https://orcid.org/0000-0002-6323-167X
Abstract
To test the hypothesis that particular tissues can control root growth, we analysed the mechanical properties of cell walls belonging to different tissues of the apical part of the maize root using atomic force microscopy. The dynamics of properties during elongation growth were characterized in four consecutive zones of the root. Extensive immunochemical characterization and quantification were used to establish the polysaccharide motif(s) related to changes in cell wall mechanics. Cell transition from division to elongation was coupled to the decrease in the elastic modulus in all root tissues. Low values of moduli were retained in the elongation zone and increased in the late elongation zone. No relationship between the immunolabelling pattern and mechanical properties of the cell walls was revealed. When measured values of elastic moduli and turgor pressure were used in the computational simulation, this resulted in an elastic response of the modelled root and the distribution of stress and strain similar to those observed in vivo. In all analysed root zones, cell walls of the inner cortex displayed moduli of elasticity that were maximal or comparable with the maximal values among all tissues. Thus, we propose that the inner cortex serves as a growth-limiting tissue in maize roots.
Introduction
Plant cell growth is an irreversible increase in volume and/or accumulation of structures. Each plant cell has an extracellular matrix made mainly of polysaccharides called the cell wall. Within plant tissues, the walls of neighbouring cells are connected and cells grow synchronously to maintain these contacts. Isotropic growth is when an individual cell grows equally in all directions, while anisotropic growth is when growth rate differs in different directions (Baskin et al., 1999). Plant cell growth is dependent on turgor pressure and the mechanical properties of cell walls (Schopfer, 2006). Cells within axial plant organs usually grow much faster in the direction that is parallel to the organ axis, exemplifying coordinated anisotropic growth, which is also known as elongation.
Since turgor pressure acts isotropically, anisotropic growth occurs only through the specific properties of the cell wall that allow for differential growth patterns (Baskin, 2005). Indeed, anisotropic cell wall properties within one cell were shown in the internodal cells of Nitella opaca (Probine and Preston, 1962), onion, and Kalanchoe leaves (Kerstens et al., 2001), pollen tubes (Zerzour et al., 2009), the epidermis of Arabidopsis hypocotyls (Peaucelle et al., 2015; Daher et al., 2018), and cotyledons (Majda et al., 2017). The growth and morphogenesis of plant organs are also determined by variations in cell wall mechanical properties. Softening of the rhizodermis and stele cell walls in maize roots was detected in the elongation zone when compared with the meristem (Abeysekera and McCully, 1994; Kozlova et al., 2019). Extensibility of cucumber hypocotyls, oat coleoptiles, and maize roots is correlated with their local growth rates (McQueen-Mason et al., 1992; Cosgrove and Li, 1993; Wu et al., 1994). Softening of the cell walls on the flanks of Arabidopsis shoot meristem precedes primordia formation (Milani et al., 2011; Peaucelle et al., 2011), and both elastic and plastic compliance decline basipetally in Arabidopsis inflorescence stem, together with the growth rate (Phyo et al., 2017).
Plant organs are not just a conglomerate of anisotropically growing individual cells, and mechanical influences of different tissues upon each other should be taken into account when investigating plant growth and development (Tomos et al., 1989). The coleoptiles, hypocotyls, and internodes in different plant species bend outward when longitudinally cut, isolated epidermal layers contract, while peeled segments of stems do not contract or even expand; outer tissues represent the primary target for auxin action and react to it by loosening their cell walls. These observations led to the formation of the ‘tensile skin’ theory for stem elongation (Kutschera and Niklas, 2007). This theory proposes that the thick outer epidermis cell walls restrain inner tissues of aerial plant organs to limit their growth.
Outer cell walls of the rhizodermis are also thick, but the ‘tensile skin’ theory is not applicable to plant roots. In experiments investigating root growth, longitudinally split roots showed inward bending (Burstrom, 1971; Sachs, 1882; Abeysekera and McCully, 1994). Peeled roots continued to elongate at the same rate and zone distribution as control roots (Yang et al., 1990; Bjorkman and Cleland, 1991). These observations led to the conclusion that some inner tissue controls anisotropic root growth. Tomos et al. (1989) proposed that the endodermis may play that role. However, no unequivocal evidence to support this hypothesis has been reported.
Mechanical properties of plant cell walls are dependent on their composition and architecture. Orientation of cellulose microfibrils (Baskin, 2005; Anderson et al., 2010), the degree of methylation of homogalacturonan (Zerzour et al., 2009; Peaucelle et al., 2011, 2015; Daher et al., 2018), the side chains of rhamnogalacturonan I (RGI; McCartney et al., 2000; Ulvskov et al., 2005; Majda et al., 2017; Torode et al., 2018), and the presence of other cell wall polysaccharides (Cavalier et al., 2008; Vega-Sanchez et al., 2012; Smith-Moritz et al., 2015; Xiao et al., 2016) all influence cell wall mechanics and are regulated during plant growth.
We recently developed an atomic force microscopy- (AFM) based approach to characterize mechanical properties of primary cell walls of the inner tissues of axial plant organs (Kozlova et al., 2019). To test the hypothesis that particular tissues can control root growth, we analysed 14 types of cell walls belonging to different tissues of the apical part of maize root. The dynamics of these properties during elongation growth were characterized in four consecutive zones of the root. The results were verified through computational modelling, and extensive immunochemical characterization and quantification were used to establish the polysaccharide motif(s) related to changes in cell wall mechanics.
Materials and methods
Plant material
Maize (Zea mays, cultivar Interkras 375) seedlings were grown hydroponically for 4 d in the dark at 27 °C. The primary root was subdivided into zones similar to the pattern previously described (Kozlova et al., 2020): meristem (0–1 mm from the root cap junction), early elongation (1–2 mm), elongation (2–6 mm), and late elongation (6–10 mm) (Fig. 1A). Seven tissues were examined: rhizodermis (Rh), exodermis (Exo), outer and inner cortex (OC and IC), endodermis (End), pericycle (Per), vascular parenchyma (VP), and pith (P) (Fig. 1B). Phloem and xylem were excluded from the analysis. Cells of these tissues are barely recognizable at early stages of development; they are scattered in vascular parenchyma whose contribution to observations is difficult to estimate and they do not form a continuous ring, and thus cannot be modelled axisymmetrically.
Root zones (A) and cell wall types (B) of maize that were investigated in the current study. ORh, outer rhizodermis; Rh, rhizodermis; Exo, exodermis; OC, outer cortex; IC, inner cortex; End, endodermis; Per, pericycle; VP, vascular parenchyma; P, pith.
Measurement of cell wall mechanical properties
Measurement of cell wall mechanical properties was conducted on agarose-embedded 400 μm thick cross-sections of four zones of maize roots (Fig. 1A) using AFM. Fourteen types of cell walls were studied in each zone (Fig. 1B). Sections were made close to the middle part of each zone using a Leica VT 1000S vibratome (Leica Biosystems, Germany) as was described previously (Kozlova et al., 2019). AFM experiments were performed at room temperature in a liquid cell using an NTEGRA Prima microscope (NT-MDT, Russia). AFM HybriD mode was used to obtain the elasticity map (Supplementary Fig. S1A, B) using Biosphere B150-FM tips (NanoTools, Germany) with a typical resonance frequency of 75 kHz, an average spring constant of 2.8 N m–1, and an apex radius of 150 nm. The thermal tune procedure was performed for each new cantilever to determine its unique spring constant. Deflection sensitivity was determined at room temperature in water on a fresh cover glass for each new cantilever, between samples, and after every laser adjustment. Scanning was conducted at a speed of 5 s per line in both forward and backward directions. The typical scanning area was 60×60 µm with 256×256 points resolution. The large variation in the height of cross-sections resulted in many artefact force–indentation (F–I) curves (Kozlova et al., 2019). F–I curves devoid of artefacts were selected manually from the elasticity map and fitted to the Derjaguin–Muller–Toporov (DMT) model of contact between a sphere and a half-space.
Immunohistochemistry
Cross- and longitudinal sections (50 µm thick) were prepared with a vibratome. Sections were incubated for 10 min in 4% (w/v) paraformaldehyde solution made in 0.2 M phosphate-buffered saline (PBS, pH 7.2). The immunolabelling procedure was performed using BG1 (Australian Biosupplies Pty Ltd, Australia); AX1 and INRA-RU2 (INRA, France); and LM5, LM6, LM11, LM12, LM19, LM20, LM25, LM26, LM27, and LM28 (Leeds University, UK) antibodies as previously described (Kozlova et al., 2020). Antibody and dye specificities, dilutions, and corresponding secondary antibodies are listed in Supplementary Table S1. For cellulose staining, sections were incubated in Carbotrace™680 (Ebba Biotech, Sweden) and then washed three times with PBS. Sections were observed using a Leica DM1000 epifluorescence microscope (Leica Biosystems, Germany) fitted with a mercury lamp and appropriate filter cubes [excitation filter 460–500 nm, barrier filter 512–542 nm for fluorescein isothoiocyanate- (FITC) conjugated antibodies; excitation filter 540–580 nm, barrier filter 608–683 nm for Carbotrace™680] and constant exposure times. The mean fluorescence signal intensity was measured on a single representative cross-section at high magnification using ImageJ2 Fiji software (https://imagej.net/). The procedure for these measurements is described in Supplementary Fig. S2. Fluorescence signal intensity depends on the cell wall thickness (Haas et al., 2020). The intensity of cellulose staining by Carbotrace™680 was used as a proxy for normalization.
Turgor pressure evaluation
Turgor pressure was calculated using a Python script developed by Beauzamy et al. (2015). The input parameters included the curvature of turgescent cells, their apparent elastic modulus, and cell wall thickness. The maize root apex (15 mm) was embedded in low-melting point agarose to avoid displacement (Supplementary Fig. S3). The rhizodermis was investigated by contact mode AFM in water using B100-CONT tips (Nanotools, Germany) with a typical 13 kHz resonant frequency, average spring constant of 0.2 N m–1, and 100 nm apex radius. The scanning area was 60×60 µm with 64×64 points resolution. Scanning was conducted at a speed of 5 s per line. The curvature was calculated from topography images. The apparent elastic modulus was calculated from F–I curves fitted to the DMT model of contact between a sphere and a half-space only if the curves were free of artefacts. The thickness of maize rhizodermis cell walls from different root zones was taken from the literature (Clarke et al., 1979; Chaboud and Rougier, 1986; Abeysekera and McCully, 1993).
Elastic deformation assay
The maize root apices (15 mm) embedded in low-melting point agarose (Supplementary Fig. S3) were imaged before and after 30 min incubation in a 10% (w/w) sorbitol solution. Rhizodermis was imaged under UV light using an epifluorescence microscope fitted with the UV lamp and filter cube (355–425 nm excitation filter and 470 nm barrier filter). The width and the length of rhizodermis cells were measured from micrographs in ImageJ2 Fiji (https://imagej.net/) and then used to calculate elastic strains (ε, Equation 1) in the X- and Y-directions after deflation.
where L0 is initial length and L1 is length after removal of turgor pressure.
Finite element modelling
Maize primary roots and their elastic response were simulated using ANSYS v19.1 (ANSYS Inc.). The axisymmetric model included seven types of tissues (i.e. Rh, Exo, OC, IC, End, Per, VP, and P) and four zones of root development (i.e. meristem, early elongation, elongation, and late elongation) (Supplementary Fig. S4). The initial size of the model and zone distribution were obtained from the elastic deformation assay. Tissue dimensions were measured on micrographs of 300 µm thick cross-sections for each millimetre in ImageJ2 Fiji (https://imagej.net/). For the modelling, the dimensions of each tissue were changed according to the results of the elastic deformation assay for the X-direction. Each tissue was modelled as an isotropic material with the elastic modulus of its cell walls measured in the current study. Some tissues had cell walls whose properties differed on different sides of the cells. In this case, the elastic modulus for such tissues was taken as the average of values for corresponding cell walls. For example, the modulus of the rhizodermis was the mean value of moduli measured on outer rhizodermis (ORh), radial rhizodermis (Rh/Rh) and rhizodermis/exodermis (Rh/Exo) cell walls. Three different Poisson’s ratios (0.3, 0.4, and 0.45) were tested. Roots were modelled using quadrangular 20 µm elements. Mesh independence of the results was tested by using different coarseness of meshes.
The Y-displacement of the upper edge of modelled roots was fixed, while the rest of the root was free to displace in any direction. Turgor pressure evaluated in the current study was applied to each tissue (Supplementary Fig. S4).
The relationship between stress and strain was described by a Saint Venant–Kirchhoff strain energy (W) for an isotropic material (Equation 2).
where έ is the Green–Lagrange strain tensor, and µ and λ are the Lamé constants of the material. The Lamé constants are connected to the modulus of elasticity E and the Poisson ratio ν by (Equations 3 and 4)
Statistics
Twenty-four roots were examined for cell wall mechanical properties using AFM. Twenty roots were used for turgor pressure measurements. Twenty-five roots were used in the elastic deformation assay. Four biological replicates for each antibody were used for the immunocytochemical analysis. At least six individual cell walls from each studied tissue and in each root zone were used to collect measurements of fluorescence intensity for each antibody. Mean values ±SD among the biological replicates are presented. Mean separation was performed by ANOVA followed by the Tukey test at α=0.01 using the SPSS software package (v.21, IBM Corp.).
Results
Variations of moduli of elasticity along and across the maize root
The cell walls of maize root tissues demonstrated three types of mechanical property changes along the root length (Table 1). Some cell walls (ORh, Rh/Exo, Exo/Exo, Exo/OC, IC/End, and Per/VP) had similar elastic moduli in the meristem, early elongation zone, and active elongation zone, and increased their stiffness in the late elongation zone. Following the results of ANOVA, we have named this type of dynamics of elasticity during growth the AAAB-type, where the four letters correspond to four zones of the root starting from the meristem, and A corresponds to a lower modulus value than B (Table 1). Other cell walls (Rh/Rh, OC/OC, IC/IC, End/End, End/Per, and P/P) were relatively stiff in the meristem, softened in the early elongation zone, and in the late elongation zone returned to the high modulus values similar to those that were observed in the meristem (BAAB-type of elasticity dynamics). Two types of cell walls (Per/Per and VP/VP) had reduced moduli of elasticity in the early elongation and elongation zones compared with the meristem; however, in the late elongation zone, their moduli of elasticity exceeded values of those measured in the meristem (BAAC-type) (Table 1).
The apparent elastic moduli (MPa) of different cell walls in different root zones in maize
| Meristem | Early elongation | Elongation | Late elongation | |
|---|---|---|---|---|
| Outer rhizodermis (ORh) | 1.73±0.17 abA | 1.18±0.20 abcA | 1.35±0.40 aA | 2.96±0.78 bcd B |
| Rhizhodermis/rhizodermis (Rh/Rh) | 1.88±0.26 abcB | 0.97±0.13 abA | 1.18±0.35 aA | 1.86±0.40 ab B |
| Rhizodermis/exodermis (Rh/Exo) | 1.75±0.26 abA | 1.21±0.19 abcA | 1.18±0.22 aA | 2.87±0.51 abcd B |
| Exodermis/exodermis (Exo/Exo) | 1.39±0.24 aA | 1.59±0.27 cdeA | 1.52±0.55 aA | 2.68±0.26 abcd B |
| Exodermis/outer cortex (Exo/OC) | 1.41±0.24 aA | 0.97±0.16 abA | 1.12±0.26 aA | 2.18±0.28 abc B |
| Outer cortex/outer cortex (OC/OC) | 2.72±0.36 deB | 1.50±0.12 cdeA | 1.43±0.29 aA | 2.70±0.40 abcd B |
| Inner cortex/inner cortex (IC/IC) | 3.54±0.38 eB | 1.79±0.18 deA | 1.56±0.21 aA | 3.00±0.67 cd B |
| Inner cortex/endodermis (IC/End) | 1.72±0.24 abA | 1.21±0.24 abcA | 1.49±0.20 aA | 3.16±0.24 cd B |
| Endodermis/endodermis (End/End) | 2.63±0.36 cdB | 1.86±0.26 eA | 1.54±0.35 aA | 2.79±0.45 abcd B |
| Endodermis/pericycle (End/Per) | 2.45±0.44 bcdB | 1.47±0.11 cdeA | 1.51±0.38 aA | 3.07±0.47 cd B |
| Pericycle/pericycle (Per/Per) | 2.05±0.15 abcdB | 1.38±0.11 bcdA | 1.25±0.16 aA | 3.23±0.33 cd C |
| Pericycle/vascular parenchyma (Per/VP) | 2.14±0.57 abcdA | 1.48±0.22 cdeA | 1.45±0.03 aA | 3.35±0.31 d B |
| Vasxcular parenchyma/vascular parenchyma (VP/VP) | 2.17±0.33 abcdB | 1.39±0.15 bcdA | 1.41±0.20 aA | 3.22±0.63 cd C |
| Pith/pith (P/P) | 1.53±0.36 aB | 0.92±0.12 aA | 1.02±0.09 aA | 1.82±0.19 a B |
| Meristem | Early elongation | Elongation | Late elongation | |
|---|---|---|---|---|
| Outer rhizodermis (ORh) | 1.73±0.17 abA | 1.18±0.20 abcA | 1.35±0.40 aA | 2.96±0.78 bcd B |
| Rhizhodermis/rhizodermis (Rh/Rh) | 1.88±0.26 abcB | 0.97±0.13 abA | 1.18±0.35 aA | 1.86±0.40 ab B |
| Rhizodermis/exodermis (Rh/Exo) | 1.75±0.26 abA | 1.21±0.19 abcA | 1.18±0.22 aA | 2.87±0.51 abcd B |
| Exodermis/exodermis (Exo/Exo) | 1.39±0.24 aA | 1.59±0.27 cdeA | 1.52±0.55 aA | 2.68±0.26 abcd B |
| Exodermis/outer cortex (Exo/OC) | 1.41±0.24 aA | 0.97±0.16 abA | 1.12±0.26 aA | 2.18±0.28 abc B |
| Outer cortex/outer cortex (OC/OC) | 2.72±0.36 deB | 1.50±0.12 cdeA | 1.43±0.29 aA | 2.70±0.40 abcd B |
| Inner cortex/inner cortex (IC/IC) | 3.54±0.38 eB | 1.79±0.18 deA | 1.56±0.21 aA | 3.00±0.67 cd B |
| Inner cortex/endodermis (IC/End) | 1.72±0.24 abA | 1.21±0.24 abcA | 1.49±0.20 aA | 3.16±0.24 cd B |
| Endodermis/endodermis (End/End) | 2.63±0.36 cdB | 1.86±0.26 eA | 1.54±0.35 aA | 2.79±0.45 abcd B |
| Endodermis/pericycle (End/Per) | 2.45±0.44 bcdB | 1.47±0.11 cdeA | 1.51±0.38 aA | 3.07±0.47 cd B |
| Pericycle/pericycle (Per/Per) | 2.05±0.15 abcdB | 1.38±0.11 bcdA | 1.25±0.16 aA | 3.23±0.33 cd C |
| Pericycle/vascular parenchyma (Per/VP) | 2.14±0.57 abcdA | 1.48±0.22 cdeA | 1.45±0.03 aA | 3.35±0.31 d B |
| Vasxcular parenchyma/vascular parenchyma (VP/VP) | 2.17±0.33 abcdB | 1.39±0.15 bcdA | 1.41±0.20 aA | 3.22±0.63 cd C |
| Pith/pith (P/P) | 1.53±0.36 aB | 0.92±0.12 aA | 1.02±0.09 aA | 1.82±0.19 a B |
Mean values ±SD are presented. Lower case letters correspond to significant differences within one column, while upper case letters correspond to significant differences within a row. Mean separation was determined by one-way ANOVA followed by Tukey test at α=0.01.
The apparent elastic moduli (MPa) of different cell walls in different root zones in maize
| Meristem | Early elongation | Elongation | Late elongation | |
|---|---|---|---|---|
| Outer rhizodermis (ORh) | 1.73±0.17 abA | 1.18±0.20 abcA | 1.35±0.40 aA | 2.96±0.78 bcd B |
| Rhizhodermis/rhizodermis (Rh/Rh) | 1.88±0.26 abcB | 0.97±0.13 abA | 1.18±0.35 aA | 1.86±0.40 ab B |
| Rhizodermis/exodermis (Rh/Exo) | 1.75±0.26 abA | 1.21±0.19 abcA | 1.18±0.22 aA | 2.87±0.51 abcd B |
| Exodermis/exodermis (Exo/Exo) | 1.39±0.24 aA | 1.59±0.27 cdeA | 1.52±0.55 aA | 2.68±0.26 abcd B |
| Exodermis/outer cortex (Exo/OC) | 1.41±0.24 aA | 0.97±0.16 abA | 1.12±0.26 aA | 2.18±0.28 abc B |
| Outer cortex/outer cortex (OC/OC) | 2.72±0.36 deB | 1.50±0.12 cdeA | 1.43±0.29 aA | 2.70±0.40 abcd B |
| Inner cortex/inner cortex (IC/IC) | 3.54±0.38 eB | 1.79±0.18 deA | 1.56±0.21 aA | 3.00±0.67 cd B |
| Inner cortex/endodermis (IC/End) | 1.72±0.24 abA | 1.21±0.24 abcA | 1.49±0.20 aA | 3.16±0.24 cd B |
| Endodermis/endodermis (End/End) | 2.63±0.36 cdB | 1.86±0.26 eA | 1.54±0.35 aA | 2.79±0.45 abcd B |
| Endodermis/pericycle (End/Per) | 2.45±0.44 bcdB | 1.47±0.11 cdeA | 1.51±0.38 aA | 3.07±0.47 cd B |
| Pericycle/pericycle (Per/Per) | 2.05±0.15 abcdB | 1.38±0.11 bcdA | 1.25±0.16 aA | 3.23±0.33 cd C |
| Pericycle/vascular parenchyma (Per/VP) | 2.14±0.57 abcdA | 1.48±0.22 cdeA | 1.45±0.03 aA | 3.35±0.31 d B |
| Vasxcular parenchyma/vascular parenchyma (VP/VP) | 2.17±0.33 abcdB | 1.39±0.15 bcdA | 1.41±0.20 aA | 3.22±0.63 cd C |
| Pith/pith (P/P) | 1.53±0.36 aB | 0.92±0.12 aA | 1.02±0.09 aA | 1.82±0.19 a B |
| Meristem | Early elongation | Elongation | Late elongation | |
|---|---|---|---|---|
| Outer rhizodermis (ORh) | 1.73±0.17 abA | 1.18±0.20 abcA | 1.35±0.40 aA | 2.96±0.78 bcd B |
| Rhizhodermis/rhizodermis (Rh/Rh) | 1.88±0.26 abcB | 0.97±0.13 abA | 1.18±0.35 aA | 1.86±0.40 ab B |
| Rhizodermis/exodermis (Rh/Exo) | 1.75±0.26 abA | 1.21±0.19 abcA | 1.18±0.22 aA | 2.87±0.51 abcd B |
| Exodermis/exodermis (Exo/Exo) | 1.39±0.24 aA | 1.59±0.27 cdeA | 1.52±0.55 aA | 2.68±0.26 abcd B |
| Exodermis/outer cortex (Exo/OC) | 1.41±0.24 aA | 0.97±0.16 abA | 1.12±0.26 aA | 2.18±0.28 abc B |
| Outer cortex/outer cortex (OC/OC) | 2.72±0.36 deB | 1.50±0.12 cdeA | 1.43±0.29 aA | 2.70±0.40 abcd B |
| Inner cortex/inner cortex (IC/IC) | 3.54±0.38 eB | 1.79±0.18 deA | 1.56±0.21 aA | 3.00±0.67 cd B |
| Inner cortex/endodermis (IC/End) | 1.72±0.24 abA | 1.21±0.24 abcA | 1.49±0.20 aA | 3.16±0.24 cd B |
| Endodermis/endodermis (End/End) | 2.63±0.36 cdB | 1.86±0.26 eA | 1.54±0.35 aA | 2.79±0.45 abcd B |
| Endodermis/pericycle (End/Per) | 2.45±0.44 bcdB | 1.47±0.11 cdeA | 1.51±0.38 aA | 3.07±0.47 cd B |
| Pericycle/pericycle (Per/Per) | 2.05±0.15 abcdB | 1.38±0.11 bcdA | 1.25±0.16 aA | 3.23±0.33 cd C |
| Pericycle/vascular parenchyma (Per/VP) | 2.14±0.57 abcdA | 1.48±0.22 cdeA | 1.45±0.03 aA | 3.35±0.31 d B |
| Vasxcular parenchyma/vascular parenchyma (VP/VP) | 2.17±0.33 abcdB | 1.39±0.15 bcdA | 1.41±0.20 aA | 3.22±0.63 cd C |
| Pith/pith (P/P) | 1.53±0.36 aB | 0.92±0.12 aA | 1.02±0.09 aA | 1.82±0.19 a B |
Mean values ±SD are presented. Lower case letters correspond to significant differences within one column, while upper case letters correspond to significant differences within a row. Mean separation was determined by one-way ANOVA followed by Tukey test at α=0.01.
Significant variations in moduli were also observed between cell walls in different tissues within one root zone. Pith (P/P) and exodermis (Exo/Exo and Exo/OC) cell walls were the softest in the maize root meristem (Table 1). The highest moduli among all studied cell wall types in the meristem were observed in the inner cortex (IC/IC). In the early elongation zone, the elasticity indicators of the outer (OC/OC) and inner (IC/IC) cortex, and endodermis (End/End) were significantly higher than that of other tissues. In the active elongation zone, the apparent modulus values of all tissues insignificantly differed from each other. In the late elongation zone, vascular parenchyma (VP/VP), pericycle (Per/VP, Per/Per, and Per/End), and inner cortex (End/IC and IC/IC) had significantly higher moduli of elasticity than radial cell walls of the rhizodermis (Rh/Rh) and pith (P/P) cell walls.
Immunolabelling dynamics in maize root
To identify cell wall components that could provide particular mechanical properties to maize root cell walls, we performed semi-quantitative immunological screening. The antibody for mixed-linkage glucans, BG1 (Meikle et al., 1994), labelled all tissues of maize root in all zones, with the exception of the outer cell walls of the rhizodermis in meristems (Fig. 2A). We found that the BG1 labelling was weaker in the inner cortex and endodermis of the meristem and early elongation zone compared with other tissues in these zones (Fig. 2A, B). LM25 recognizes galactoxyloglucans (Pedersen et al., 2012). The overall intensity of LM25 labelling was high in the early elongation zone and gradually decreased towards the late elongation zone. Similar to BG1, the LM25 antibody displayed a lower affinity for the cell walls of the inner cortex and endodermis in the early elongation zone when compared with other root tissues in this zone (Fig. 2B, C).
Fluorescence micrographs of maize root cross-sections immunolabelled by BG1 (mixed-linkage glucan) (A) or LM25 (galactoxyloglucan) (C), or stained with Carbotrace™680 (D). Scale bars=100 μm. Fluorescence signal intensity of different cell wall components (B). Note the darker rings between the stele and cortex, which are indicated by white arrows. ORh, outer rhizodermis; Rh, rhizodermis; Exo, exodermis; OC, outer cortex; IC, inner cortex; End, endodermis; Per, pericycle; VP, vascular parenchyma; P, pith.
Carbotrace™680 is a fluorescent dye which was used to detect cellulose (Choong et al., 2019). In the meristem of maize roots, a weaker fluorescence was observed in inner tissues when compared with outer tissues of the root (Fig. 2B, D). The signal from the pith and vascular parenchyma was strong in the early elongation zone, while cell walls of the inner cortex, endodermis, and pericycle remained weakly stained. A low fluorescence intensity was also a characteristic of the cell walls in these tissues in the active elongation and late elongation zones.
Longitudinal sections of maize roots were labelled using the same antibodies and dye to further characterize cell wall polysaccharide dynamics (Fig. 3). A weak BG1 labelling was observed in the inner cortex and endodermis starting from the quiescent centre of the root (Fig. 3A). In the elongation zone, all tissues were labelled uniformly. In contrast, LM25 labelling indicated that galactoxyloglucans were evenly distributed in the cell walls of the meristem, including the quiescent centre (Fig. 3B). However, in the early elongation zone, LM25 labelling became stronger in the outer cortex and stele tissues (Figs 2B, C, 3B) compared with the inner cortex and endodermis. In the elongation and late elongation zones, the LM25 epitope gradually disappeared from the stele and cortex, remaining in the outer cell walls of the rhizodermis and anticlinal cell walls of the metaxylem (Fig. 3B). The Carbotrace™680 signal was lower in the stele periphery than in other tissues in the early elongation, elongation, and late elongation zones (Fig. 3C). Anticlinal cell walls of the metaxylem were characterized by high intensity Carbotrace™680 staining.
Composite fluorescence micrographs of longitudinal sections of maize roots immunolabelled with BG1 (mixed-linkage glucan) (A), LM25 (galactoxyloglucan) (B), or stained with Carbotrace™680 (cellulose) (C). Scale bars=500 μm. 1, cortex; 2, endodermis and pericycle; 3, vascular parenchyma and pith. White arrows indicate the quiescent centre of the root.
The AX1 antibody recognizes arabinoxylans (Guillon et al., 2004) and was able to bind maize cell walls in all investigated zones (Fig. 4A). Outer cell walls of the rhizodermis were labelled by the AX1 antibody only in the late elongation zone, while other tissues in all root zones were labelled evenly (Supplementary Table S2). The LM28 antibody is a specific marker for glucuronoxylans (Cornuault et al., 2015) and labelled all tissues except the rhizodermis in meristems and the early elongation zones (Fig. 4B). The signal intensity decreased in root cortex during active and late elongation, while the cell walls of vascular parenchyma still possessed the LM28 epitope. Radial cell walls in the endodermis were specifically labelled at the late elongation stage (Fig. 4B). The LM11 epitope, which is believed to be a low-substituted xylan (McCartney et al., 2005), was present in the cell walls of vascular tissues in the early elongation, elongation, and late elongation zones (Fig. 4C). In the late elongation zone, the exodermis, rhizodermis, and radial endodermis cell walls contained these xylans. The LM12 antibody recognizes feruloylated polymers (Pedersen et al., 2012), which were present at the root surface and in the radial cell walls of the rhizodermis in all investigated zones. In the late elongation zone, LM12 signals were observed in all tissues, with the highest fluorescence in the outer cell walls of the rhizodermis, radial cell walls of the endodermis, and in vascular bundles.
Fluorescence micrographs of maize root cross-sections immunolabelled with AX1 (arabinoxylan) (A), LM28 (glucuronoxylan) (B), LM11 (low-substituted xylan) (C), and LM12 (feruloylated polysaccharides) (D) antibodies. Scale bars=100 μm.
The INRA-RU2 antibody recognizes the backbone of RGI (Ralet et al., 2010); we observed INRA-RU2 labelling in the root cap slime in all investigated zones of maize roots (Fig. 5A). Labelling was also observed in the cortex and pith throughout development; however, it was localized in cell corners, and thus the protocol for measuring fluorescence intensity was not applicable to the INRA-RU2 micrographs. Side chains of RGI are mainly composed of arabinose and galactose, and the LM6 antibody can be used to specifically identify those with arabinan side chains (Verhertbruggen et al., 2009). We did not observe an LM6 signal in the exodermis cell walls in the meristem (Supplementary Table S3), although other tissues showed moderate LM6 labelling (Fig. 5B). Linear 1,4-galactans recognized by the LM5 antibody (Jones et al., 1997) were detected in all investigated root zones; however, over the course of root development, the labelling disappeared from nearly all tissues. A slight increase in LM6 labelling intensity was observed in protoxylem and phloem in the elongation and late elongation zones compared with other tissues (Fig. 5C). The LM26 antibody binds to β-1,4-galactan branched at the O(6) position with another galactosyl residue (Torode et al., 2018). Among the antibodies used to detect RGI side chains, LM26 showed the highest labelling intensity in all tissues and at all stages of development (Fig. 5D; Supplementary Table S3).
Fluorescence micrographs of maize root cross-sections immunolabelled by INRA-RU2 (rhamnogalacturonan I backbone) (A), LM6 (1,5-α-arabinan) (B), LM5 (1,4-β-galactan) (C), and LM26 (1,6-branched 1,4-β-galactan) (D). Scale bars=100 μm.
The LM20 antibody recognizes methylesterified homogalacturonans, while LM19 detects unesterified homogalacturonans (Verhertbruggen et al., 2009). We observed LM20 labelling in meristems, including the quiescent centre and early elongation zone, but the signal was lost from cell walls at later stages of development (Fig. 6A). The LM19 antibody signal was observed in maize root cell walls in the meristem and early elongation zone (Fig. 6B), but not in the quiescent centre (white arrow in Fig. 6B). The intensity of LM19 labelling decreased in active elongation and late elongation zones. Neither LM19 nor LM20 antibodies labelled cells in the root cap or rhizodermis cell walls.
Composite fluorescence micrographs of longitudinal sections and cross-sections of maize roots immunolabelled by LM20 (esterified homogalacturonan) (A) and LM19 (unesterified homogalacturonan) (B) showing their fluorescence signals in different cell wall types. Scale bars=100 μm. The white arrow indicates the quiescent centre of the root.
We plotted signal intensities of all probes versus the elastic moduli of cell walls from certain tissues at particular developmental stages to test if there was any correlation between them (Fig. 7). Neither one single epitope nor a set of epitopes was associated with low or high elastic modulus values (Fig. 7), and two tissues having similar elastic moduli could possess dramatically different labelling patterns. For example, ORh and IC/End in the meristem both had an apparent elastic modulus of 1.7 MPa, but the BG1, AX1, LM25, LM19, LM20, LM5, and LM12 labelling patterns were different. Conversely, cell walls with significantly different elastic properties [i.e. P/P (0.92 MPa) and IC/IC (1.79 MPa) in the early elongation zone] were characterized by identical combinations of the epitopes that they possessed.
Occurrence of different polysaccharide motifs in cell walls of maize with different mechanical properties. The white–blue heatmap corresponds to elastic modulus values (MPa), while the white-green heatmap represents fluorescence signal intensity (arbitrary units) from the antibodies or dye used to evaluate cell wall composition. ORh, outer rhizodermis; Rh, rhizodermis; Exo, exodermis; OC, outer cortex; IC, inner cortex; End, endodermis; Per, pericycle; VP, vascular parenchyma; P, pith. BG1, mixed-linkage glucan; AX1, arabinoxylan; LM28, glucuronoxylan; LM11, xylan; LM25, galactoxyloglucan; LM19, unesterified homogalacturonan; LM20, methylesterified homogalacturonan; LM5, 1,4-galactan; LM6, 1,5-arabinan; LM26, 1,6-branched-1,4-galactan; LM12, feruloylated polymers, Carbotrace™680, cellulose.
These results indicate that among the polysaccharide motifs that were detected in maize root cell walls, there was none whose presence can be used as a proxy to predict the elastic modulus of cell walls.
Root modelling
Computer modelling requires a range of parameters to be established in advance. Turgor pressure was hypothesized to serve as a driving force for the growth process, and thus it was necessary to determine its magnitude and dynamics. The presence of turgor pressure can change the dimensions of individual cells and zones in the primary root. Thus, we performed an elastic deformation assay using hyperosmotic treatment of maize root cells to determine the dimensions of cells and tissues after they lose turgor pressure.
Elastic deformation assay
Root apices (15 mm long) were imaged before and after a 30 min incubation in a sorbitol solution (Fig. 8A, B). The loss of turgor resulted in root shrinkage in both width (X-direction negative strain) and length (Y-direction negative strain); the effect was similar in both directions (Fig. 8C). The early elongation and elongation zones displayed higher shrinkage than the meristem and late elongation zone.
Elastic deformation assay and turgor pressure in maize primary roots. The rhizodermis of maize roots in water (A) and after 30 min in a sorbitol solution (B) imaged under UV light. Scale bars=50 µm. Dashed squares indicate the same cells under the two conditions. Note the difference in cell size which indicates that the shrinkage of root cells occurred in both width and length after turgor loss. Mean values (n=25) of strains in X- (cell width) and Y- (cell length) directions observed after turgor loss (C). Different letters of the same colour correspond to a significant difference according to one-way ANOVA followed by Tukey test at α=0.01. Representative 3D topography image of rhizodermal cells (D). Note the curvature. Turgor pressure along the root is presented as mean values ±SD from n=20. No significant differences were observed from a one-way ANOVA followed by Tukey test at α=0.01 (E).
Turgor pressure in cells of maize root rhizodermis
Only the rhizodermal cells were used for evaluating changes in turgor pressure along the root because there is no radial gradient of turgor pressure in the apical part of maize primary roots (Spollen and Sharp, 1991). The first millimetre of root is covered by root cap cells and viscous mucilage, and thus could not be investigated. Changes in turgor pressure along the root were not found to be significant (Fig. 8E); therefore, the average value accounted for 0.6 MPa and was used for all root zones in the computational modelling.
Finite element method-based model of maize root
The mechanical properties of cell walls, dimensions of non-turgescent root zones, and turgor pressure values obtained in this study formed the basis for computational modelling of maize root elastic behaviour (Supplementary Fig. S4). In fact, we simulated a plasmolysed root and its response to turgor application. Three models built with different Poisson’s ratios were tested. All of the modelled roots were elongated under action of turgor. Total longitudinal deformation occurring upon application of turgor was compared with that observed in the elastic deformation assay (650 nm shrinkage of a 10 mm long root segment). Total longitudinal deformation at Poisson’s ratios of 0.3, 0.4, and 0.45 accounted for 1300, 636, and 316 nm, respectively. Thus, the Poisson’s ratio of 0.4 allows for better prediction of the real elastic response in maize root. Stress and strain distribution along and across the root were studied further on the model with a Poisson’s ratio of 0.4.
The stress in the X-direction averaged 0.7 MPa, with uniform distribution in all root zones (Fig. 9A), while stress in the Y-direction was distributed unevenly throughout the maize root (Fig. 9B). All root tissues were under tensile stress, but the highest values of stress were detected in the inner cortex in the meristem and early elongation zone.
Axisymmetric model of the apical region of primary maize root. Four consecutive zones from meristem to late elongation zone each having seven tissues were modelled to establish stress–strain distribution under action of turgor pressure. The stressed condition at application of turgor pressure is shown. Stress in the X-direction (A), stress in the Y-direction (B), strain in the X-direction (C), and strain in the Y-direction (D). Rh, rhizodermis; Exo, exodermis; OC, outer cortex; IC, inner cortex; End, endodermis; Per, pericycle; VP, vascular parenchyma; P, pith.
Positive strain values (extension) in the X-direction were observed in the pith and rhizodermis of the early elongation and elongation zones (Fig. 9C). However, the positive strain in these tissues was equilibrated by negative strain values (compression) in the inner cortex. Thus, the total change of root width was low. All tissues in all root zones had positive values of strain in the Y-direction, which demonstrated root elongation under the action of turgor pressure. Lower positive strain values were characteristic of tissues in the meristem and late elongation zone, and higher values of strain were observed in the early elongation and elongation zones (Fig. 9D), which was consistent with the results from the elastic deformation assay (Fig. 8D). Moreover, strain values in both the X- and Y-directions for each of the investigated root zones were similar to those observed in the elastic deformation assay (Table 2). Thus, the maize root, which was modelled with the mechanical properties of cell walls obtained within this study, does not collapse upon the application of turgor pressure but responds elastically in accordance with in vivo observations.
Comparison of average stains in different zones along the maize root obtained in elastic deformation assay and modelling (μm/μm)
| Elastic deformation assay | Modelling | |||
|---|---|---|---|---|
| Average X-directional strain | Average Y-directional strain | Average X-directional strain | Average Y-directional strain | |
| Meristem | 0.053±0.009 | 0.056±0.015 | 0.059 | 0.047 |
| Early elongation | 0.148±0.036 | 0.138±0.034 | 0.098 | 0.087 |
| Elongation | 0.114±0.023 | 0.081±0.033 | 0.083 | 0.086 |
| Late elongation | 0.035±0.012 | 0.033±0.009 | 0.040 | 0.043 |
| Elastic deformation assay | Modelling | |||
|---|---|---|---|---|
| Average X-directional strain | Average Y-directional strain | Average X-directional strain | Average Y-directional strain | |
| Meristem | 0.053±0.009 | 0.056±0.015 | 0.059 | 0.047 |
| Early elongation | 0.148±0.036 | 0.138±0.034 | 0.098 | 0.087 |
| Elongation | 0.114±0.023 | 0.081±0.033 | 0.083 | 0.086 |
| Late elongation | 0.035±0.012 | 0.033±0.009 | 0.040 | 0.043 |
Mean values ±SD are presented for the elastic deformation assay.
Comparison of average stains in different zones along the maize root obtained in elastic deformation assay and modelling (μm/μm)
| Elastic deformation assay | Modelling | |||
|---|---|---|---|---|
| Average X-directional strain | Average Y-directional strain | Average X-directional strain | Average Y-directional strain | |
| Meristem | 0.053±0.009 | 0.056±0.015 | 0.059 | 0.047 |
| Early elongation | 0.148±0.036 | 0.138±0.034 | 0.098 | 0.087 |
| Elongation | 0.114±0.023 | 0.081±0.033 | 0.083 | 0.086 |
| Late elongation | 0.035±0.012 | 0.033±0.009 | 0.040 | 0.043 |
| Elastic deformation assay | Modelling | |||
|---|---|---|---|---|
| Average X-directional strain | Average Y-directional strain | Average X-directional strain | Average Y-directional strain | |
| Meristem | 0.053±0.009 | 0.056±0.015 | 0.059 | 0.047 |
| Early elongation | 0.148±0.036 | 0.138±0.034 | 0.098 | 0.087 |
| Elongation | 0.114±0.023 | 0.081±0.033 | 0.083 | 0.086 |
| Late elongation | 0.035±0.012 | 0.033±0.009 | 0.040 | 0.043 |
Mean values ±SD are presented for the elastic deformation assay.
Discussion
Immunolabelling cannot be used as a proxy to predict cell wall mechanical properties in maize root
Several studies have reported the correlation between the presence of a particular polysaccharide in cell walls and their mechanical properties. In the coordinated growth of vegetative tissues of dicots, the presence of unesterified homogalacturonan corresponded to softer cell walls, while methylesterified homogalacturonans were characteristic of stiffer cell walls (Peaucelle et al., 2011, 2015; Braybrook and Peaucelle, 2013; Majda et al., 2017; Daher et al., 2018). An opposite pattern was described for tip-growing pollen tubes, where less methylation of homogalacturonans was attributed to the stiffer cell wall parts, while more methylation was observed in softer parts (Geitmann and Parre, 2004; Parre and Geitmann, 2005; Palin and Geitmann, 2012). None of these patterns has been observed in maize roots. Both stiff and soft cell walls may contain epitopes of high- or low-methylated homogalacturonans, of both forms, or none at all (Figs 6, 7; LM19 and LM20). However, the transition of cell wall properties from generally stiff in the meristem to relatively soft in elongation zone may occur in part due to an overall decrease in the degree of methylation of homogalacturonans, and thus may function in a manner similar to the coordinated growth of tissues in dicots.
Another pectic polysaccharide that may modulate cell wall mechanics is RGI, which has various side chains, and β-1,4-galactan chains in particular. Galactans have been associated with the hardening of pea cotyledon cell walls during seed maturation (McCartney et al., 2000). Transgenic potato tubers with a low level of β-1,4-galactans in their cell walls were more brittle when subjected to uniaxial compression tests than the wild type (Ulvskov et al., 2005). Additionally, Majda et al. (2017) found that β-1,4-galactans and α-1,5-arabinans in anticlinal cell walls of Arabidopsis pavement cells were associated with their decreased mechanical stiffness. Mechanical probing of transverse sections of Miscanthus giganteus stems and leaves indicated that phloem companion cells with abundant linear 1,4-galactan were less elastic than sieve elements that accumulated 1,6-branched 1,4-galactan. However, both phloem cell types had more elastic cell walls than those in the sclerenchyma or xylem (Torode et al., 2018). We found that none of the antibodies that recognize the different parts of RGI (e.g. INRA-RU2 for backbone, LM5, LM6, and LM26 for side chains) showed any preference of binding to more or less elastic cell walls in maize roots (Figs 5, 7). This may indicate that RGI structure variations are not exploited by maize cells to adjust cell wall mechanical properties.
Carbotrace™680 detects cellulose quantitatively (Choong et al., 2019). Weaker Carbotrace™680 signals were observed in the inner cortex and endodermis when compared with other tissues in all investigated root zones (Figs 2, 3). A ring of weak mixed-linkage glucan antibody labelling was observed between the stele and cortex in the meristem and early elongation zone (Figs 2, 3; BG1). Lower affinity for the inner cortex and endodermis when compared with other tissues was also shown by LM25 in the early elongation zone, indicating a lower detectable amount of galactoxyloglucan in these cells (Figs 2, 3; LM25). Mutant plants deficient in either mixed-linkage glucans or xyloglucans showed compromised growth, reduced tensile strength, and altered cellulose deposition (Cavalier et al., 2008; Vega-Sanchez et al., 2012; Smith-Moritz et al., 2015; Xiao et al., 2016). A reduced fluorescence signal does not necessary indicate a low level of corresponding polysaccharide. It may rather reflect the specificity of cell wall architecture in these cells. The inner cortex represents the stiffest tissue of maize root in the meristem and early elongation zone (Table 1). Further studies are required to understand whether there is any connection between the deposition of polymers based on the β-glucan backbone and mechanical properties of cell walls in the inner cortex of maize roots.
Xylans are abundant in the secondary cell walls of vascular plants and represent the link between cellulose and lignin (Kang et al., 2019). Despite their abundance in plant cell walls, to our knowledge, xylans have never been studied using any kind of mechanical testing in combination with immunological labelling in plant cells. Arabidopsis plants deficient in xylans and xylan glucuronosyl side chains (irx9, irx14, and triple gux1/2/3 mutants) had lower stem strength than wild-type plants (Lee et al., 2012; Chiniquy et al., 2013). However, this decreased stem strength is more likely to be the result of disturbed secondary cell wall deposition than the low amount of xylan per se. Grasses contain xylans in both primary and secondary cell walls but, surprisingly, no influence on plant biomechanics was reported for grass xylan synthesis mutants.
Distribution of arabinoxylans and glucuronoxylans (Figs 4A, B, 7; AX1 and LM28) did not correlate with changes in cell wall mechanical properties. Two antibodies that were expected to label rigid cell walls were LM12 (feruloylated polysaccharides) and LM11 (low-substituted xylans). Low-substituted xylans are typical components of secondary cell walls, which are generally stiffer than primary cell walls (Burgert and Dunlop, 2011; Cosgrove and Jarvis, 2012), and polysaccharide feruloylation is associated with growth cessation and cell wall stiffening in Festuca leaves, rice internodes, and Avena coleoptiles (Buanafina, 2009). Indeed, almost all tissues that became stiffer in the late elongation zone compared with the meristem (Table 1; AAAB- and BAAC-type tissues) were labelled with both antibodies (Figs 4C, D, 7; LM11 and LM12). Thus, the stiffening of cell walls in the late elongation zone can be partially due to the change in the glucuronoarabinoxylan structure through the removal of some side chains (formation of the LM11 epitope) and feruloyl-esterification of other side chains (formation of the LM12 epitope). However, the presence of both LM11 and LM12 epitopes in cell walls cannot be used as a rule of thumb to predict cell wall properties. Cell walls of the vascular parenchyma and pericycle were the stiffest in the late elongation zone (Table 1; Fig. 7) but had no LM11 epitope and were only moderately labelled by LM12. Intense LM12 labelling was characteristic of outer rhizodermis cell walls, which were among the softest at all tissues and developmental stages (Fig. 7). Relatively stiff meristematic cell walls were usually not labelled by LM12 and LM11 antibodies.
In the maize root, the presence of specific epitopes did not predict certain mechanical properties of the cell walls. This may be partially explained by limitations of the immunohistochemical approach. First, homogalacturonans and mixed-linkage glucans can mask other cell wall components (Marcus et al., 2010; Kozlova et al., 2014). Their abundance in maize root cell walls potentially affects immunohistological results. Secondly, polysaccharides possessing the same epitope can differ in their morphology, and their ability to interact with other cell wall components, and may have different functional significance (Gorshkova et al., 2013; Petrova et al., 2019). Thus, similar sets and intensities of labelling do not necessarily indicate the same set or quantity of polymers. Thirdly, immunolabelling and staining cannot provide full information about cell wall architecture because the same set of polymers may be arranged in different ways (Haas et al., 2020). The difference in cell wall thickness in different tissues potentially represents the fourth issue. Thick cell walls with a low content of a particular component may have the same fluorescence intensity as much thinner walls with a high content of this component. To overcome this issue, we normalized fluorescence intensity to cellulose abundance measured by Carbotrace™680 staining (Supplementary Fig. S5). However, no correspondence between the occurrence of different epitopes and cell wall elasticity was revealed. Therefore, immunochemistry alone should not be used as a proxy to predict mechanical properties of plant cell walls. Nevertheless, using this approach revealed that the transition from division to elongation and the overall decrease in the elastic modulus of cell walls between the meristem and early elongation zone of maize root was coupled with the reduction in the degree of homogalacturonan methylation. The inner cortex, which is the stiffest tissue of maize root in the meristem and early elongation zone, was characterized by low Carbotrace™680 staining and mixed-linkage glucan and galactoxyloglucan labelling. The stiffening of cell walls during late elongation was associated with the accumulation of low-substituted xylans and feruloyl-esterification of cell walls.
A hypothetical model for root anisotropic growth
Maize root is one of the well-studied examples of coordinated anisotropic growth. However, we believe that this is the first report of an in-depth mechanical characterization of cell walls in different tissues along the elongating root (Table 1). AFM-based examination of mechanical properties on vibratome-derived sections has several limitations (Kozlova et al., 2019). In part, these issues originate from the sample preparation procedure (possible solubilization of cell wall components in a vibratome water bath and AFM liquid cell). Partly they come from the assumptions that are used for calculations (the cell wall is considered as an isotropic material representing a half-space contacting an ideally manufactured and described sphere of a cantilever tip). Nonetheless, measured moduli of elasticity for different cell wall types and root zones were of the same order of magnitude as the values for primary plant cell walls which were obtained by other groups using AFM (Bidhendi and Geitmann, 2019). At the same time, these values were significantly lower than those obtained for plant tissues by other methods of mechanical testing (Bidhendi and Geitmann, 2019). The validity of our findings is supported by modelling. In contrast to some other models of plant cells and tissues (Wang et al., 2004; Pieczywek and Zdunek, 2014), our model was sensitive to Poisson’s ratio changes. However, maize root, modelled using the mechanical properties found in this study and Poisson’s ratio of 0.4, not only withstood the application of turgor pressure but also showed an elastic response to it, consistent with that observed in vivo (Figs 8, 9; Table 2).
In agreement with previous findings and the general hypothesis that a higher modulus corresponds to a lower extensibility, and vice versa, we report that cell walls of root tissues became softer as they transitioned from the meristematic region to the early elongation zone. Relatively low apparent elastic modulus values (suggesting higher extensibility) were retained in the active elongation zone. The late elongation zone in general was characterized by significantly increased modulus values as compared with the elongation zone (Table 1; BAAB and BAAC tissue types). Cell walls of the rhizodermis and exodermis belonged to the AAAB-type and did not show significant changes at the transition from meristem to elongation. This result corresponds to the report by Fernandes et al. (2012) who did not find changes in the elastic modulus of the Arabidopsis rhizodermis along the elongating root.
Along with changes in the elastic moduli that occur between developmental stages, radial gradients of stiffness were found within each zone (Table 1; Fig. 7). Tensiometry tests have shown that the central cylinder in the elongation zone of maize roots had a higher plasticity and lower ultimate strength than the cortex (Pritchard and Tomos, 1994; Chimungu et al., 2015). Thus, different tissues in roots do indeed have different mechanical properties, and the most rigid tissues belong to the cortical part of the root. The elastic modulus of growth-limiting tissue should be higher than those of other tissues (Tomos et al., 1989). At all stages of cell development, cell walls of the inner cortex displayed elastic moduli that were maximal or comparable with the maximal values among all other tissues. We propose that the inner cortex serves as a growth-limiting tissue in maize roots. Introducing measured moduli into the model showed that upon turgor pressure application, the inner cortex developed the highest stress in the Y-direction among all tissues (Fig. 9B). This is consistent with the inward bending observed in longitudinally split whole roots of maize (Burstrom, 1971; Sachs, 1882). Compression of the inner cortex equilibrates extension in softer tissues (e.g. pith and rhizodermis) in the X-direction, and thus supports the preferable direction of growth (Fig. 9C, D). Differences in elastic moduli between tissues may develop as a result of histogenesis in maize roots. Cells in the pith and outer cell layers (rhizodermis, exodermis, and outer cortex) finish their cell division and begin to elongate earlier than the cells of the inner cortex, endodermis, and pericycle (Baluška et al., 2001). Correspondingly, these latter tissues have smaller cells, which also may influence their extensibility (Bassel et al., 2014).
Earlier reports on anisotropic growth assumed that the direction of growth was determined by the difference in cell wall properties on different sides of individual cells (Tomos et al., 1989). Only periclinal cell walls were mechanically studied within this work, but we found a significant difference between radial and tangential cell walls in the endodermis of meristems and early elongation zones of the maize root (Table 1; IC/End and End/End). Casparian bands appear in the radial cell walls of maize root endodermis further up than the late elongation zone (Karahara et al., 2004; Shen et al., 2015). Polar distribution of different transporters, receptors, and other proteins in the endodermis already appears in the meristem (Alassimone et al., 2010, 2012), but characterizes stele-facing and soil-facing sides of the cell and not the adjacent sides. In addition to the high elastic modulus at early stages of development, radial cell walls of the endodermis were labelled by LM28 (glucuronoxylan), LM11 (xylans), and LM12 (feruloylated polymers) more intensively than tangential cell walls in the late elongation zone (Fig. 4). Another example of differential labelling was observed in metaxylem (Fig. 3). Anticlinal cell walls were characterized by more intensive Carbotrace™680 staining and LM25 (galactoxyloglucan) labelling than periclinal cell walls at all stages of root development (Fig. 3). Thus, there are several examples of mechanical and chemical anisotropy in cell walls within one maize root cell.
When measured values of elastic moduli and turgor pressure were used in the computer model, this resulted in a Y-strain distribution similar to those observed in vivo (Liang et al., 1997; Kozlova et al., 2012). Thus, we suggest that the softening and stiffening of cell walls along the root may determine the location and length of the elongating region. The differences in mechanical properties between tissues, and the stiff inner cortex in particular, may dictate the direction of growth and prevent root swelling. Distinct cell wall mechanical properties of different surfaces of the cell may also make a contribution to root growth anisotropy, but this has still to be verified.
Supplementary data
The following supplementary data are available at JXB online
Fig. S1. Atomic force microscopy on cross-sections of maize root cell walls.
Fig. S2. Removal of fluorescence background and signal intensity measurement on a micrograph of a maize cross-section.
Fig. S3. Scheme of root blocking in agarose for turgor pressure evaluation and elastic deformation assay.
Fig. S4. Scheme of maize root model and zone dimensions for a non-turgescent root used for modelling.
Fig. S5. Occurrence of different polysaccharide motifs normalized by cellulose staining using Carbotrace™680 in cell walls of maize with different mechanical properties.
Table S1. Antibodies and dye used in the current study.
Table S2. The relative intensity of labelling and staining of cell walls in maize root with various antibodies and the dye specific for mixed-linkage glucan, glucuronoarabinoxylan, galactoxyloglucan, and cellulose.
Table S3. The relative intensity of labelling of cell walls in maize root with various antibodies specific for pectins.
Acknowledgements
We would like to express our gratitude to Professor Paul Knox (University of Leeds, UK), Dr Marie-Christine Ralet (French National Institute for Agricultural Research, Nantes, France), and Professor Ewa Mellerowicz (Umeå Plant Science Centre, Umeå, Sweden) for kindly providing antibodies and dye used in this study. Maize seeds were provided by Dr Dmitry Suslov (Saint-Petersburg State University). Work was partially supported by the Russian Science Foundation (project number 18-14-00168, to AP and LK). Some of the work was performed with financial support from the government assignment for the Federal Research Center Kazan Scientific Center of Russian Academy of Sciences (TG).
Author contributions
Conceptualization, LK; supervision, LK and TG; funding acquisition, LK; project administration, LK; investigation, AP; visualization, AP; formal analysis, AP; writing—original draft preparation, LK and AP; writing—review and editing, LK, AP, and TG.
Data availability
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary data.
Comments