Suberin in Monocotyledonous Crop Plants: Structure and Function in Response to Abiotic Stresses

Abstract

Apoplastic barriers, formed by Casparian bands and suberin lamellae, represent important means of plant roots to adapt water and nutrient homeostasis to changing environmental conditions. To understand and evaluate the precise physiological role of suberin lamellae in water and nutrient transport characteristics, it is important to understand root anatomy, including main deposition sites and microstructure of suberin. Here we review suberin localization, chemistry, biosynthesis, and differential implementation in dependence of different abiotic stimuli in roots of monocotyledonous crop plants. Furthermore, we add results on the formation of suberized barriers in barley roots under nitrogen and phosphate deficiency, as well as ABA treatments. We conclude that the degree of suberin accumulation is essentially independent of absolute root length, while endodermal plasticity strongly and differentially responds to external environmental stimuli and thus affects plant physiology.

1 Introduction

Suberin and cutin are chemically closely related biopolymers forming lipophilic barrier structures for water, solutes, gases, and pathogens in specialized plant-environment interfaces (Schreiber 2010). Suberin is found in specialized tissue layers including periderms of shoots and the rhizo-, hypo-, and endodermis of roots, whereas cutin is restricted to the epidermis of above-ground organs (Ranathunge et al. 2011b). Since deposition sites of suberin and its chemical composition in plants have been very well characterized (Bernards 2002; Brundrett et al. 1991; Graça and Santos 2007; Kolattukudy and Espelie 1989; Schreiber et al. 1999), research of the past decade has focused on elucidating its three-dimensional structure, biosynthesis, genetic regulation, and functional implications. Oligomeric building blocks (Graça et al. 2015; Graça 2015), as well as many key enzymes and reaction steps (Vishwanath et al. 2015) orchestrated by a multitude of different transcription factors (Capote et al. 2018; Cohen et al. 2020; Kosma et al. 2014; Krishnamurthy et al. 2021), are known to date. Also, the general effect of suberization processes on transport physiology has been very well described (Barberon 2017; Kim et al. 2018; Ranathunge et al. 2017; Ranathunge and Schreiber 2011). All of this allowed for a slow and still ongoing transition and transfer from the suberin model species cork oak (Quercus suber L.), potato (Solanum tuberosum L.), and Arabidopsis thaliana to further important crop species employed in agriculture (Kreszies et al. 2018) and agroforestry (Rains et al. 2017).

Here we summarize and discuss studies on root suberization of monocotyledonous crop species in response to abiotic stresses (excess and deficiency) and possible effects on root transport physiology. We conclude that apoplastic barriers, formed by the Casparian bands and suberin lamellae, are important means for a plant root to adapt to changing environmental conditions. To understand and evaluate the precise physiological role of suberin lamellae development, it is important to consider root anatomy, including main deposition sites and microstructure of suberin, biosynthesis, genetic regulation, and transport physiology.

2 Suberin

2.1 Localization

Suberin is a plant-unique biopolymer deposited in various organs and tissues throughout the organism, displaying a multitude of physiological functions. It is probably best known as an integral component of cork oak (Quercus suber L.) periderm (Graça and Pereira 2000b), even providing this plant with its specific epithet. But it can also be found in tubers (Kolattukudy et al. 1975), seeds (Compagnon et al. 2009), wounding tissue (Yang and Bernards 2006), trichomes and glands (Kolattukudy 2001), and roots (Schreiber 2010). A noteworthy difference between suberin and cutin, its counterpart which is only to be found in above-ground tissues, is that suberin may be deposited into, whereas cutin is deposited onto cell walls (Andersen et al. 2015). Suberin is typically found incrusted in primary cell walls or deposited in lamellate structures between the plasma membrane and carbohydrate cell walls (Peterson and Cholewa 1998). In roots, suberin is frequently observed in endodermal and hypo-/exodermal cell layers, where it serves as a bidirectional apoplastic transport barrier for water and solutes as well as against pathogen invasion (Enstone et al. 2003). If a hypodermis exhibits Casparian bands, it is by definition referred to as exodermis (Perumalla and Peterson 1986). Cells of the endodermis may species-dependent originate from periclinal divisions of cortex-endodermis or epidermis-endodermis initial cells (Dolan et al. 1993; Pauluzzi et al. 2012) in the root apical meristem under influence of the SHORT-ROOT and SCARECROW transcription factors (Gallagher et al. 2004; Koizumi et al. 2012a, b). After cell divisions and maturations, endodermal development is characterized by three distinct stages, firstly described by Krömer (1903). Stage I: deposition of Casparian bands in transverse and radial cell walls. Stage II: accumulation of protoplast-enclosing suberin lamellae, and stage III: tertiary thickening of cell walls. Stage I differentiation is reported to be under the main control of the MYB36 transcription factor (Kamiya et al. 2015), whereas the master regulator of stage II yet remains elusive (Andersen et al. 2015). The third stage, however, may only be observed in certain species and is not ubiquitously found (Zeier and Schreiber 1998). These three stages are typically associated with the maturity of a given root segment and may develop earlier (i.e. closer to the root apex) or even later under stress conditions (Kreszies et al. 2019; Melino et al. 2021; Ranathunge et al. 2015; Stoláriková et al. 2012). Single cells in the endodermis lacking suberin lamellae are called passage cells, which have also been identified in the exodermis, and are hypothesized to aid in retaining transport abilities (Andersen et al. 2018; Peterson and Enstone 1996). In contrast to the endodermis, such a precise chronological order of differentiation has not been reported for the exodermis (Tylová et al. 2017). Furthermore, an exodermis may form constitutively as it has been observed in many rice (Oryza sativa L.) cultivars (Pedersen et al. 2020), optionally only due to stress as in maize (Zea mays L.) (Zimmermann et al. 2000), or mature even before the endodermis as in some wetland species (Soukup et al. 2002). Some other plant roots appear to be entirely devoid of exodermis formation (Nawrath et al. 2013). In many modern and some wild barley (Hordeum vulgare L.) cultivars no exodermis could be induced even under severe environmental stresses (Armand et al. 2019; Coffey et al. 2018; Kreszies et al. 2020a; Ranathunge et al. 2017), but not necessarily the whole species is incapable of forming this protective feature (Kreszies et al. 2020a; Reissinger et al. 2003). Nonetheless, an endodermis and hypo-/exodermis do share the most important common features such as developmental plasticity in response to biotic and abiotic stress, function as apoplastic barriers, and band plasmolysis due to Casparian bands (Enstone et al. 2003; Enstone and Peterson 1997; Hose et al. 2001; Karahara et al. 2004).

2.2 Composition and Structure

Patterns of suberin deposition are investigated by light microscopy using different types of (fluorescent) dyes, as Sudan red or fluorol yellow (Brundrett et al. 1991). In contrast, qualitative investigation of polymer building blocks and true quantification may best be achieved by precise analytical methods such as mass spectrometry and gas chromatography (Zeier and Schreiber 1997). Hints on the chemical nature of suberin already originate from the nineteenth century, where the renowned French chemist Michel Eugène Chevreul firstly used the term “subérine” to describe non-soluble compounds characteristic of cork (Chevreul 1815). This non-solubility, even in modern analyses, is used to specifically isolate suberized tissues for subsequent monomer extraction. Today, suberin is known as a biopolyester made of a polyaliphatic and a polyaromatic domain with glycerol as its backbone (Bernards 2002; Franke and Schreiber 2007; Graça 2015; Graça and Pereira 2000a; Rains et al. 2017).

The polyaliphatic fraction (aliphatics) consists of saturated or unsaturated long-chain mono- or bi-functional fatty acids, ω-hydroxy acids, α,ω-dicarboxylic acids, mid-chain oxygenated fatty acids, 2-hydroxy acids, and alcohols of species-dependent chain lengths, most commonly between C₁₆ and C₂₆ and in varying relative amounts (Holloway 1983; Kolattukudy 2002). In some species, such as sweet potato, even chain lengths of up to C₃₂ in the alcohol compound class have been reported (Bernards 2002; Kolattukudy et al. 1975). Chain length distribution as well as the abundance of individual monomers are not only species-specific but may also depend on the tissue of origin. It is documented, that bark suberin of a given species not necessarily has to share the same composition as suberin being deposited in roots (Matzke and Riederer 1991). ω-hydroxy acids and α,ω-dicarboxylic acids represent the most important aliphatic monomers (Graça 2015) and have therefore been termed as “suberin diagnostic”.

The polyaromatic domain (aromatics) is composed of monolignols and/or hydroxycinnamic acids, yet still subject to frequent debates (Bernards and Razem 2001; Rains et al. 2017). Due to solubilized monolignols, the aromatic fraction of suberin has historically been attributed to the lignin polymer (Kolattukudy 1980). However, modern approaches consider it to be distinctly different from lignin due to relatively higher amounts of hydroxycinnamic acids (Bernards 2002). The latter functional group is typically composed of ferulic and coumaric acid (Ranathunge et al. 2015), of which ferulic acid frequently represents <1% of the fractionalized suberin polymer (Graça 2010). Coumaric acid may also lack entirely, as in bark suberin of poplar (Rains et al. 2017) or roots of Arabidopsis (Molina et al. 2009). In many recent studies, released monolignols appear to be considered as co-solubilized, and only hydroxycinnamic acids are reported as core aromatic domain (Graça 2015; Kreszies et al. 2019; Ranathunge et al. 2017; Shiono et al. 2014b). Nonetheless, when interpreting aromatic compounds, great care needs to be taken due to several reasons. On the one hand, typically employed chemical reactions aimed at investigating suberin are best in depolymerizing the aliphatic fraction (Graça and Pereira 2000b; Ranathunge et al. 2011b; Zeier and Schreiber 1999), resulting in non-representative amounts of aromatics (Graça 2015) in some species. On the other hand, especially Poaceae species such as barley, maize, and rice are known to have high amounts of aromatic molecules bound to all cell walls (Carpita 1996; Chabbert et al. 1994), which may lead to a strong overestimation of truly suberin-specific aromatic compounds. For example, if comparing these monocotyledonous species to the dicot model plant Arabidopsis, their aromatic suberin fraction appears to be disproportionally large (Fleck et al. 2015; Franke et al. 2005; Ranathunge and Schreiber 2011; Schreiber et al. 2005b). The puzzle of how exactly aliphatics, aromatics, and glycerol are interlinked and three-dimensionally organized is still not solved and restricted to models (Graça 2015; Graça and Santos 2007; Ranathunge et al. 2011b) since the observation of the suberin polymer in its unaltered native form has not yet been achieved.

2.3 Biosynthesis

To obtain a fully functional suberin polymer several crucial reaction steps are needed: (i) biosynthesis of glycerol, hydroxycinnamic acids, and long-chain fatty acid precursors; (ii) elongation of precursors to very-long-chain fatty acids; (iii) α,ω-bifunctionalization by the introduction of additional hydroxy- and carboxy-groups and mid-chain oxygenation of fatty acids; (iv) reduction of fatty acids to alcohols; (v) conjugation of acyl chains to glycerol and ferulic acid; (vi) export of monomers and oligomers out of the cell into the apoplast; and (vii) polymerization to a three-dimensional structure. To achieve this, an array of highly orchestrated enzymes has to be recruited. In the following, we will focus on the most commonly described genes. Suberin C₁₆ and C₁₈ fatty acid precursors of the polyaliphatic domain are synthesized by the fatty acid synthase (FAS) complex in the plastids and subsequently transported to the endoplasmic reticulum (Li-Beisson et al. 2016). In the endoplasmic reticulum, most importantly fatty acid elongation (FAE) and various functionalization steps are taking place (Franke and Schreiber 2007). Elongation with C₂ units is achieved by β-ketoacyl-CoA synthases (KCS) (Franke et al. 2009; Lee et al. 2009; Serra et al. 2009b) and may, at various intermediate chain lengths, be succeeded by oxidation reactions to introduce additional hydroxy- and carboxy groups, resulting in the most characteristic suberin monomers ω-hydroxy acids and α,ω-dicarboxylic acids. Oxidations are carried out, depending on the chain length of their substrate, by members of the cytochrome P450 (CYP) monooxygenase/ω-hydroxylase enzyme family, two members of which are most renowned as HORST (CYP86A1) and RALPH (CYP86B1) in Arabidopsis suberin biosynthesis (Compagnon et al. 2009; Höfer et al. 2008). Alternatively, fatty acid precursors of any given chain length may be reduced by fatty acid reductases (FAR) to yield primary alcohols (Vishwanath et al. 2013) or just remain unmodified. Entirely independent of this, hydroxycinnamic acids of the polyaromatic fraction are synthesized in the phenylpropanoid pathway (Bernards 2002; Bernards and Razem 2001; Tsai et al. 2006). First esterification reactions even within the cell at the endoplasmic reticulum are carried out by glycerol 3-phosphate acyltransferases (GPAT) (Beisson et al. 2007) and aliphatic suberin feruloyl transferases (ASFT) (Molina et al. 2009) to conjugate acyl chains with glycerol or ferulic acid, respectively, before secretion of the suberin building blocks into the apoplast. Very little is known about the export process from the endoplasmic reticulum through the plasma membrane and even less about polymerization outside of the protoplast. ABC transporters have been characterized in suberin monomer secretion (Panikashvili et al. 2010; Shiono et al. 2014b; Yadav et al. 2014). Also, lipid transfer proteins (LTP) (Plett et al. 2016; Shiono et al. 2014a), as well as Golgi-mediated vesicle trafficking (Vishwanath et al. 2015) are speculated to participate in suberin transportation. Once the building blocks reached their destination, assembly to a functional polymer by esterification has to be executed. However, no candidate genes have yet been characterized by true experimental confirmation to be involved in this very crucial step of suberin biosynthesis (Ranathunge et al. 2011b).

2.4 Regulation of Suberin Biosynthesis

So far only limited knowledge is available about the phytohormonal and genetic regulation of the spatiotemporal tightly controlled suberization process. As discussed in the section environmental stimuli (Sect. 4), a multitude of external stimuli has to be sensed and translated to adapt the root physiology to the changing environmental conditions. Early studies investigating supplementation effects of abscisic acid (ABA) on suberization in potato tubers were able to show that fairly high concentrations of 100 µM ABA significantly increased suberin amounts and diffusion resistance (Cottle and Kolattukudy 1982; Soliday et al. 1978). Enhanced suberization was later also confirmed with roots of maize subjected to 10 µM ABA (Zeier 1998). In Arabidopsis, it could be demonstrated that ABA significantly enhances suberization processes, whereas ethylene appears to be involved in the delay of suberin lamellae development, even under non-stress conditions (Barberon et al. 2016). Delayed suberization, which simultaneously goes along with an increased number of passage cells (Ogden et al. 2018), was shown to also be mediated by auxin-influenced cytokinin signaling as its suppression in the root apical meristem resulted in increased numbers of passage cells (Andersen et al. 2018). This effect was surprisingly not to be antagonized by additional ABA treatment, suggesting that cytokinin determines the responsiveness of endodermal cells to ABA (Andersen et al. 2018). Based on co-expression studies, members of the MYB, NAC, and WRKY transcription factor gene families have been suggested to take part in orchestrating suberization processes (Ranathunge et al. 2011b), and as of today, at least nine independent genes of these families have been described in greater detail (Capote et al. 2018; Cohen et al. 2020; Kosma et al. 2014; Krishnamurthy et al. 2021; Lashbrooke et al. 2016; Legay et al. 2016; Mahmood et al. 2019; Verdaguer et al. 2016). Aside from this fast progress in elucidating suberin biosynthesis and regulation, still, no suberin mutants of barley and maize and only one for rice (Shiono et al. 2014b) have been reported, whereas most studies were done with Arabidopsis or potato (Bernards 2002; Franke et al. 2005; Lulai and Corsini 1998; Schreiber et al. 2005c; Serra et al. 2009a; Vogt et al. 1983; Yang and Bernards 2006). Future studies are needed to transfer the acquired knowledge on suberin biosynthesis and regulation from model to crop plants. In contrast, environmental stimuli (Sect. 4) have been widely described for both, model and crop plants. Therefore, we will focus on findings based on the monocotyledonous crop species barley, maize, and rice.

2.5 Function of Suberin

Suberin displays an array of important functions, which may depend on the site of deposition. It acts as a sealing agent in wounding tissue after physical injuries (Yang and Bernards 2006), protects barks against fire (Dantas and Pausas 2013), serves as a diffusion barrier for atmospheric gases in bundle sheaths (Mertz and Brutnell 2014), leads to abscission of specific plant organs (van Doorn and Stead 1997) and may even be employed by humans for insulation and several further industrial applications (Gandini et al. 2006; Graça 2015). In plant roots, it is well established for many species that suberization of the endodermis and/or exodermis continuously increases over the length of the root (i.e. its maturity) (Andersen et al. 2015; Kotula et al. 2009, 2017; Kreszies et al. 2019; Ranathunge et al. 2017; Schreiber et al. 1999). Differential development, being induced or delayed suberization, has been observed as a reaction to certain environmental stimuli, as will be thoroughly discussed later (Sect. 4, Tables 1, 2, 3, and 4). Due to these findings, it was concluded that suberin lamellae must play an important role in plant water and nutrient homeostasis. It does act as an apoplastic transport barrier that influences water conductivity and nutrient uptake dynamics, which has repeatedly been proven by the employment of suited measurement techniques such as the root pressure probe or ion bypass measurements (Krishnamurthy et al. 2011; Ranathunge et al. 2005; Zimmermann et al. 2000).

The question of whether the endodermis or the exodermis, if present, greater contributes to the apoplastic barrier properties, was investigated extensively with studies on rice roots. It could be shown that an increase of exodermal suberization did not significantly decrease water transport conductivities whereas the endodermis was the rate-limiting factor for water flow (Ranathunge et al. 2003; Ranathunge et al. 2011a). The endodermis, which is always present, fulfills a bi-directional function. Besides controlling the uptake of solutes, it must as well prevent leakage of solutes from the stele into the cortex (Barberon 2017; Enstone et al. 2003). The exodermis, if developed, will have the same function as the endodermis. In addition, due to its deposition in the hypodermis, suberin can also serve as a barrier against pathogens (Reissinger et al. 2003) and has been proven to represent a barrier against radial oxygen loss of many wetland species (De Simone et al. 2003; Kotula et al. 2009, 2017; Soukup et al. 2007). When rice plants are flooded, the exodermis forms a strong barrier against radial oxygen loss around the developing aerenchyma, which in combination greatly facilitate the diffusion of oxygen from the shoot to the growing root tip (Pedersen et al. 2020).

Even though one might conclude, as it has also been indicated by some studies (Kotula et al. 2009; Ranathunge et al. 2017; Zimmermann et al. 2000), that more suberin always yields a stronger transport barrier, this is not always the case. Not only the amount of suberin but probably also its microstructure needs to be considered (Kreszies et al. 2018). The findings of Ranathunge et al. (2003) indicated that despite a strong suberization of the rice exodermis, its water conductivity was still comparably high. This could be due to a porous structure allowing small water molecules to pass, whereas diffusion of larger oxygen molecules and NaCl ions are hindered (Schreiber et al. 2005b). This was later verified by sealing potential wall pores with different precipitates and particles which in turn significantly decreased the conductivity of the outer part of the root (Ranathunge et al. 2004, 2005). Comparable conclusions were drawn after investigating mutant lines of Arabidopsis. Even though enhanced suberin1 (esb1) mutants (Baxter et al. 2009) exhibit two-fold more suberin than their wildtypes, this ectopic suberin accumulation failed to significantly reduce water and NaCl permeabilities (Ranathunge and Schreiber 2011).

It is well accepted that especially the hydrophobic aliphatic domain conveys water and gas repellency in roots (Graça and Pereira 2000b; Hose et al. 2001; Ranathunge et al. 2011a; Shiono et al. 2014b; Zimmermann et al. 2000). Impregnation with associated waxes, especially in above-ground organs, can greatly enhance barrier properties, as it was shown for aerial roots of Monstera deliciosa and air-exposed potato tubers (Schreiber et al. 2005c; Vogt et al. 1983; Zeier and Schreiber 1998). However, little to no wax has yet been reported for isolated endodermal and exodermal cell walls from soil-grown roots (Schreiber et al. 1999). In turn, there is still speculation about the precise role of aromatic suberin compounds. It has been suggested that the aromatic monomers link the “core suberin” (i.e. being aliphatics) to the primary cell wall and lignified cell walls (Graça 2015). In addition, the aromatic fraction might contribute to mechanical stability and act as a barrier especially for pathogens (Bernards and Razem 2001; Lulai and Corsini 1998).

3 The Effect of Suberized Barriers on Water and Solute Transport

Three pathways for radial solute and water transport, summarized as composite transport model (CTM), have been described (Steudle et al. 1993; Steudle and Brinckmann 1989) and this was further expanded and refined in the following years (Kim et al. 2018; Ranathunge et al. 2017; Steudle 2000a, b; Steudle and Peterson 1998). The model suggests three theoretical pathways: (i) the apoplastic (cell wall) pathway, (ii) the symplastic (cellular) pathway, and (iii) the transcellular pathway, which represents a combination of the former two pathways (apo- and symplastic). Experimentally, the symplastic and the transcellular pathway cannot be measured separately to date (Steudle and Peterson 1998). Purely symplastic and purely apoplastic pathways would represent the two extremes of water and solute uptake. If water and solute transport would exclusively take place symplastically, molecules would enter the protoplast across the plasma membrane of the rhizodermis and travel through plasmodesmata from cell to cell, without ever directly crossing further membranes, before being exported into the xylem vessels (Maurel and Chrispeels 2001; Steudle 2000a). With a purely apoplastic pathway, water and solutes would diffuse across the porous cell wall continuum without ever crossing any cell membrane. However, latest at the endodermis further radial transport is controlled by the symplast via membrane-bound transporters and channels regulating which solutes are taken up into the central cylinder (Ranathunge et al. 2017). This sealing of the apoplast may only be interrupted at sites where lateral roots emerge, thus breaking the continuum, or at the very root tip where no functional Casparian bands and suberin lamellae have yet been developed (Kreszies et al. 2018; Steudle and Peterson 1998). In contrast, the transcellular pathway is characterized by constant vectorial influx and efflux processes (Geldner 2013).

Axial (i.e. longitudinal) movement may be neglected at this point since the resistance of dead xylem vessels is low (Steudle and Peterson 1998) and radial uptake has proven to be the rate-limiting step (Frensch and Steudle 1989). This is different for immature root tips where functional xylem has not yet formed (Ranathunge et al. 2017).

Sophisticated experimental approaches, for example, root pressure probes, exudation experiments, or pressure chambers, have to be applied for measuring and quantifying radial transport across roots (Miyamoto et al. 2001; Steudle et al. 1987; Suku et al. 2014). Thus, hydraulic conductivities (Lp_r, in m³ m⁻² s⁻¹ MPa⁻¹ or simplified m s⁻¹ MPa⁻¹), which are a measurement of water flow per unit surface area perpendicular to flow path and per unit biophysical driving force (Knipfer et al. 2020; Kreszies et al. 2018), as well as membrane permeabilities (P_sr, in m s⁻¹) and reflection coefficients (σ_sr, no dimension) of given solutes are obtained (Ranathunge et al. 2017). The latter two are inversely correlated (Steudle and Peterson 1998), which means that, for example, a high P_sr indicates a low σ_sr. The reflection coefficient is by definition a dimensionless value between zero and unity, where zero resembles non-selectivity for a solute and one (unity) indicates perfect semi-permeability of a given barrier, as would be the case in an ideal osmometer (Steudle and Brinckmann 1989). In roots, σ_sr of the apoplast is virtually zero whereas the semi-permeable cell-to-cell path may theoretically reach values close to unity (Kim et al. 2018; Steudle and Peterson 1998). The root hydraulic conductivity (Lp_r) may be dissected into hydrostatic Lp_r(HY) and osmotic Lp_r(OS) by considering the two driving forces hydrostatic (i.e. transpirational tension) and osmotic (i.e. osmolyte gradients) pressure, which are influencing the Lp_r (Steudle and Peterson 1998). Hydrostatic hydraulic conductivity reflects composite water flow along all three pathways in parallel (Steudle 2000b). In turn, osmotic pressure gradients may not impact the non-selective apoplastic route, which is why Lp_r(OS) solely resembles the combined cell-to-cell pathways (Kreszies et al. 2018). The ratio of both (Lp_r(HY):Lp_r(OS)) yields information of relative contributions. If it is close to one, as for Arabidopsis (Ranathunge and Schreiber 2011), a dominant cell-to-cell pathway is indicated, whereas a value substantially greater than one depicts a higher contribution of the apoplastic pathway (Steudle and Peterson 1998). In the case of rice, for example, ratios may easily reach values of up to 20 resembling a strong apoplastic contribution (Miyamoto et al. 2001). All of the above-described water and solute uptake processes are highly dependent on several variables, for example, plant species (Kreszies et al. 2020a; Schreiber et al. 2005b), plant age, developmental state (Ranathunge et al. 2017), formation of an exodermis (Zimmermann et al. 2000), or specific growth conditions (Kreszies et al. 2019).

A controversially discussed topic is the detailed composition and exact physiological role specifying Casparian bands and suberin lamellae as apoplastic barriers in water and nutrient relations of a plant, which has already been mentioned shortly after their discovery (Krömer 1903). Undeniably, primary cell walls exhibit pores with diameters being too big to selectively regulate water and low molecular weight solutes (Marschner 1995; Ranathunge et al. 2005, 2011b). Secondary modifications in the form of lignin or suberin, however, might be sufficient to shrink pore sizes to a degree that allows higher semi-permeability of solutes and potentially affect the small water molecules to a lower extent (Kreszies et al. 2019; Schreiber et al. 2005b). The contribution of Casparian bands in effectively sealing the apoplastic pathway at the endodermis is indisputable (Caspary 1865; Clarkson 1993; Peterson 1987). However, for Casparian bands, it was only shown experimentally that the apoplastic movement of large fluorescent dyes is blocked (Kamiya et al. 2015; Naseer et al. 2012; Peterson 1975). To our best knowledge, experiments proving that small water molecules are completely blocked by Casparian bands are still missing. In contrast for suberin lamellae, which are chemically hydrophobic, reductions of radial water transport under osmotic stress or radial oxygen loss under anaerobic conditions have been shown (Kotula et al. 2009, 2017; Kreszies et al. 2019). A further compelling example is the potato periderm, which is composed of suberin and completely lacks Casparian bands. Potato periderm has barrier properties comparable to a leaf cuticle. Thus all cell walls, even the primary cell walls, must obviously be suberized and encrusted with waxes. Otherwise, such efficient barrier properties could not be obtained (Schreiber et al. 2005c).

Moreover, the precise chemical composition of Casparian bands is still a matter of debate. A broad consensus is achieved in their structure being mainly made of lignin, but the exact contribution of suberin to their chemical properties remains yet elusive and might even be species-dependent (Soukup and Tylová 2018; Tylová et al. 2017). On one hand, analytical methods, as well as Raman scattering microscopy, showed that suberin aside from lignin was identified in endodermal Casparian bands of many species, even if no developmental stage II had been observed (Man et al. 2018; Schreiber et al. 1994; Schreiber 1996; Thomas et al. 2007; Zeier et al. 1999; Zeier and Schreiber 1998). On the other hand for Arabidopsis, reporter gene systems and expression studies indicated no involvement of suberin in endodermal Casparian band biosynthesis (Kamiya et al. 2015; Naseer et al. 2012). Yet additional endodermis-specific transcriptomic and chemical analyses are still missing to provide further reliable data in this controversy (Kreszies et al. 2019). Recently, direct analysis of Casparian bands via high-resolution Raman spectroscopy indicated the simultaneous deposition of both, lignin as well as suberin, in Casparian bands of maize (Man et al. 2018) and Chinese fir (Song et al. 2019). It can be summarized here, that neither Casparian bands nor suberin lamellae establish perfect barriers but rather convey increased resistances against diffusion of water as well as solutes (Barberon et al. 2016; Ranathunge et al. 2003, 2005; Schreiber et al. 2005b; Schreiber 2010; Steudle 2000a; Vogt et al. 1983), which in turn is highly dependent on the environmental conditions the plant is confronted with.

4 Environmental Stimuli

For suberin lamellae development, it has repeatedly been shown that biotic (Borg-Olivier and Monties 1993; Lulai and Corsini 1998; Ranathunge et al. 2008; Reissinger et al. 2003; Salas-González et al. 2021; Thomas et al. 2007) and abiotic (Tables 1, 2, 3, and 4; Enstone et al. 2003; Stoláriková et al. 2012) environmental stimuli are capable of differentially affecting suberization. Some may induce suberization, whereas others delay the development of suberin lamellae. Inoculation of Arabidopsis roots with microbiota was recently shown to convey enhanced performance under abiotic stress treatments (Salas-González et al. 2021). This greatly emphasizes the need for future investigations on potentially beneficial microbiome-root interactions and how they influence suberization processes of crop species in field trials. In root research, plants are typically grown under axenic gel-based or hydroponic conditions, as it allows precise control over treatments and provides easy access to the roots at harvest. Especially for agar plates, which are frequently employed to cultivate only several-day-old seedlings of Arabidopsis, illumination of roots should generally be avoided, but at the very least be considered as a potential additional stimulus (Baluška et al. 2009). Unconscious secondary outcomes might also be produced by the employment of multi-faceted stresses such as salinity, which not only exerts ion toxicity upon roots but also considerable osmotic effects. Studies comparing hydroponics with additional methods of cultivation (Table 1), such as growth in aeroponics or most naturally on soil, have proven that cultivation alone yields a stimulus strong enough to differentially affect suberization and root morphology (Krishnamurthy et al. 2009; Miyamoto et al. 2001; Ranathunge et al. 2015; Redjala et al. 2011; Zimmermann et al. 2000). Hydroponic growth generally seems to induce less suberin development than the other two methods of cultivation (Table 1). When comparing results, this should always be taken into account. Also, just unilateral subjection to stimuli was shown to have specific effects, as gel-grown maize roots developed asymmetrical suberization after treatment with cadmium and air exposure (Líška et al. 2016). It must be emphasized that also under control conditions even different parts of the same root along its developmental axis may exhibit distinct features and properties (Ranathunge et al. 2017). Lastly, genotypic differences between cultivars or wild types of the very same species might lead to different reactions (Kreszies et al. 2020a). To better compare specific developmental conditions, the percentage distance of roots measured from the apex should be chosen, as this ensures root segments to have the same age, irrespective of potentially varied lengths under environmental stress conditions (Ranathunge et al. 2015). Due to all of the above, a “detailed map” (see Fig. 7) over the length of the root should be created for the most reliable and conclusive interpretation of experimental results (Kreszies et al. 2018). Besides, such a map would also allow feasible and reliable comparisons to other species or cultivars/genotypes within the same species.

Table 1 Summary of studies reporting about effects of different cultivation methods and various treatments on root morphology, suberization reactions, and hydraulic properties of monocotyledonous crop species and Arabidopsis

The sensing of external conditions by plants is closely integrated into phytohormone signaling pathways as it has been mainly investigated in the dicotyledonous model plant Arabidopsis (Sect. 2.4). The phytohormone signaling network of Arabidopsis is highly reactive to nutrient availability (Table 1): manganese (Mn), iron (Fe), zinc (Zn) and phosphorus (P) deficiency lead to decreased suberization by affecting ethylene and cytokinin activity, whereas potassium (K) and sulfur (S) starvation, as well as sodium chloride (NaCl) addition, stimulated suberization by triggering ABA signaling in Arabidopsis (Andersen et al. 2018; Barberon et al. 2016). The exact sensing mechanism of nutrient availability appears to be still unknown (Barberon et al. 2016). However, partly due to the nature of thin and fragile Arabidopsis roots, adequate chemical quantification after nutrient stress exposure and analysis of resulting transport properties have not or rarely been performed. Data on various crop species may be taken as assistance to further elucidate the effects of specific environmental stimuli.

The following sections are focused on the developmental processes of rice, maize, and especially barley roots. Until now, most studies investigated seminal and/or adventitious main roots but explicitly not their laterals. However, studies by Tylová et al. (2017) and Knipfer et al. (2020) have emphasized that lateral roots must get detailed consideration in the future, and more research is needed to draw further conclusions on this often-overlooked part of the plant root system.

4.1 Water Deficiency and Osmotic Stress

Precisely controlled drought stress (i.e. water deficit) in hydroponic systems (Table 2), as opposed to water withdrawal on soil, is most often mimicked by different types of polyethylene glycol (PEG) addition to the medium (Michel 1983). This non-toxic and inert polymeric agent introduces osmotic stress by allowing the precise adjustment of the nutrient solution’s water potential (Ψ, in MPa). Investigating effects of a water potential of −1.07 MPa on maize roots found decreased root lengths but simultaneously increased amounts of suberin in the endo- and exodermis (Zeier 1998). For barley roots, this general setup was refined with PEG8000 by not only dissecting roots into three zones according to their developmental stage but also investigating stepwise decreased water potentials of −0.4, −0.8, and −1.2 MPa (Kreszies et al. 2019). Due to the fine nature of barley roots, separation into the central cylinder (endodermis) and outer part of the root (hypo-/exodermis) was not feasible. Nonetheless, since in the modern cultivar Scarlett no exodermis was ever observed, changes of suberin amounts could solely be attributed to specific effects in the endodermis. It was shown that root lengths decreased with decreasing water potential of the medium and aliphatic suberin amounts conversely increased, most remarkably in the root zones of 25–50 and 50–100%. These findings were supported by a significant upregulation of suberin biosynthesis genes. The aromatic fraction, in contrast, was found to be non-responsive to the imposed stresses. Interestingly, the onset of suberization remained at a similar distance of 25%, not resulting in a shift of suberization towards the root tip. Water and nutrient transport analyses of the −0.8 MPa treatment revealed a significant reduction in Lp_r(HY) with no effects on Lp_r(OS) and P_sr of NaCl if compared to control conditions (Kreszies et al. 2019). The addition of silicon (Si) at a water potential of −0.8 MPa was not able to show additional effects on root lengths or the degree of suberization in cultivar Scarlett (Kreszies et al. 2020b). These investigations were subsequently expanded to further barley cultivars and wild barley accessions (Kreszies et al. 2020a). Most findings for the cultivars were highly similar to that of the cultivar Scarlett, but especially the wild barley accessions reacted differently. In wild barley accessions, suberization was more restricted to the most basal root zone (50–100%) and also aromatic compounds did show significant increases. Furthermore, one of the wild accessions from Jordan was observed to exhibit a properly developed exodermis. Still, when comparing wild accessions with modern cultivars, the induction of core suberin genes was less pronounced, potentially reflecting the overall slightly lower aliphatic suberin amounts of wild accessions (Kreszies et al. 2020a).

Table 2 Summary of studies reporting about effects of osmotic stress, salinity, and ABA supplementation on root morphology, suberization reactions, and hydraulic properties of monocotyledonous crop species

4.2 Salt Stress

Salinity, which frequently goes along with drought conditions as salts (e.g. NaCl) accumulate in the soil (Table 2), is not only capable of exerting osmotic stress similar to PEG but may also impose the additional stress of severe ion toxicity on root organs, especially in the case of NaCl. Quantitative investigations on maize roots indicated that suberin amounts increased upon 100 mM NaCl treatment (Zeier 1998). Further and more elaborate studies were carried out on roots of rice with concentrations of 50, 100, and 200 mM NaCl (Krishnamurthy et al. 2009, 2011). These concentrations exhibit theoretical water potentials of −0.25, −0.5, and −1.0 MPa, respectively. Taken together, salt treatments were shown to increase the expression of suberin genes within only 30 min, but control levels were reached again after 4 h of exposure. This upregulated gene expression coincided with significantly enhanced suberization after 7 d of treatment, where aliphatics as well as aromatics accumulated in both the endo- and exodermis. Interestingly, the most pronounced changes of the endodermis were observed in the apical half (0–50%), whereas the exodermis reacted more severely in the basal half of the root (50–100%). It was further observed that a just two-day-long treatment with the highest 200 mM concentration was not sufficiently long to yield increased suberin amounts (Krishnamurthy et al. 2011). This shows that the formation of enhanced suberin lamellae needs to be considered as a long-term adaptation to abiotic stress. Pre-conditioning with lower concentrations of sodium chloride as well as supplementation of calcium (Ca) greatly improved the survival rate of rice plants. Subsequent investigations of water and solute transport after salt exposure showed a significant reduction in Lp_r(HY) and P_sr of NaCl, whereas Lp_r(OS) could due to methodological limitations not be determined (Krishnamurthy et al. 2011). This data may be complemented with findings of barley, where 100 mM of NaCl significantly reduced root lengths and, in histochemistry, comparably increased suberization especially in the endodermis of the root tip (Knipfer et al. 2020). Here, in addition to a decreased Lp_r(HY), the Lp_r(OS) was found to be negatively affected as well. Furthermore, the σ_sr of NaCl was close to 1.00 (Knipfer et al. 2020). This is in strong contrast to our studies on barley where we reported reflection coefficients for sodium chloride of 0.29 up to 0.69 (Kreszies et al. 2019, 2020a; Ranathunge et al. 2017). These differences may be explained by different experimental procedures, calculations, and points of view of whether roots behave as perfect osmometers or not (for detailed reviews see Kim et al. 2018; Kreszies et al. 2018).

4.3 Exogenous Abscisic Acid Treatment

Most of the environmental stress conditions presented here will most probably influence ABA homeostasis, as ABA is commonly described as the most important phytohormone regulating abiotic stress reactions (Aasamaa et al. 2002; Barberon 2017; Bauer et al. 2013; Chen et al. 2014; Kosma et al. 2009; Macková et al. 2013). As endogenous ABA was shown to be involved in wound-induced suberization of potato tubers (Lulai et al. 2008; Lulai and Suttle 2009), one might consider investigating its effects after artificial exogenous application. However, surprisingly very little is known for monocot crop species (Table 2). In maize, external application of 10 µM ABA resulted in significantly reduced root lengths and an increased suberin accumulation in endo- and hypo-/exodermis (Zeier 1998), but no subsequent effects on transport physiology were investigated. This was performed by Schraut et al. (2005), where a physiological dosage of 0.1 µM ABA significantly enhanced water flow (J_v) within minutes even after various long-term nutrient starvation treatments in maize. These short-term responses of root hydraulic conductivity might substantially be regulated by aquaporin activity (Kaneko et al. 2015) rather than suberization effects occurring after several days.

A proof of concept analysis of ABA application to roots of barley has been performed in our laboratory (see Fig. 1 and Supplementary for experimental details) to confirm results previously reported for Arabidopsis (Barberon et al. 2016) and maize roots (Zeier 1998). The average root length of 12 d old plants under control conditions in the ABA experiment was 19.3 ± 3.7 cm (Fig. 2). An addition of 10 µM ABA did not affect the root length (19.8 ± 3.5 cm) compared to its control. In contrast, subjection to 50 µM ABA for 6 d significantly reduced the average root length (14.2 ± 2.9 cm). Suberin lamellae development has been affected differentially in between the treatments (Fig. 3). Under control conditions, no suberization was visible until 25% of the root length (zone A), whereas patchy suberization was observed in zone B (25–50%) and developed into full suberization in zone C (50–100%) (Fig. 3a, e, i, m). Suberization of zone C did not seem to vary considerably in either treatment, since microscopy indicated consistent full suberization (Fig. 3a, b). However, if roots were subjected to ABA, staining with fluorol yellow revealed an increased suberization in zone A and B, because full suberization was reached already very close behind the root apex (Fig. 3j, n). Microscopic observations have only been carried out for the 50 µM ABA concentration, which severely affected root length (Fig. 3b, f, j, n). Under influence of ABA, no exodermis formation could be observed in barley seminal roots. Aliphatic and aromatic suberin monomer distribution of barley roots were exactly as previously described elsewhere (Kreszies et al. 2019; Ranathunge et al. 2017). The aliphatic fraction consisted of ω-hydroxy acids, α,ω-dicarboxylic acids, fatty acids, and alcohols of chain-lengths between C₁₆ and C₂₆. Aromatic suberin was composed of ferulic and coumaric acid. Furthermore, relative amounts of individual monomers, as well as substance class composition did not differ significantly in between growth conditions. Due to these abovementioned reasons, only sums of aliphatic (Fig. 4a) and aromatic (Fig. 4b) suberin are shown. It is evident, that with increasing maturity of the root (i.e. from tip to base), the suberin content of both fractions steadily increased, which is perfectly in line with the histochemical observations (Fig. 3). The addition of both 10 and 50 µM ABA considerably increased the total aliphatic suberin amount, most remarkably in zone A of 50 µM ABA treatment and zone B of both ABA applications. Zone C was not affected much, with only a slight increase by the addition of 50 µM ABA. It seems, that between 5 to 8 µg cm⁻² a plateau of aliphatic suberin accumulation has been reached. The aromatic fraction behaved accordingly, but with less steep increases of amounts if compared to aliphatics, especially in zone A and B.

Fig. 1

General hydroponic cultivation procedure. After 3 d of germination, seedlings were transferred into (modified) ½ Hoagland nutrient solution. In control and deficiency (-nitrogen, -phosphorus) treatments, seedlings were left to grow for further 9 d. In excess ABA treatments, ABA was added to the nutrient solution at day 6, and plants were grown for another 6 d to ensure comparability to previous osmotic stress studies (Kreszies et al. 2019, 2020a, b). All plants were harvested at 12 d of age

Fig. 2

Seminal root lengths of 12 d old barley plants grown in control or ABA supplemented (+10/50 µM ABA) conditions. The number beside each box (n = 35–45 individual roots) represents the mean. Significant differences at p ≤ 0.05 based on t-tests are indicated by asterisks

Fig. 3

Development of endodermal suberin lamellae in barley seminal roots. Suberin lamellae are indicated by yellow fluorescence through staining with fluorol yellow 088. Apical zone A (0–25%) and intermediate zone B (25–50%) are depicted, as suberization in distances of >50% (basal zone C) was always fully developed. Under control conditions, suberization started at approx. 25%, developed in a patchy manner and became fully suberized at 50% (a, e, i, m). 50 µM ABA (+ABA) and nitrogen deficiency (−N) treatment showed full suberization already at 12.5% distance (n, o), whereas phosphorus starvation (−P) led to similar suberin lamellae development as under control conditions. Size bars = 50 µm

Fig. 4

Total amounts of aliphatic (a) and aromatic (b) suberin components under control ABA supplemented (+10/50 µM ABA) conditions. Seminal roots were divided into three zones: apical zone A (0–25%), intermediate zone B (25–50%), and basal zone C (50–100%). Bars represent means ± standard deviation of n = 2 replicates. Due to the lack of a third biological replication, no statistical tests were performed

ABA treatment did not exert osmotic stress on roots, which means that observable effects are solely due to the excess conditions employed, without having to consider secondary effects. Since no exodermis formation could be observed, quantitative changes in suberin amounts can be attributed exclusively to differences in endodermal development. Root lengths measured and suberin lamellae deposition observed in control conditions fit very well to that reported previously for the very same growth conditions (same plant age and climate chamber) for barley cultivar Scarlett (Kreszies et al. 2019, 2020b). However, the newly introduced environmental stimulus employed in this study provoked significantly different reactions regarding root morphology and suberin deposition. Barley seminal root lengths were found to be significantly decreased only at concentrations higher than 10 µM ABA (Fig. 2). Differently in maize, this dose of 10 µM was sufficient to severely affect the average lengths of primary roots (Zeier 1998). The suberization was strongly enhanced (Fig. 3f, j, n). Already at the very root tip, full suberization had been reached, which continuously persisted to the root base. Fluorol yellow signals for ectopic suberin could not be observed in cortex cells of any distance, as it had previously been reported for roots of Arabidopsis (Barberon et al. 2016). Roots also had no exodermis, which indicates that the barley cultivar Scarlett might just not be able to additionally suberize the hypodermal layer, even in the most severe environmental stress conditions (Kreszies et al. 2019, 2020b). Chemical analysis confirmed the strong suberization induced by ABA (Fig. 4), as it has already been reported before in different species (Barberon et al. 2016; Zeier 1998). It is noteworthy that root zones A and B were most enhanced with suberin deposition and amounts were similar to root zone C (Fig. 4). This was the strongest reaction of this cultivar ever observed yet (Kreszies et al. 2019, 2020b). Total aliphatic suberin amounts seem to approach a threshold value, which is also found in zone C of control roots. This was not found for the aromatic fraction. Different from nitrogen and phosphorus deficiency (Sect. 4.4), synthesis of the aliphatic fraction seemed to be induced more intensely than aromatic compounds. The effect of ABA-induced suberization will probably decrease root hydraulic properties as shown earlier in various species in response to abiotic stress (Armand et al. 2019; Kreszies et al. 2019; Krishnamurthy et al. 2011; Ranathunge et al. 2015; Zimmermann et al. 2000).

4.4 Nitrogen, Phosphorus, and Potassium Excess and Deficiency

Comparatively, extensive research has been performed investigating the effects of nitrogen (N) excess and deficiency (Table 3). Nitrogen is typically available to the plant in the form of nitrate (NO₃⁻) or, less frequently, ammonium (NH₄⁺) (Bang et al. 2021). Phenotypic reactions induced by excess or deficiency conditions may not only depend on the nitrogen source but also the species exposed (Armand et al. 2019; Melino et al. 2021; Ranathunge et al. 2015). Non-optimal dosages of NH₄⁺ differentially affected root morphology and water and solute permeability of rice. Low concentrations and deficiencies resulted in increased root lengths and decreased amounts of aliphatics and aromatics, whereas excess concentrations led to shorter roots and enhanced accumulation of suberin in both endo- and exodermis, at the root tip as well as in the basal part (Ranathunge et al. 2015). These findings were perfectly reflected in qPCR analyses: suberin gene expression was down-regulated under low ammonium conditions but increased in high concentrations. Effects on root water and solute transport appeared well correlated in this case. Accordingly, reduced amounts of suberin resulted in increased Lp_r (both osmotic and hydrostatic) and P_sr of NaCl. Increased suberin concentration was able to decrease P_sr, even though the Lp_r remained unchanged (Ranathunge et al. 2015). In strong contrast to these findings on ammonium are the effects of low concentrations or absence of nitrate with maize and barley, which independently and repeatedly led to opposite findings (Armand et al. 2019; Melino et al. 2021; Plett et al. 2016; Schraut et al. 2005). First indications for this might be extrapolated from measurements of radial water flows (J_v) that were found to be significantly reduced after treating maize plants with nitrate deficiency (Schraut et al. 2005), indicating a potentially increased deposition of suberin. Later, maize transcriptomic data revealed co-expression of lipid metabolism genes that were attributed to a suberization response to nitrogen supply and demand (Plett et al. 2016). Two studies on nitrate deficiency that investigated several barley cultivars observed that roots were also longer under starvation conditions compared to control (Armand et al. 2019; Melino et al. 2021). However, compared to rice, the opposite behavior regarding suberization and transport properties was found. In barley, based on microscopy and exudation experiments, suberization of the endodermis was increased at 25 and 50% distance from the tip, and Lp_r(OS) was significantly reduced (Armand et al. 2019). The increased suberization due to nitrogen limitation was also quantitatively confirmed recently (Melino et al. 2021). Most significant increases of both the aliphatic and aromatic fraction of suberin were found in the endodermis at 25–50 or 50–75%, depending on the cultivar investigated. Additionally, the expression of suberin genes was significantly upregulated in all cultivars (Melino et al. 2021).

Table 3 Summary of studies reporting about effects of plant mineral nutrients on root morphology, suberization reactions, and hydraulic properties of monocotyledonous crop species

Surprisingly, less is known about the effect of phosphorus (P) and potassium (K) deficiency on suberization and subsequent effects on transport processes of monocot crop plants, even though they are two macronutrients in plant mineral nutrition (Table 3). In Arabidopsis, potassium deficiency resulted in enhanced suberization (Barberon et al. 2016). However, phosphorus starvation has not been investigated in more detail. Maize grown in phosphorus- or potassium-limited conditions were found to have an increased radial water flow (J_v) (Schraut et al. 2005), which might point towards reduced suberin amounts. Conversely, for barley roots, an increased degree of suberization at 25 and 50% distance after phosphorus deficiency treatment was histochemically observed. This coincided with significantly decreased rates of Lp_r(OS) (Armand et al. 2019). Also, barley roots were found to be decreased in length, but due to thicker roots under phosphorus depletion, the combined root surface area remained stable. The root:shoot surface area ratio had increased significantly (Armand et al. 2019). A comparative study but with potassium starvation yielded entirely different findings (Coffey et al. 2018). Root lengths had significantly increased, but no changes in suberization could be observed via microscopy. Nonetheless, the Lp_r(OS) was found to be significantly reduced, which was argued to be due to aquaporin activity to counterbalance the increased surface area of roots (Coffey et al. 2018). Unfortunately, we cannot conclude the final quantitative suberin amounts from the studies mentioned about phosphorous and potassium deficiency at the moment. Thus, we add and discuss new data about phosphorus deficiency. In general, more suberin observed by microscopy such as with staining via fluorol yellow can be supported by findings of enhanced suberin amounts via analytical methods (Kreszies et al. 2019). However, all microscopy staining methods have the disadvantage that there is a certain threshold needed to bind and show a signal and no fine differences, for example, for a reduced suberin amount besides more passage cells can be detected (Kreszies et al. 2020b).

To further complement the set of intensively studied abiotic stresses in barley roots (Tables 1, 2, 3, and 4), additional histochemical, as well as chemical analyses (see Fig. 1 and Supplementary for experimental details), have been carried out on nitrogen and phosphorus deficiency in our laboratory. Control roots of 12 d old plants in the deficiency experiment (Fig. 5) were similar in length (21.4 ± 4.5 cm) to the control roots of the ABA treatment (19.3 ± 3.7 cm). In contrast, root lengths of both deficiency treatments were significantly longer than their control. Roots subjected to nitrogen deficiency (35.7 ± 10.6 cm) were additionally considerably longer than those of phosphorus deficiency (30.0 ± 6.2 cm). Microscopic investigations of Casparian bands did not show any differences between control and treatments (data not shown). Starting in the middle of zone A, all roots displayed continuous Casparian bands in the radial endodermal cell walls and no Casparian bands or suberin lamellae have been observed in the hypodermis. Thus, no exodermis has been formed. Full suberization of zone C was consistent throughout all treatments (Fig. 3a, c, d). It is evident, that nitrogen deficiency (Fig. 3c, g, k, o) induced suberization patterns very similar to that of the ABA treatment (Fig. 3b, f, j, n). Treatment with phosphorus deficiency, in contrast, did not seem to induce the development of suberin lamellae (Fig. 3l, p) and suberin visualizations in zone B appeared to be comparable to that of the control. Qualitative and relative suberin composition was consistent with previously published data (Kreszies et al. 2019; Ranathunge et al. 2017) and also in between growth conditions, which is why only sums of aliphatics and aromatics are shown (Fig. 6a, b). Suberin amounts of both fractions of control conditions are in agreement with the control in the ABA treatment. Comparing aliphatic amounts of the nitrogen deficiency treatment to the control, significant increases could be observed in zone A and zone B. Zone C showed a slight trend of increased suberization, which was not confirmed statistically. Similar trends were observed for aromatics, however, none of these differences were significant. In strong contrast to this, phosphorus deficiency treatment resulted in significantly decreased aliphatic suberin amounts in zone A and B and the aromatic fraction, again, insignificantly followed this trend. Same as with nitrogen starvation, no change of aliphatic suberin could be observed in zone C. All mentioned observations are supported by the histochemical investigation (Fig. 3).

Fig. 5

Seminal root lengths of 12 d old barley plants grown in control or nutrient deficiency (-nitrogen, -phosphorus) conditions. The number beside each box (n = 40–66 individual roots) represents the mean. Significant differences at p ≤ 0.05 based on t-tests are indicated by asterisks

Fig. 6

Total amounts of aliphatic (a) and aromatic (b) suberin components under control or nutrient deficiency (-nitrogen, -phosphorus) conditions. Seminal roots were divided into three zones: apical zone A (0–25%), intermediate zone B (25–50%), and basal zone C (50–100%). Bars represent means ± standard deviation of n = 3 biological replicates. Significant differences at p ≤ 0.05 based on t-tests are indicated by asterisks

None of the deficiency stimuli of this study exerted osmotic stress on roots, which means that observable effects are solely due to the deficiency conditions employed, and no secondary effects in this regard need to be considered. Quantitative changes in suberin amounts can be attributed exclusively to differences in endodermal development, since no exodermis formation could be observed. Root lengths and Casparian band and suberin lamellae deposition in the control conditions fit very well to that reported previously for barley cultivar Scarlett (Kreszies et al. 2019, 2020b) cultivated under the very same growth conditions (same plant age and climate chamber). However, the newly introduced nutrient deficiency stimuli employed in this study provoked significantly different reactions regarding root morphology and suberin deposition. Upon nitrogen starvation, roots were significantly longer (Fig. 5), which is in agreement with nitrate as well as ammonium deficiency treatments previously investigated with rice, maize, and barley (Armand et al. 2019; Melino et al. 2021; Plett et al. 2016; Ranathunge et al. 2015). Low nitrogen conditions also significantly enhanced aliphatic suberin amounts in zone A and zone B (Fig. 6a), fitting very well to the histochemical observations indicating earlier and stronger suberization (Fig. 3g, k, o). Such enhanced suberization of the root tip has only been reported via microscopy in salt-stressed barley (Knipfer et al. 2020), which is a combination of osmotic and ionic stress. In contrast, under severe osmotic stress zone A remained entirely unsuberized, while in zone B and C the suberin lamellae were enhanced (Kreszies et al. 2019). Under nitrogen depletion, there was not only a shift of onset of suberization to even before 12.5% distance. Additionally, the endodermis was fully suberized near the root tip (Fig. 3o). This increased suberin accumulation resulting from NO₃⁻ limitation is in line with other studies on barley (Armand et al. 2019; Melino et al. 2021) and also indicated on the gene expression level for maize (Plett et al. 2016), but conflicting with opposite findings on NH₄⁺ in rice (Ranathunge et al. 2015) and NO₃⁻ in castor bean (Schreiber et al. 2005a). These reported differences could be attributed to the species investigated; especially dicotyledonous plants might react differently as monocots. In the case of rice, low dosages but not an entire deficiency of ammonium have been investigated (Ranathunge et al. 2015), whereas most other studies mentioned focused on nitrate reduction, and ammonium was rarely even supplemented in the hydroponic nutrient solution. As was hypothesized for rice and castor bean, a reduced amount of suberin lamellae could help to maintain a high uptake of the lacking essential nutrient, but this may not be valid in the case observed here. As the endodermis serves as a bidirectional barrier (Enstone et al. 2003), one might imagine that increased suberization in turn also aids in retaining nitrogen, which can still be taken up via high-affinity nitrate transporters into the root (Melino et al. 2021). Endodermal suberization appears to be highly dependent on plant species and precise environmental conditions. The fact that no changes in suberin contents took place in zone C despite strong reactions in zone A and B may indicate that a certain threshold value of suberin for its physiological function has been reached. In the mature root part of zone C the endodermis needs to be already completely suberized because water and solute transport in this root region is mainly longitudinal via the xylem to the shoot (Ranathunge et al. 2017). This would be supported by the finding that suberin amounts of zone B approach, but never surpass, those of the most developed zone C. Still, the increased suberin amounts due to changes in nitrogen status might lead to physiologically important decreased water and solute transport properties (Armand et al. 2019; Ranathunge et al. 2015).

Phosphorus deficiency, which similarly to nitrogen starvation resulted in increased root lengths (Fig. 5), contrarily affected root suberization. Especially in zone A and B aliphatic suberin amounts were significantly reduced (Fig. 6a), which was also reflected in fluorol yellow staining of suberin lamellae (Fig. 3h, l, p). Identical to nitrogen limitation, zone C was not significantly affected, and the aromatic suberin fraction exhibited the same yet insignificant trend of the aliphatic domain. These findings are conflicting with microscopy-based reports (Armand et al. 2019) where decreased root lengths, increased root:shoot surface areas, and an increased suberization of the endodermis, most remarkably at 25 and 50% distance upon phosphorus deficiency treatment with barley were found. This resulted in significantly decreased osmotic hydraulic conductivity (Armand et al. 2019). In contrast, Andersen et al. (2018) histochemically identified low phosphorus to significantly reduce endodermal suberization in Arabidopsis, and Schraut et al. (2005) found phosphorus starvation to not statistically affect, if not even slightly increase, water flow (J_v) in maize roots. However, quantitative chemical suberin analysis, as provided here for the first time in the context of phosphorous deficiency, should be more specific than solely microscopic investigations (Kreszies et al. 2020b).

4.5 Heavy Metal Accumulation

Most studies to date investigating the effects of heavy metal accumulation in crop plants have been based on histochemical investigations (partly reviewed in Kreszies et al. 2020b), and very little is known about quantitative effects and root transport properties (Table 3). The studies have in common, that exposure of roots to heavy metals such as cadmium (Cd) always reduced the root length and enhanced the development of suberin lamellae (Líška et al. 2016; Lukačová et al. 2013; Redjala et al. 2011; Vaculík et al. 2009, 2012). For example, Líška et al. (2016) were able to show that gel-grown maize plants exhibited unilateral suberization of the endo- and exodermis after unilateral treatment with 50 µM cadmium, which points towards a highly elaborate sensing and reaction mechanism. By using 1 µM radiolabeled CdCl₂, Redjala et al. (2011) reported that growth in hydroponics, which by microscopy was found to induce lesser suberin deposition as aeroponic cultivation, resulted in increased uptake of cadmium if compared to aeroponics. This indicates increased membrane permeability towards heavy metal ions induced just by the method of cultivation (Redjala et al. 2011). The only study known to deliver quantitative information is that of Zeier (1998), who found 100 µM of CdCl₂ to significantly increase the suberin content in the endodermis and the hypo-/exodermis of maize, clearly confirming the notion of previously mentioned histochemical analyses.

4.6 Silicon Fertilization

Silicon (Si) supplementation has been reported to differentially influence the deposition of suberin lamellae in maize and rice but was typically only based on microscopical observations. Based on these, some studies reported enhanced suberization (Fleck et al. 2011, 2015; Lukačová et al. 2013), whereas others found no effect or even reduced suberin lamellae development (Vaculík et al. 2009, 2012). Very interestingly, only a few chemical analyses were carried out for silicon addition to maize and rice roots. In these cases, no effects on aliphatic suberin could be observed, but the aromatic fraction was even significantly decreased sometimes (Fleck et al. 2015; Hinrichs et al. 2017). In barley, slightly increased root lengths upon silicon treatment, but no significant effects on suberin amounts by analytical methods were observed. The addition of silicon in osmotic stress conditions (−0.8 MPa induced by PEG8000) (Table 2) did not affect root lengths or the degree of suberization if compared to a PEG8000 treatment without silicon (Kreszies et al. 2020b). Contradictory observations of histochemistry obtained by many studies might be attributed to possible formations of silica aggregates that could either be able to interfere with the binding of fluorol yellow stain or lead to quenching of the fluorol yellow signals. This emphasizes the importance of combining qualitative microscopy with quantitative chemical analyses (Kreszies et al. 2020b). Transport properties after silicon application have been investigated in sorghum roots with and without osmotic and salt stress (Liu et al. 2014, 2015). In both studies it was found, that silicon treatment alone did not influence root hydraulic conductivity. However, the supplementation of silicon was able to significantly alleviate reductions in root hydraulic conductivity that are normally induced by osmotic and salt stress. These effects were attributed to enhanced aquaporin gene expression induced by silicon under stress conditions (Liu et al. 2014, 2015).

4.7 Hypoxia

The effects of oxygen deprivation are best described with rice plants (Table 4). Oxygen deficiency led to increased suberin amounts in the exodermis, both of the aliphatic and aromatic fractions, and in parallel radial oxygen was decreased starting from 20 mm behind the root tip (Kotula et al. 2009). Ranathunge et al. (2011a) confirmed these findings of effects on root morphology and suberin content of the exodermis, but also investigated changes in the endodermis as well as water and nutrient transport properties. They were able to show, that in addition to the exodermis also the endodermis is reinforced by suberin deposition under stagnant conditions. This did not correlate with decreased hydraulic conductivity (Lp_r(HY) as well as Lp_r(OS)). However, solute permeability of NaCl was significantly reduced, which was attributed to resulting specific pore sizes in the suberin lamellae, which would be capable of filtering Na⁺ ions, but not water molecules (Ranathunge et al. 2011a). The fact that oxygen leakage through the cortex (Kotula et al. 2009) but not water transport was reduced by increased suberization was argued to be caused by the differential pathways employed by dissolved oxygen and water: oxygen travels in a diffusional manner, whereas water moves in hydrostatic bulk flow (Ranathunge et al. 2011a). Also in rice, the first mutant- and microdissection-based genetic evidence for a barrier formation in the hypo-/exodermis has been reported (Shiono et al. 2014a, b). By using permeability tests, reduced culm number1 (rcn1) mutants of rice, which were defective in an ABC transporter gene, were shown to be incapable of forming an efficient exodermal barrier under stagnant conditions. Interestingly, the endodermal development was similar to the wildtype and it represented a barrier to apoplastic tracers. Suberin biosynthesis genes were found to be most upregulated near the root tip, but no effect on lignin-associated genes could be observed. Quantitative suberin evaluation indicated that the absence of an exodermal barrier was occurring in parallel with a significant reduction of the aliphatic suberin fraction. However, this decrease of aliphatics was accompanied by an increase of the aromatic suberin amount, which in turn was not able to compensate for the lost barrier properties (Shiono et al. 2014a, b).

Table 4 Summary of studies reporting about effects of hypoxia on root morphology, suberization reactions, and hydraulic properties of monocotyledonous crop species

Another study incorporating chemical analyses based on a further monocotyledonous plant was carried out with two accessions of Hordeum marinum, a close relative of H. vulgare growing close to sea water (Kotula et al. 2017). Their findings were similar to that observed in rice plants and showed reduced root lengths and increased amounts of aliphatic suberin upon hypoxia. Microscopy indicated the development of an exodermis. This newly formed barrier resulted in significantly reduced radial oxygen losses in the accession that was shown to have the most pronounced enhancement of the exodermis upon oxygen deficiency (Kotula et al. 2017). Colmer et al. (2019) tried to identify molecules that could be involved in the perception of low oxygen conditions. They focused on small organic acids that are produced by anaerobic microorganisms upon hypoxia, being acetic, propionic, butyric, and hexanoic acid. Indeed, most of the acids in a specific concentration were able to significantly decrease root lengths and radial oxygen loss. Histochemical analysis, however, was not able to identify considerable changes in suberization in endodermis as well as hypo-/exodermis. Gene expression studies indicated an upregulation of suberin genes after treatment with propionic and butyric acid, which could potentially be responsible for providing the increased barrier properties to oxygen diffusion (Colmer et al. 2019). In a recent review focusing on the effect of low soil oxygen on root morphology and anatomy of maize, wheat, and rice it was summarized that an enhanced suberin formation under waterlogged conditions is always observed (Pedersen et al. 2020).

5 Conclusion

Responses in development and suberization of barley seminal roots (cv. Scarlett) towards different environmental stress factors are highly variable (Fig. 7). In response to osmotic stress and ABA treatment, the aliphatic suberin fraction exhibited significant increases, whereas the aromatics did show no or only weak increases (this study, Kreszies et al. 2019, 2020b). Especially root zone B (25–50%) showed the most intens responses towards environmental stimuli (Tables 1, 2, 3, and 4). Some stimuli specifically also triggered reactions in the most apical root segment (zone A; i.e. ABA and nitrogen deficiency, this study), which indicates the high plasticity of roots adapting their endodermal development to the variable environmental stress factors (Fig. 7). The weakest responses were observed in the basal root parts (50–100% or zone C), where suberization was already very high under control conditions. It is also obvious that barley cv. Scarlett seems to be incapable of forming an exodermis in response to any of these stress conditions (Fig. 7). This, however, may only be concluded for cv. Scarlett, as other genotypes (e.g. wild barley) or barley species (Kotula et al. 2017; Kreszies et al. 2020a; Reissinger et al. 2003) have been reported to form an exodermis as a reaction to biotic as well as abiotic stimuli. We conclude, that the degree of suberin accumulation is essentially independent of absolute root length, while it strongly and differentially responds to external environmental stimuli (Fig. 7).

Fig. 7

Schematic representation of various abiotic environmental stimuli on seminal root development of barley cv. Scarlett. Under control conditions, barley seminal roots show in the youngest root zone from 0–25% relative root length only Casparian bands and no suberin lamellae. At 25–50% follows patchy suberin lamellae including passage cells. From approximately 50% relative length to the base of the seminal root the whole endodermis is completely suberized. In response to ABA, osmotic stress, and osmotic stress with silicon supplementation barley seminal root lengths are decreased and the fully suberized zone is shifted more towards the tip region (red arrow) because passage cells get suberized. Thus, under osmotic stress with and without silicon supplementation the patchy suberin lamellae root segment gets smaller, while it is completely missing after the addition of ABA. Furthermore, ABA treatment resulted in an earlier onset of suberization at around 12.5% relative distance. In response to supplementation with silicon or under nitrogen or phosphate deficiency barley seminal roots are significantly longer compared to the control. However, there was no effect on the suberization pattern by additional silicon supplementation. Under nitrogen deficiency, suberization is enhanced (red arrow) along the whole root similar to the ABA treatment. In contrast under phosphate deficiency, suberin amounts were reduced compared to the control (green dotted lines) in the younger half of the root. Data of this study was combined with those of Kreszies et al. (2019, 2020a, b). Only main roots and no lateral roots are shown for simplification. Red dots indicate Casparian bands, yellow lines indicate suberin lamellae (SL). +ABA, abscisic acid addition; −Ψ, osmotic stress by PEG8000; −Ψ & +Si, osmotic stress with silicon supplementation; +Si, silicon supplementation; −N, nitrogen deficiency; −P, phosphorus deficiency