the lateral-root-less phenotype. After generating one F2 (Col-0 × Pyl-1)-segregating population, a mapping-by-sequencing strategy [55] could be employed on two pooled genomes showing contrasting lateral root outgrowth. A further molecular characterization could involve monitoring the expression of the genes involved in cell separation and lateral root emergence. The molecular dissection of the lateral root development in Pyl-1 could greatly contribute to drawing strategies for optimizing the root system architecture and improving N capture in plants.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/nitrogen3030029/s1, Table S1: Variance components extracted from mixed linear models, Figure S1: Arabidopsis accessions in the Arcachon Bay area, France, Figure S2: Comparison of the shape of lateral root primordia in Columbia-0 and Pyla-1 accessions.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.