Hormonal regulation of root hair growth and responses to the environment in Arabidopsis

This review presents a molecular framework for how phytohormones regulate normal root hair development and how this is affected by changes in the rhizosphere, to enable greater understanding of the specialized functions of root hairs and their developmental and environmental plasticity.


Introduction
Roots provide structural anchorage and access to water and nutrients, which is vital to plant survival. In addition, they form the site of symbiotic interactions with soil-living microorganisms (Haling et al., 2013;Grierson et al., 2014). As roots encounter heterogeneous soils in which nutrients and water are not equally distributed, their root system architecture is constantly modified. Dynamic modification of root hair growth, length, density, and morphology allows the plant to meet its nutrient demand in a variety of soil conditions, with the aim of creating an optimal soil-sampling root volume (Forde and Lorenzo, 2001;López-Bucio et al., 2003;Morris et al., 2017). The importance of root hairs is illustrated by the fact that overall plant fitness is negatively affected in several loss-of-function root hair mutants when grown in challenging soil conditions (Wu et al., 2007;Miguel et al., 2015). Consequently, dynamic root hair morphogenesis is a key trait for improving the acquisition of essential nutrients in heterogeneous soils. Therefore, a better understanding of the root hair developmental pathway is crucial. Various studies, mostly based on the model organism Arabidopsis, have led to a better understanding of the genetic and molecular cascade that lies at the base of root hair development (Balcerowicz et al., 2015;Schoenaers et al., 2017).
Nevertheless, it is far from clear how different plant hormones feed into this genetic framework. In addition, many environmental conditions influence different stages of root hair morphogenesis, in particular nutrient availability (phosphorus, nitrogen, iron, etc.), water deficiency, and microbe interaction. Currently, it is not fully understood how these different signals This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details) are integrated by the plant root, and how they fine-tune root hair development. Here, we review the current state-of-the-art regarding hormonal regulation and environmental impact on root hair development.

General hormonal regulation of root hair development
Root hair development starts with the determination of whether an epidermal cell becomes a root hair (H; trichoblast) or non-root hair (N; atrichoblast) cell, giving rise to distinct hair and non-hair cell files in the Arabidopsis root ( Fig. 1; Dolan et al., 1994). Epidermal cell fate determination depends on a position-dependent signal originating from the underlying cortical cells. Signal perception in epidermal cells that overlie two cortical cells triggers a transcription factor cascade that ultimately inhibits the expression of GLABRA2 (GL2), a core non-root hair-determining transcription factor. As a result, ROOT HAIR DEFECTIVE 6 (RHD6) expression is initiated, leading to the initiation of a root hair bulge (Fig. 2). This bulge then begins tip growth before ultimately maturing. Root hair initiation and tip growth are each characterized by distinct yet interconnected molecular pathways (extensively reviewed in Balcerowicz et al., 2015;Schoenaers et al., 2017).
This genetic framework, which guides root hair development (Fig. 2), is heavily controlled by hormonal cues. In particular the influence of auxin and ethylene is well documented. These hormonal signals are key to the plant's ability to dynamically regulate root hair form and function in response to the changing properties of the surrounding soil environment. In addition, crosstalk exists between several hormone classes (Wang and Irving, 2011;Zhang et al., 2016;Liu et al., 2017;Zhang et al., 2018).

Auxin
Auxin is a well-characterized hormone that influences many plant developmental processes and acts as a positive regulator of root hair development (Paque and Weijers, 2016). Most auxin responses occur as a result of transcriptional and translational changes. This signalling pathway is known as the canonical pathway. To the contrary, other responses occur too fast to be dependent on nuclear transcription. These non-canonical pathways (Kubeš and Napier, 2019) will not be discussed in this review. The TRANSPORT INHIBITOR RESISTANT1/ AUXIN SIGNALING F-BOX (TIR1/AFB) receptor complex that binds to auxin is the substrate-recognition subunit of the SUPPRESSOR OF KINETOCHORE PROTEIN 1 (SKP1)/CULLIN1/F-Box (SCF) E3 ubiquitin ligase complex. Auxin-TIR1 binding promotes the interaction between TIR1/ AFB and AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) co-repressor proteins, which triggers the polyubiquitination and degradation of these Aux/IAAs. The latter causes dissociation of Aux/IAAs from AUXIN RESPONSE FACTORS (ARFs), allowing these transcription factors to regulate gene transcription ( Fig. 3A; Teale et al., 2006;Weijers and Wagner, 2016;Ma et al., 2018).
Overexpression of ARF5, 7, and 8 enhances root hair growth (Mangano et al., 2017), while root hair-specific overexpression of ARF1-4, 9-11, and 16 inhibits root hair growth, indicating that root hair development depends on the relative transcriptional activity of ARFs, which function as activators (aARFs) or repressors (rARFs) of gene transcription (Choi et al., 2018). Several auxin-signalling mutants, including aux1, axr1, axr2/ iaa7, arf5, arf7/arf19, axr3/iaa17, slr/iaa14, and iaa28 have altered root hair development (Masucci and Schiefelbein, 1996;Rahman et al., 2002;Okushima et al., 2005;Li et al., 2016), and supplementation or overproduction of auxin [e.g. gainof-function mutant SHORT HYPOCOTYL 2 (SHY2/ IAA3] leads to longer root hairs (Pitts et al., 1998;Knox et al., 2003). In addition, loss-and gain-of-function mutants of DIOXYGENASE FOR AUXIN OXIDATION 1 (DAO1), a protein that plays a major role in auxin oxidation and thus in regulating the cellular homeostasis of auxin, display altered root hair length (Porco et al., 2016). Together, these findings clearly illustrate the pivotal role of auxin in controlling root hair morphogenesis. Masucci and Schiefelbein (1996) have shown that auxin is capable of restoring root hair growth even in the hairless rhd6 mutant, which lacks the core root hair-initiating basic helix-loop-helix (bHLH) transcription factor RHD6 (Fig. 2). According to Yi et al. (2010), this occurs through the induction of the bHLH transcription factor ROOT HAIR DEFECTIVE 6-LIKE4 (RSL4), which acts downstream of RHD6. As such, RSL4 is considered to be the first auxin-responsive gene in the timeline of root hair morphogenesis. During root hair initiation, a 4 h-long RSL4 expression pulse determines final root hair length. As shown in Fig. 2, auxin induces RSL4 expression through a direct interaction of ARF5 with the RSL4 promoter (Mangano et al., 2017). Surprisingly, to date, none of the genes functioning upstream of RSL4 has been shown to be auxin regulated, suggesting that auxin is not involved in early root hair development. Conversely, it has been shown that auxin is able to directly regulate the expression of genes downstream of RSL4. For instance, ARF7 and ARF19 directly interact with the promoter sequence of the RSL4-target gene ERULUS, a trichoblast-specific receptor-like kinase and core regulator of the root hair tip growth process (Schoenaers et al., 2018). The relative importance of direct (through ARFs) and indirect (through RSL4) auxin-controlled gene expression remains to be investigated.

Ethylene
The gaseous hormone ethylene is a positive regulator of root hair development, and frequently found to functionally interact with auxin signalling (Dubois et al., 2018). A family of five receptors perceives ethylene, ETHYLENE RESPONSE 1 (ETR1), ETR2, ETHYLENE RESPONSE SENSOR 1 (ERS1), ERS2, and ETHYLENE INSENSITIVE 4 (EIN4), which are functionally redundant (Dong et al., 2010;Ju and Chang, 2015). In the absence of ethylene, the ethylene receptors phosphorylate and activate the Ser/Thr kinase CONSTITUTIVE TRIPLE RESPONSE 1 (CTR1). Subsequent phosphorylation of ETHYLENE INSENSITIVE 2 (EIN2) by CTR1 leads to the activation of EIN2 TARGETTING PROTEIN 1 and 2 (ETP1/2), which mediates ubiquitination-dependent degradation of the C-terminal domain of EIN2. As C-terminal EIN2 is required for stabilization of ETHYLENE INSENSITIVE3 (EIN3) and EIN3-LIKE 1 (EIL1) transcription factors in the nucleus, EIN3 and EIL1 remain inactive and are targeted for proteasome-dependent degradation, which requires EIN3 BINDING F-BOX 1 and 2 (EBF1/2) when ethylene is absent (Binder et al., 2007). The binding of ethylene inactivates the receptor signalling. As a consequence, EIN2 is not phosphorylated by CTR1, and a cytoplasmic C-end part is proteolytically cleaved off. The latter then moves to the nucleus, stabilizes EIN3 and EIL1, and facilitates expression of ethylene-responsive (ERF) genes (Fig. 3B;Ju and Chang, 2015;Moniuszko, 2015;Shakeel et al., 2015).
Loss of ETR1 and EIN2 function (leading to ethylene insensitivity) or blocking of ethylene biosynthesis inhibits root hair growth (Masucci and Schiefelbein, 1996;Rahman et al., 2002). Supplementation of the ethylene precursor aminocyclopropane carboxylic acid (ACC), on the other hand, stimulates root hair growth. Importantly, ACC supplementation also induces the formation of ectopic root hairs (Pitts et al., 1998). Together, these data show that ethylene is a positive regulator of root hair development that, unlike auxin, also regulates the epidermal cell fate determination pathway.
Ethylene stimulates the expression of different key root hair genes such as RHD6, RSL4, and RSL2, which regulate the later stages of root hair initiation and tip growth (Masucci and Schiefelbein, 1994;Yi et al., 2010. Feng et al. (2017) showed that EIN3 physically interacts with RHD6, and that both transcription factors coactivate the expression of RSL4 in the presence of ethylene (Fig. 2). RHD6 forms a transcriptional complex with RSL1, another bHLH transcription factor of the RSL family that seems to be an upstream regulator of RSL2 (Pires et al., 2013). Together with the complex EIN3-EIL1, they are currently considered to be key regulators of root hair initiation and elongation (Feng et al., 2017). Whether EIN3 also interacts with RSL1 and whether EIL1 acts similarly is currently unknown. Furthermore, ethylene signalling seems to go specifically through RHD6 and RSL1 because root hair initiation cannot be rescued by ethylene in the rhd6 rsl1 double mutant. Transcriptional analysis revealed the genes that are regulated by EIN3-EIL1 and RHD6-RSL1, which potentially control root hair initiation and/or elongation (Feng et al., 2017). Further research is needed to better understand ethylene-controlled root hair morphogenesis and where and how auxin and ethylene intersect in the root hair formation pathway.
There are few reports regarding the role of cytokinins in regulating root hair development. Recent findings have shed new light on the molecular basis of the cytokinin signalling pathway. Cytokinin binds to His kinase-like proteins such as ARABIDOPSIS HISTIDINE KINASE 2/3/4 (AHK2/ AHK3/AHK4) triggering their autophosphorylation. Subsequent phosphate-group transfer towards an ARABIDOPSIS HISTIDINE PHOSPHOTRANSFER PROTEIN (AHP) leads to AHP nuclear translocation and the activation of type-B ARABIDOPSIS RESPONSE REGULATORS (ARRs), which in turn induce cytokininmediated gene transcription. Finally, cytokinin signalling is subject to negative feedback regulation by expression of type-A ARRs, which are negative regulators of type-B ARRs ( Fig. 3C; Hutchison and Kieber, 2002;Hwang et al., 2012;Kieber and Schaller, 2014;Huang et al., 2018b).
The role of cytokinin in root hair development remained unclear until the transcription factor ZINC FINGER PROTEIN5 (ZFP5) was discovered. Loss-of-function zfp5 mutants exhibit a short root hair phenotype, caused by slower growth, which cannot be rescued by cytokinin addition. In wild-type plants, however, cytokinin supplementation induces an increase in final root hair length, while lowering endogenous cytokinin levels results in a short root hair phenotype. ZFP5, which is cytokinin-inducible, seems to lie at the base of cytokinin-regulated root hair morphogenesis and directly promotes the expression of CPC, a positive root hair regulator ( Fig. 2; An et al., 2012;Zhang et al., 2016). At the moment, it needs to be investigated whether other genes in the root hair developmental pathway are (direct) targets of cytokinin signalling and whether cytokinin acts through ZFP5 only.
Abscisic acid is a negative regulator of root hair growth since supplementation results in reduced root hair growth, a response that is absent in ABA-insensitive mutants. Recent work shows that ABA controls root hair growth by inhibiting the expression of the key root hair transcription factor RSL2 through the up-regulation of OBF BINDING PROTEIN4 (OBP4; Fig. 2). This transcription factor in turn binds to the RSL2 promoter sequence and negatively regulates its transcription. In line with the latter, root hairs of rsl2 were found to be irresponsive to ABA treatment (Rymen et al., 2017).
BRs seem essential for position-dependent epidermal cell fate determination in roots (Wei and Li, 2016). BR-deficient mutants display ectopic root hair formation, whereas BR-signal-enhanced mutants produce fewer root hairs. In line with this, treatment with brassinolide or bikinin produced fewer root hairs, while brassinazole, a BR-synthesis inhibitor, led to a change in epidermal cell fate determination resulting in the formation of more trichoblasts and thus more root hairs (Cheng et al., 2014). Kuppusamy et al. (2009) found that BRs positively regulate the expression of WER and GL2, both negative regulators of early root hair development (Fig. 2). Root hair formation is determined by transcription factor complex formation and lateral communication between H-and N-cells (Bernhardt et al., 2005;Savage et al., 2008). Cheng et al. (2014) have demonstrated that BRs interfere with this process, leading to aberrant cell-specific GL2 and WER activity. At low BR concentrations, the formation and activity of the transcription factor complex (WER-GL3/EGL3-TTG1) is inhibited in both H-and N-cells since WER expression is lower and the kinase BIN2 phosphorylates EGL3 and TTG1, which reduces the functionality of the transcription factor complex. This suppresses GL2 expression and thus results in formation of more root hairs as hair cell fate is promoted. At high BR concentrations, WER expression is increased in both N-and H-cells, the WER-GL3/EGL3-TTG1 transcription factor complex is fully functional (since BIN2 is not active), promoting GL2 expression and determination of the non-hair cell fate in all epidermal cells (Cheng et al., 2014;Wei and Li, 2016).

Strigolactones
Strigolactones (SLs) were recently classified as a class of phytohormones (Brewer et al., 2013). DWARF 14 (D14) receptor proteins perceive strigolactones, leading to a conformational change and subsequent signal transduction. Similar to other hormone-perception mechanisms, SL signal transduction depends on the initial ubiquitin-dependent degradation of SL-suppressor proteins (e.g. SUPPRESSOR OF MORE AXILLARY GROWTH 2 (MAX2)-LIKE family proteins; SMXLs). SMXL degradation allows downstream signalling through TCP-domain transcription factors ( Fig. 3F; Machin et al., 2020). Kapulnik et al. (2011a,b) have described that supplementation with synthetic SL (GR24) enhances root hair growth. This positive effect of SLs depends, at least in part, on MAX2 F-box proteins that function in ubiquitin-dependent protein degradation. MAX2/3/4 loss-of-function plants display a short root hair phenotype. However, contrary to max3 and max4 mutants, the max2 root hair length can only be restored at high GR24 concentrations. These data highlight an emerging role for SLs in controlling root hair morphogenesis.

Gibberellin
Among other things, gibberellins (GA) control cell elongation, root apical meristem (RAM) growth, and the formation of lateral roots (Ogawa et al., 2003;Tyler et al., 2004;Achard, 2013, 2016;Hu et al., 2018), yet little is currently known on GA involvement in root hair development. GA signal perception works through three orthologs of GIBBERELLIN INSENSITIVE DWARF 1 (GID1A-C), soluble nuclear receptors, which consist of a GA-binding pocket and a flexible N-terminal extension. GID1-GA binding leads to a GID1 conformational change that enables subsequent interaction with DELLA repressor proteins, leading to the formation of a GA-GID1-DELLA complex. Upon GA-GID1-DELLA complex formation, the DELLA proteins undergo small conformational changes that enhance recognition with the F-box proteins SLEEPY 1 (SLY1) and SNEEZY (SNE/ SLY2). This promotes DELLA ubiquitination and subsequent destruction of DELLAs by the 26S proteasome. As a result, the transcription factors (PIFs, JAZs) that were inactivated by DELLAs become de-repressed and GA-induced signal transduction is initiated (Fig. 3G; Davière and Achard, 2013).
There are no reports on direct root hair growth responses towards GA in Arabidopsis. Nevertheless, Devlin and Brown (1969) have shown that GA has a positive effect on the elongation rate of Agrostis alba root hairs. A similar effect was seen in Datura innoxia, where exogenous GA resulted in more hairy roots and increased root hair elongation (Ohkawa et al., 1989). How GA feeds into the genetic root hair growth framework needs further investigation.
The complex crosstalk among phytohormones during root hair development Hormonal crosstalk is instrumental in plant development. The root hair developmental pathway is fundamentally controlled by input from several hormonal classes. Hence, understanding how hormone signalling pathways are interconnected is a crucial step towards building an integrated view of root hair morphogenesis.

Auxin and ethylene
A large effort has gone into understanding the heavily interconnected auxin and ethylene signalling pathways during root hair growth (Vanstraelen and Benková, 2012;Muday et al., 2012;Robles et al., 2013;Lee and Cho, 2013;Qin and Huang, 2018). A functional interaction is illustrated by the finding that auxin biosynthesis is up-regulated by ethylene in the root tip and during polar root hair initiation (Stepanova et al., 2005(Stepanova et al., , 2007Růzicka et al., 2007;Swarup et al., 2007;Ikeda et al., 2009). Vice versa, Strader et al. (2010) demonstrated that the aux1 mutation suppresses the long hair phenotype of the eto1 mutant, suggesting that ethylene needs auxin transport during the regulation of some phases of root hair development. In addition, the rhd6 root hair-less phenotype can be rescued by both auxin and ACC (Masucci and Schiefelbein, 1994), showing that both hormones can influence root hair development downstream of RHD6-controlled gene transcription.
The importance of both auxin and ethylene signalling is further illustrated by the fact that root hairs are defective in the aux1 ein2 double mutant (Rahman et al., 2002) and that transcriptome analysis identified a large group of common root hair genes that are both up-regulated by auxin and ethylene supplementation (Bruex et al., 2012). The nature of auxin-ethylene crosstalk is, however, complex. Hence, auxin and ethylene signalling are not per definition mutually interdependent. For instance, the phenotype of aux1 etr1 double mutants can be suppressed by administration of IAA, but not by ACC addition (Masucci and Schiefelbein, 1996). To the contrary, ein3 eil1 rhd6 rsl1 quadruple mutant roots remain hairless after auxin application, suggesting that the link between ethylene and auxin signalling is not that straightforward after all during root hair morphogenesis (Feng et al., 2017). Nonetheless, auxinethylene crosstalk is embedded throughout a variety of plant developmental processes (Muday et al., 2012). In general, it is now believed that functional auxin signalling is required for optimal ethylene responses. In line with the latter, the aux1 and tir1 auxin mutants show reduced sensitivity to ethylene treatment. Vice versa, it was found that the expression of auxin homeostasis and polar transport genes is increased as a result of increased ethylene levels (Zemlyanskaya et al., 2018).
Mechanistically, the interconnection of auxin and ethylene signalling can be explained by the presence of an EIN3 binding region in the AUXIN RESISTANT 1 (AUX1) promoter (Zažímalová et al., 2007). Similarly, ACC transport is promoted by LYSINE HISTIDINE TRANSPORTER 1 (LHT1), which in turn is under the control of auxin (Lewis et al, 2013). Moreover, the transcription of 1-AMINOCYCLOPROPANE-1-CARBOXYLATE SYNTHASE (ACS) genes is up-regulated by auxin in an ARFdependent manner (Stepanova et al., 2007). Auxin and ethylene seem to impinge on the same core root hair pathway, through RSL4 and RHD6, respectively, suggesting that both hormonal signals need to be considered in order to understand root hair growth.

Cytokinin, auxin and ethylene
A functional auxin-cytokinin interaction during plant development is emerging. Surprisingly, cytokinin and auxin were found to interact in a synergistic or antagonistic manner during plant development. For instance, both hormones were found to positively regulate root/stem growth and flowering (Srivastava, 2002;Moubayidin et al., 2013;El-Showk et al., 2013;Schaller et al., 2015;Liu et al., 2017). Contrastingly, auxin and cytokinin have an antagonistic role in lateral root formation. On one hand, type-B ARR1 stimulates the expression of SHY2, which encodes IAA3 (Tian and Reed, 1999) and negatively influences auxin signalling. This indirectly causes degradation of PIN1 and thereby suppresses auxin transport (Dello Ioio et al., 2008;Marhavý et al., 2014), which leads to an antagonistic relationship between auxin and cytokinin. On the other hand, auxin can inhibit SHY2/ IAA3 through degradation via the SCF complex, resulting in PIN expression and subsequent auxin transport. In addition, Schaller et al. (2015) have shown that auxin directly activates the transcription of ARR7 and ARR15, both type-A ARRs known to be negative regulators of the cytokinin response. During early lateral root development auxin promotes the cytokinin inhibitor AHP6, which represses cytokinin signalling by modulating PIN1 and, as such, auxin distribution (Moreira et al., 2013).
Cytokinin, together with ethylene, stimulates the abscission of different plant organs. In fact, the stability of ACC synthase can be increased by cytokinin, resulting in higher ethylene production (Chae et al., 2003;Hansen et al., 2009). For root hair development, however, the effect of cytokinin might be independent of auxin and ethylene, especially since cytokinin can rescue root hair elongation when auxin (axr1) and ethylene signalling (etr1-1) are disturbed . Furthermore, auxin and ethylene can stimulate root hair elongation when cytokinin is defective in an overexpression line of CKX2, a cytokinin oxidase . It is therefore plausible that several common genes in the root hair-forming network can be targets for auxin, ethylene, and cytokinin. This idea is strengthened by the observation that the phenotype of the rhd6 mutant can be rescued by exogenous auxin and ethylene, but also by cytokinin (Masucci and Schiefelbein, 1994;Zhang et al., 2016). Candidates downstream of RHD6 are RSL1/2/4 (Fig. 2). In addition, auxin regulates cytokinin biosynthesis (Nordström et al., 2004) as the auxin response factors ARF5 and ARF7 control cytokinin biosynthesis by up-regulating CYTOKININ RESPONSE FACTOR (CRF) genes and cytokinin biosynthetic enzymes Liu et al. (2017). Furthermore, it has been demonstrated that cytokinin can regulate ethylene biosynthesis (El-Showk et al., 2013), showing a complex interplay of cytokinin with other hormones and as such influencing root hair development.

Strigolactones, auxin, and ethylene
Strigolactones have emerged as major players in regulating plant growth and development (Brewer et al., 2013). Several studies have shown that SLs functionally interact with auxin transport through effects on PIN and TIR1 protein abundance (Kapulnik et al., 2011a,b;Ruyter-Spira et al., 2011;Mayzlish-Gati et al., 2012;Machin et al., 2020). These changes in auxin transport lead to changes in auxin levels, which can impact on root hair development. However, how exactly SLs regulate PIN abundances remains unclear.
Treatment of mutants showed that SL signalling is not required for ethylene-responses in root hairs and that SL responses need ethylene synthesis and signalling. In line with this, SL treatment induced ACC synthase genes, encoding key rate-determining enzymes in ethylene synthesis (Yamagami et al., 2003). Besides, SL signalling is not required for auxin to elicit its root hair growth stimulating effect, while auxin perception is needed to some extent for the SL effect on root hairs. In addition, auxin and SL have an additive effect on root hair development. Treatment of aux1-7 ein2-1 mutants, which are impaired in sensitivity towards auxin and ethylene, showed reduced sensitivity to SLs. These observations, together with the finding that SLs can regulate PIN abundances in an unknown manner, strongly suggest that there is intimate crosstalk between SLs, ethylene, and auxin in the regulation of root hair elongation. Evidence indicates that SL affects root hair elongation via up-regulation of the ethylene pathway, which in turn acts on root hair development and interacts with the auxin pathway. The latter is supported by the fact that SLs affect PIN abundances, auxin transport capacities, and thus cellular auxin (Kapulnik et al., 2011a,b;Ruyter-Spira et al., 2011;Mayzlish-Gati et al., 2012;Machin et al., 2020).
These descriptions clearly show that hormonal regulation of root/root hair development cannot be described by a linear signalling cascade. Instead, the mode of action of different hormones is integrated in a complex, interconnected network of which the developmental output is the result of crosstalk between the latter and direct and indirect effects. It is crucial to emphasize that there are still many outstanding questions regarding hormone interactions during root hair development.

Influence of environmental stressors on root hair development and their interaction with hormones
Rhizosphere composition dramatically impacts root hair development. Nutrients and abiotic stressors are unevenly distributed throughout the soil. Hence, roots encounter heterogeneous conditions along their growth axis, forcing them to dynamically regulate root system architecture and, for example, root hair morphogenesis (Morris et al., 2017;Shahzad and Amtmann, 2017).

Environmental stressors affecting root hair development
Phosphorus (P i ) is a crucial soil-immobile macronutrient that is taken up and recycled by the plant root. It is compartmentalized specifically to the top-soil, as its main source is plant detritus. Plants have evolved specialized mechanisms for P i sensing, uptake and transport. Mechanistically, P i status is sensed by SPX-domain-containing proteins. They act as cellular receptors for inositol pyrophosphate molecules (PP-InsP), a proxy for cellular P-content, and regulate P i homeostasis (Puga et al., 2017). SPX proteins negatively regulate PHOSPHATE STARVATION RESPONSE (PHR) transcription factor and its homologs PHR1-LIKE 1 (PHLs) (Puga et al., 2014;Qi et al., 2017). The SPX-PHR1 complex dissociates in the absence of PP-InsPs, releasing PHR1 to interact with the promoters of P i starvation induced genes (PSI) (Bustos et al., 2010;Wild et al., 2016). Intriguingly, Ried et al. (2019, Preprint) recently described that InsP 8 /SPX-induced PHR1 oligomerization is required for PHR-mediated transcriptional regulation during P i homeostasis. Choi et al. (2017) showed that sustained exposure to P i -deficient conditions results in a decrease of cellular ATP levels. The latter triggers PP-InsP generation and degradation by VIH1/2, two bifunctional inositol pyrophosphate kinases/ phosphatases , allowing the plant to maintain its P i /ATP homeostasis and regulate the P i starvation response.
In low phosphate conditions the root is shorter, more lateral roots are formed, and root hairs grow much longer and denser (Fig. 4), facilitating top-soil foraging and increased nutrient uptake (Ma et al., 2001;Koevoets et al., 2016;Zhang et al., 2018). These adaptations occur through local up-regulation of auxin biosynthesis (Bhosale et al., 2018) and ARF7/19induced PHR1 expression, which promotes root hair growth (Huang et al., 2018a). Recently Bhosale et al. (2018) found that P i deficiency is first sensed at the root tip, causing up-regulation of TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS 1 (TAA1), a crucial factor in the auxin biosynthetic pathway. AUX1-mediated auxin influx in the lateral root cap and adjacent epidermal cells is then required for increased auxin transport towards the epidermal root hair zone. Subsequently, an auxin-induced genetic cascade is initiated, which involves ARF19, RSL4, and RSL2, enhancing root hair elongation (Fig. 5A). Conversely, high phosphate conditions are able to strongly repress RSL4 expression and root hair growth. Mangano et al. (2018) revealed that when root hairs encounter two conflicting growth signals, i.e. high phosphate levels (inhibits root hair growth) and high auxin levels (enhances root hair growth), RSL2 plays an important role in restoring reactive oxygen species homeostasis and concomitantly in root hair growth recovery. The importance of RSL2 rather than RSL4 in the integration of conflicting growth signals is further illustrated by the finding that the rsl4 mutant displays auxin-enhanced root hair length in either low or high P i conditions, whereas in the presence of high phosphate, auxin supplementation of rsl2 roots has no phenotypic effect.
According to Song et al. (2016), P i starvation indirectly leads to ectopic root hair formation by elevation of EIN3 protein levels, in turn up-regulating root hair initiation and elongation genes (Fig. 5A). This was found in a dominant negative mutation of an ethylene receptor (ERS1). It is currently unknown whether the EIN3 level increases due to an increase in ethylene synthesis or ethylene-dependent auxin signalling. Besides auxin, also SLs play a role in P i -regulated root hair growth. Hence, SL concentrations increase under low P i and nitrogen levels (Pérez-Torres et al., 2008). In the presence of MAX2, the transcription of the auxin receptor TIR1 is stimulated. This promotes auxin signalling and results in an increase in root hair density and induction of PSI genes (Ma et al., 2001;Chiou and Lin, 2011). Mayzlish-Gati et al. (2012) have shown that the supply of GR24 restores this response in SL-synthesis mutants, pointing to the importance of SL (Fig. 5A).
Iron (Fe) is a crucial micronutrient for plants as it is a key cofactor for various enzymatic reactions. However, Fe solubility is low in aerobic and neutral pH environments, leading to limited bioavailability and consequent Fe deficiency (Ivanov et al., 2012). Arabidopsis roots alter Fe solubility by acidifying the rhizosphere through the action of the H + -ATPase AHA2. Iron is then reduced from ferric (Fe 3+ ) to ferrous (Fe 2+ ) iron by FERRIC REDUCTASE-OXIDASE 2 (FRO2), and the bivalent ion is then imported into root cells by IRON-REGULATED TRANSPORTER 1 (IRT1). In Arabidopsis, UPSTREAM REGULATOR OF IRT1 (URI) becomes phosphorylated under Fe deficiency and forms heterodimers with specific bHLH transcription factors to induce expression of many Fe-deficiency-induced genes, including BRUTUS (BTS), BRUTUS-LIKE 1 (BTSL1), POPEYE (PYE), and other class bHLHs. The latter induce the expression of the transcription factor FER-LIKE IRON DEFICIENCY-INDUCED TRANSCRIPTION FACTOR (FIT), which forms heterodimers with bHLH38/39/100/101  and regulates various genes involved in Fe uptake, such as IRT1 and FRO2 (Lan et al., 2012;Kobayashi and Nishizawa, 2015). In parallel, PYE regulates the expression of genes that are involved in Fe homeostasis (Kim et al., 2019). Under low Fe, BTS/BTSLs are inactive, but when Fe availability increases or is restored, Fe binding to the proteins activates their ubiquitin ligase activity (Kobayashi and Nishizawa, 2015) and they interact with the different transcription factor complexes, leading to their degradation (Kim et al., 2019) (Fig. 5B).
Under low Fe conditions, rather than forming ectopic root hairs, plant roots increase their absorptive surface by inducing root hair branching (Müller and Schmidt, 2004;Lan et al., 2012). Several lines of evidence suggest that both auxin and ethylene are involved in this response. The auxin-insensitive mutant axr2-1 (Wilson et al., 1990), mutated in IAA7 (Nagpal et al., 2000), also displays branched root hairs, as under Fe-deficient conditions. Transcriptome analysis showed that Fe deficiency causes differential expression of several genes involved in ethylene biosynthesis and auxin/ethylene signalling (Thimm et al., 2001). It is plausible that these ethylene-and auxin-related genes are targets of the FIT and PYE transcription factors. The latter would explain the altered root hair morphology (due to changes in auxin responsiveness) and (low) percentage of ectopic root hair formation (by ethylene) (Fig. 5B). Importantly, Lan et al. (2011) reported that post-transcriptional and posttranslational processes are also involved in the low Fe response.
Nitrogen is a major limiting mineral nutrient that is available in the soil as ammonium (NH 4 + ) and nitrate (NO 3 − ) (Forde, 2002). Besides being an essential nutrient, nitrate also functions as a signalling molecule regulating many different aspects of development, including root development (Walch-Liu et al., 2006). This mobile nutrient is distributed by vertical water flow, known as leaching, and accumulates in deeper soils (Jobbágy and Jackson, 2001). Expectedly, low nitrate induces primary root growth yet inhibits lateral root development, enabling root systems to reach deeper soil layers (Bouguyon et al., 2015;Araya et al., 2016). In contrast, high nitrate conditions lead to slower primary root growth and an increase in root branching, which aids in the foraging for nutrients in that particular nitrate-rich soil patch or layer. When external nitrate concentrations are even higher, lateral root development is arrested in the early stages after emergence. Because of its high concentration in the soil, nitrate accumulates in the shoot and limits basipetal auxin fluxes, which results in lower root auxin and negative regulation of lateral root development (Walch-Liu et al., 2006). Root hairs seem to play an important role in nitrate uptake, as root hair defective mutants accumulate significantly less nitrate than wild-types, while mutants with ectopic root hairs show an increase in nitrate content. Furthermore, root hairs seem important in detecting or sensing changes in nitrogen (Shin et al., 2005). In addition, nitrate stimulates an increase in root hair density (Vatter et al., 2015) by decreasing trichoblast cell lengths and not by inducing ectopic root hair formation, nor by stimulating root hair elongation (Canales et al., 2017).
Two families of transporters participate in the uptake of nitrate in the root, NITRATE TRANSPORTER 1 and 2 (NRT1/2), where NRT1.1 and NRT2.1 are thought to be the most important (Okamoto et al., 2003). To be in the active form as a nitrate transporter, NRT2.1 has to interact with NAR2.1 (also called NRT3.1) and form an NRT2.1-NAR2.1 hetero-oligomer (Yong et al., 2010), whose activity is under post-translational control (Jacquot et al., 2019, Preprint).  Currently, NRT1.1 is the only known transporter for nitrate in Arabidopsis (Ho et al., 2009;Giehl and von Wirén, 2015) and functions both as a transporter and as a receptor, called a transceptor. In line with the latter, NRT1-1 loss-of-function mutants failed to show an increase in nitrate-induced root hair density (Vatter et al., 2015). Downstream of NRT1.1, two transcription factors, TGACG SEQUENCE-SPECIFIC BINDING PROTEIN 1 and 4 (TGA1/TGA4), play a redundant role in increasing the expression of CPC, which seems to be required for the reduction of trichoblast cell length and the subsequent increase in root hair density (Canales et al., 2017). In addition, both transcription factors regulate the expression of NRT2.1 and NRT2.2, which are required for the induction of lateral root development but not for the reduction of cell elongation in the primary root (and resulting increase in root hair density). Furthermore, the nitrate transport function of NRT1.1 is needed for the increase in expression of the auxin receptor AFB3 in pericycle cells and its downstream target, NAC4 Vidal et al., 2014). The involvement of auxin is further evidenced by the fact that NRT1.1 can facilitate the uptake of both nitrate and auxin. Consequently, in low nitrate conditions, auxin is taken up, while in higher nitrate conditions, nitrate outcompetes auxin leading to decreased auxin uptake through NRT1.1 and, consequently, auxin accumulation. This increased auxin level then affects lateral root development (Krouk et al., 2010). Tian et al. (2009) have described a rapid rise in ethylene production upon exposure to high nitrate levels. It is known that increased ethylene concentrations lead to reduced root epidermal cell elongation and thus an increase in root hair density (Swarup et al., 2007;Markakis et al., 2013). The observation that both ethylene and auxin impact epidermal cell length could also explain the involvement of NRT1.1 in the shorter trichoblast cell length in high nitrate conditions (Fig. 5C).
In nitrate-deficient conditions, simultaneous NRT1.1, NRT2.1, and NRT2.2 expression is required for maintaining root growth (Ye et al., 2019). Zheng et al. (2013) have demonstrated that NTR2.1 is strongly up-regulated in low nitrate conditions in an attempt to enhance nitrate uptake. In parallel, ethylene biosynthesis is up-regulated and dependent on NRT2.1. Surprisingly, ethylene reduced the expression of NRT1.1 and NRT2.1 and concomitantly the uptake of nitrate, which acts as a negative feedback loop ( Fig. 5C; Tian et al., 2009). No data on root hair responses are available.
During drought, plants are confronted with osmotic stress, causing cell dehydration and membrane/macromolecule damage (Lisar et al., 2012). Primary root growth, lateral root formation, and root hair development are reduced upon osmotic stress, a process that is heavily controlled by ABA in a complex interacting network with cytokinin and other phytohormones (De Smet et al., 2003;Nakashima et al., 2014;Singh and Laxmi, 2015;Rowe et al., 2016;Tiwari et al., 2017). ABA stimulates the expression of the gene for ABSCISIC ACID INSENSITIVE4 (ABI4) transcription factor, among many other genes, which in turn inhibits PIN1 expression, leading to reduced polar auxin transport (Shkolnik-Inbar and Bar-Zvi, 2010) and thus lower auxin levels in the root. Besides having a direct inhibitory effect on root hair development, ABA could thus also act indirectly through changing the levels of other hormones. Under progressive drought stress, root hairs seemed to branch more often, which was also mediated by ABA and auxin, probably in a similar (concentration-dependent) manner (Bobrownyzky, 2016).
Increased salinity inhibits plant growth by creating osmotic stress and inhibiting photosynthesis in the shoot (Koevoets et al., 2016). In addition, primary and lateral root growth and root hair development are suppressed during salt stress (Wang et al., 2008;Mabrouk and Heikal, 2008). Salt exposure causes down-regulation of PIN genes and stabilization of AUXIN RESISTANT3 (AXR3)/IAA17, reducing polar auxin transport and inhibiting auxin signalling, respectively. The resulting reduction in auxin levels and signalling then negatively affects root hair development (Liu et al., 2015).
Biotic factors, such as fungi, bacteria, insects, and pathogens can influence plant development. The main actors in the defence against such biotic factors are salicylic acid (SA), which offers biotroph resistance, and jasmonic acid (JA), which mediates necrotroph resistance (Takatsuji and Jiang, 2014). From the six major hormones discussed in this review, only ethylene and cytokinin positively interact with SA and JA and provoke the synthesis of pathogenesis-related proteins (Yang et al., 2015). Pathogenic strains of Pseudomonas syringae inhibit primary root elongation and stimulate lateral root and root hair formation/elongation, which requires functional ethylene signalling and does not seem to be associated with changed auxin levels (Pecenková et al., 2017). The same effect is noticeable in environments containing beneficial rhizobacteria (Pecenková et al., 2017) and mycorrhizal, saprotrophic, and pathogenic fungi. In their presence, a peculiar clumped root phenotype is observed, marked by a short primary root and more or longer lateral roots (Casarrubia et al., 2016). Unfortunately, the effect on root hair development has not been reported.

Environmental stressors with no reported effect on root hair development
Temperature changes greatly impact on root functioning. For instance, seedling growth at 26-29 °C stimulates primary root elongation (Hanzawa et al., 2013;Wang et al., 2016;Martins et al., 2017;Feraru et al., 2019). Although both auxin and BR were reported to be involved in this response, the underlying hormonal mechanism is currently under debate. Yet, it was recently shown that enhanced primary root growth at higher temperatures coincided with increased auxin signalling. The latter is linked to the inhibition of PIN-LIKES6 (PILS6) proteins that limit nuclear auxin by sequestering auxin to the ER (Feraru et al., 2019). The effect of high temperature exposure on root hair development has not been studied/reported to date. Shibasaki et al. (2009) have shown that acute cold stress (4 °C) leads to inhibition of root elongation due to blocked basipetal auxin transport. The latter was caused by reduced intracellular trafficking of PIN2 and PIN3. After low temperature acclimation, the elongation rate recovers. It is plausible to think that cold-affected auxin concentrations regulate root hair development, but the exact effect of low temperatures on root hair development is still largely unexplored.

Conclusion and future remarks
A large effort has gone into understanding the regulatory mechanisms controlling root hair development, with the aim of harnessing this knowledge to improve root functioning in future environmental conditions (e.g. nutrient depletion, water deprivation). The combination of state-of-the-art tools to study the molecular-genetic background of plant growth has led to a drastically improved understanding of how different plant hormones impinge on the root hair growth pathway.
In this review we have highlighted the general root hair developmental scheme and how hormones control this. Combined, it is clear that highly complex hormonal crosstalk and sensing of rhizosphere conditions lie at the base of root hair morphogenesis. At this stage, constructing an integrated view of the root hair signalling cascade, including hormonal and environmental control, is too optimistic. Yet, with the scientific community increasingly embracing the complexity and interconnectivity of the involved pathways, and harnessing the power of -omic datasets and hormone marker and sensor lines, we are closing in on deciphering the role of novel genes in the different phases of root hair development. Integration of all these data allows us to link the environmental changes to changes in hormone levels and finally to their effect on the (sub)cellular processes that actually guide root hair growth.