Plant Sciences/Arabidopsis root development

This Article is a draft, you can participate to expand it and clean it up.

To do:

summarize interest of Arabidopsis root for biology
fill the sections more consistently

Here is a temporary plan about what this review will talk about:

Overview of the Arabidopsis seedling
Overview of the root
Tissues
Developmental zones
Structure of the root meristem
Cell initials
Primary vs lateral roots (embryonic vs post-embryonic origin)
Hormones in the Arabidopsis seedling root
Biotic and abiotic factors affecting root development

Intro

Plant roots are responsible for nutrient and water uptake and provide physical support to the plant. Most of the root system is made of lateral roots that originate postembryonically. Lateral root development is controlled by different factors including nutrient concentration in the plant and the soil (Lopez-Bucio et al., 2003; Malamy, 2005, for a review). This plasticity allows adaptation of the root system to the soil, a very heterogeneous and changing environment, and is consequently very important for the survival of the plant (Grime et al., 1986; Hodge, 2004). The root of the model plant Arabidopsis_thaliana is a very useful tool for investigating the basis of plant development.The Arabidopsis root has a simple structure and the root meristem is an excellent system in which to study stem cell biology.

The Root Apical Meristem (RAM)

Primary root growth occurs at the root apical meristem (RAM). Unlike the SAM, the RAM produces cells in two directions. The RAM produces a cap of tissue called the root cap, which covers the distal tip of roots. The root cap protects the root tip as it grows through the soil. 1999). The RC perceives and processes many environmental stimuli, and mediates the direction of root growth accordingly. Gravity (gravitropism), light (phototropism), obstacles (thigmotropism), gradients of temperature (thermotropism), humidity (hydrotropism), ions and other chemicals (chemotropism) are all examples of environmental stimuli that are perceived and processed by the cap (Hasenstein & Evans 1988; Ishikawa & Evans 1990; Okada & Shimura 1990; Fortin & Poff 1991; Takahashi 1997;Eapenet al. 2003).

Cells are continuously sloughed off the outer surface of the root cap. The RAM also produces cells proximally that contribute to the root proper, but unlike the SAM, the RAM produces no lateral appendages.

The root meristem is a largely invariant structure made up of few tissue types that undergo predictable divisions and do not produce lateral structures (Dolan et al., 1993). Primary root tissues are organized in concentric cylinders of epidermis, ground tissue (cortex and endodermis), and stele (pericycle and vasculature) from outside to in. These, in turn, are made up of longitudinal cell files that originate from single cells termed initials Scheres et al., 1994). Initials fulfill the minimal definition of a stem cell by producing two cells in every division: the regenerated initial, and a daughter cell that differentiates progressively upon displacement by further rounds of division. Two terminal tissues, the columella (central) and lateral root cap, are also produced by the activity of initials. Together, initials for all tissue types surround a group of four to seven mitotically less active cells in Arabidopsis thaliana known as the quiescent center (QC). Nearly every animal system relies on local signaling to form a stem cell niche, or microenvironment, that promotes stem cell status (reviewed in Spradling et al., 2001; Fuchs et al., 2004). Laser ablation experiments have shown that this is also the case for the plant root and have identified the QC as the source of a signal that inhibits differentiation of the contacting initials (van den Berg et al., 1997). QC cells position the stem cell niche but also behave as stem cells in their own right. Occasional QC divisions are self-renewing and replenish initials that have been displaced from their position (Kidner et al., 2000). Little is known about the molecular mechanisms that determine the properties of the QC or initial cells. Stem cell fate has been correlated with the position of a local maximum of auxin phytohormone perception in the QC and columella root cap initials (Sabatini et al., 1999). Auxin signaling is necessary for QC initiation in the embryo (Hardtke and Berleth, 1998;Hamann et al., 2002)

RAM structure

In many root meristems, cell files can be traced back to a small group of initials. Often these initials are organized in a tiered arrangement, with each tier giving rise to particular tissues. In radish, tier 1 produces root cap and epidermis, tier 2 cortex and tier 3 vasculature. In maize, tier 1 produces only root cap, tier 2 epidermis and cortex, tier 3 vasculature. Other plants show no tiered arrangement or only a single initial layer. Many lower plants such as ferns have a single large pyramidal cell called the apical cell. The apical cell alternates cleavages along its four faces and is the ultimate source of all cells in the meristem.

The center of the RAM is occupied by a quiescent center which has low mitotic activity. The quiescent center is most apparent in actively growing roots and is lost during dormancy, carbohydrate starvation or root cap removal. This is the same region of the meristem to which cell files trace and was postulated to contain the root initials. Evidence suggests the quiescent center does function as the zone of initials. Colchicine induces polyploidy in any cells undergoing mitosis at the time of treatment. In actively growing roots treated with colchicine, sectors of polyploid cells were transient indicating that no initial cells were dividing at the time of treatment. If the quiescent center is eliminated by removal of the root cap, prior to colchicine treatment, then stable sectors of polyploid cells are produced and persist after re-establishment of the quiescent center. Thus infrequent division of initial cells in the quiescent center is the source of cells for the RAM. More recent cell lineage experiments have used transposon activation of a GUS reporter to mark root cell lineages, leading to the same conclusions.

Surgical manipulations have provided information about RAM organization. Quiescent centers surgically explanted to sterile culture regenerated a meristem and grew as a root. Bisected RAMs formed 2 meristems and roots. These experiments demonstrate the organizing capacity for RAMs. Experiments with maize which has a highly tiered meristem suggest the layered structure is not required for normal function. Portions of the maize RAM were removed by glancing incisions and when the RAM regenerated, the layered arrangement was no longer present on the cut side, but was still apparent on the undamaged side. Although the structural arrangement of cells differed from one side to the other, the meristem still produced a normal root. The layered structure may therefore reflect “status quo” patterns of cell division but not have functional significance in the normal production of root tissues.

Embryonic origin of the RAM

The RAM contains the only cells in the plant embryo that are derived from the suspensor rather than the embryo proper. Part of the RAM is derived from the embryo proper but the quiescent center and columella initials (see below) are derived from the very last cell of the suspensor, called the hypophesis.

Arabidopsis root

The Arabidopsis root has a simple structure. There are an unusually small number of tissue and cell types in the Arabidopsis root. It possesses a single layer of cells within its epidermal, cortical, endodermal, and pericycle tissues, as well as a simple set of cells in its vascular tissue. In addition, the number of cells within each layer is fairly constant. For example, the Arabidopsis primary root always possesses eight files of cortex cells, eight files of root-hair cells, and approximately 10-14 files of non-hair cells in the epidermis. Furthermore, the developmental origin of each of the tissue types within the root meristem has been defined.

The Arabidopsis root is a highly ordered affair, consisting of concentric cylinders of cells, with new cells being added at the distal root tip [4]. This radial pattern emanates, by a series of stereotyped divisions, from a small group of stem cells which surround four ‘quiescent centre’ cells at the heart of the root meristem (Figure 1). The quiescent centre has very little mitotic activity itself but functions to maintain the stem cell status of adjacent cells. Cells laid down in front of the advancing tip differentiate to form the root cap which protects the quiescent centre and stem cell niche as they push through the soil. In the cells left behind by the growing tip, division eventually gives way to a phase of rapid cell expansion without division which marks the end of the meristematic zone and the beginning of the elongation zone (Figure 1). Elongated cells then begin to differentiate, marked most clearly by the appearence of root hairs [4]. Thus, as well as its radial organisation, the root has a more general proximo-distal pattern of activities.

Arabidopsis roots are extremely small and simple: the radial pattern consists of concentric arrangements of 6 different tissues, each a single cell layer thick. This allows every cell to be traced back to a specific progenitor in the meristem. All cells trace back to a small group of cells called the promeristem (see handout) and the pattern of cell divisions in the promeristem is invariant. Fate mapping showed that the central cells and columnella root cap cells of the embryonic RAM are derived from the basal cell formed by the first zygotic division, while the remainder of the promeristem is derived from the apical cell.

Cell signaling and RAM organization

Although cell divisions in the arabidopsis RAM typically follow a set pattern, generating relatively invariant cell lineages, laser ablation studies indicate that position rather than lineage is the important factor for determining cell fate (identity). For example, if a cortical initial is ablated, a pericycle cell will invade the voided space, assume the pattern of cell division characteristic of a cortical initial and give rise to cortical and endodermal derivatives. Thus the cell changes fate from pericycle to cortical initial when it’s position is altered.

These studies also demonstrated that the differentiated cells proximal to the RAM influence the fate of cells in the RAM. The cortical/endodermal initial undergoes a transverse division after which the daughter undergoes a periclinal division to give rise to the endodermis and cortex. If three adjacent cortical initial daughter cells were ablated, such that a cortical initial was no longer in contact with any daughter cell, the initial still divided to produce a daughter but now the daughter did not undergo the typical asymmetric periclinal division to generate the cortical and endodermal cell files. Therefore, the more mature, differentiated cells direct the pattern of division and differentiation of the initial cells (van den Berg, 1995).

Two mutants show the same phenotype: endodermis and cortex are replaced by a single cell layer. In the short root mutant the single layer has characteristics of cortex thus shr is required to specify endodermal identity. In contrast, scarecrow has both endodermal and cortical characteristics in the single layer and thus appears to be involved in specifying the cell division that generates the two layers.

The quiescent center regulates the differentiation of neighboring cells. If one or 2 QC cells were ablated, the columnella initials that were no longer in direct contact with a QC cell ceased dividing and underwent differentiation as a columnella cell. Similarly, cortical initial cells that were no longer in direct contact with the a QC cell behaved as cortical daughter cells and underwent an asymmetric division to generate cortical and endodermal cell files. Therefore, direct contact with the QC inhibits the differentiation of initial cells in the RAM (van den Berg, 1997).

Developmental Zones

A developmental gradient is apparent at the growing tip of roots. This gradient can be subdivided into 3 zones. The apical tip is the meristem or zone of cell division. The next zone proximal to the meristem is the zone of elongation where cell division ceases and there is rapid cell growth by elongation. Then comes the zone of differentiation or specialization, where cells assume their final fate. The zone of differentiation is made obvious by the appearance of root hairs in the epidermis and lignification of the xylem that can be visualised by autofluorescence. The remaining part of the root called mature zone is able to produce lateral roots.

Radial Organization of the Root

The outer layer of cells is the epidermis, next layer is the cortex, followed by the endodermis, pericycle and vasculature. Near the root tip, there is also a layer of lateral root cap cells outside the epidermis. In arabidopsis, each of these tissues corresponds to a single layer of cells.

Hormones and root development

Auxin and root development

Sites and regulation of auxin biosynthesis in Arabidopsis roots

Auxin has been shown to be important for many aspects of root development, including initiation and emergence of lateral roots, patterning of the root apical meristem, gravitropism, and root elongation. Auxin biosynthesis occurs in both aerial portions of the plant and in roots; thus, the auxin required for root development could come from either source, or both. To monitor putative internal sites of auxin synthesis in the root, a method for measuring indole-3-acetic acid (IAA) biosynthesis with tissue resolution was developed. We monitored IAA synthesis in 0.5- to 2-mm sections of Arabidopsis thaliana roots and were able to identify an important auxin source in the meristematic region of the primary root tip as well as in the tips of emerged lateral roots. Lower but significant synthesis capacity was observed in tissues upward from the tip, showing that the root contains multiple auxin sources. Root-localized IAA synthesis was diminished in a cyp79B2 cyp79B3 double knockout, suggesting an important role for Trp-dependent IAA synthesis pathways in the root. We present a model for how the primary root is supplied with auxin during early seedling development.

Auxin and RAM organization

Previous work from the Scheres lab [3] showed a high level of induction of auxin responsive genes in a tightly focused region just below the quiescent centre, suggesting a point source of auxin. Various experimental manipulations to increase, reduce or move this point were found to bring about concomitant changes in cell fate, consistent with a morphogen-like role for auxin in root tip patterning [3,5].

A GUS reporter for auxin concentration shows a concentration of auxin just at the distal end of the vasculature, just proximal to the RAM. Disruption of this localized accumulation by auxin transport inhibitors, exogenous auxin application or mutations in genes involved with auxin response, disrupts the organization of cell types in the root apex, as visualized by promoter trap cell markers. Thus the proper localization and perception of this auxin maximum are important for organizing the RAM (Sabatini, 1999).

In maize it has been reported that cells of the QC are characterized by their highly oxidized status. Glutathione and ascorbic acid occur predominately in the oxidized forms in the QC. This is contrasted with the status of these redox intermediates in adjacent, rapidly dividing cells in the root meristem, in which the reduced forms of these two species are favored. Using a redox sensitive fluorescent dye it is possible to visualize an overall oxidizing environment in the QC, and make comparisons with the adjacent, rapidly dividing cells in the root meristem. Altering the distribution of auxin and the location of the auxin maximum in the root tip activates the QC, and cells leave G1 and enter mitosis. Commencement of relatively more rapid cell division in the QC is preceded by changes in the overall redox status of the QC, which becomes less oxidizing. (Jiang, et al, 2003)

the PIN network in root development

Auxin provide positional information. Cell polarity in the meristem disrupted by Auxin transport inhibitors (TIBA, NPA), vesicular traficking inhibitors such as brefeldin A, or actin inhibitors (cytochelasine ...) See mainly Blilou et al (2005), Friml et al (2002, 2003), Benkova et al (2004)

Local accumulation of the plant growth regulator auxin mediates pattern formation in Arabidopsis roots and influences outgrowth and development of lateral root- and shoot-derived primordia. However, it has remained unclear how auxin can simultaneously regulate patterning and organ outgrowth and how its distribution is stabilized in a primordium-specific manner. Blilou et al (2005) have shown that five PIN genes collectively control auxin distribution to regulate cell division and cell expansion in the primary root. Furthermore, the joint action of these genes has an important role in pattern formation by focusing the auxin maximum and restricting the expression domain of PLETHORA (PLT) genes, major determinants for root stem cell specification. In turn, PLT genes are required for PIN gene transcription to stabilize the auxin maximum at the distal root tip. Their data reveal an interaction network of auxin transport facilitators and root fate determinants that control patterning and growth of the root primordium.

The auxin gradients are established, maintained, modified or even completely reversed by a complex interacting network of auxin transporters distributed throughout the region. Evidence for the dynamic nature of auxin gradients has come from a series of seminal papers, the most recent of which [Blilou et al 2005] demonstrates the existence of a network of active auxin transport that recycles auxin around the root

Auxin and ethylene interaction

Ponce et al (2005) found Auxin and ethylene interactions control mitotic activity of the quiescent centre, root cap size, and pattern of cap cell differentiation in maize.

Cytokinins and root development

Gravitropism (Aloni et al 2004), development (Miyawaki et al 2003, Takei et al 2004) nitrate response, control of root meristem activity (Werner et al 2003)

Regulation of the root meristem activity by cytokinins

Cytokinin are essential for cell division, therefore very important for root meristem activity. Produced in root tip (Aloni et al 2004, Miyawaki et al 2004) and vasculature of root mainly. Altough essential for cell proliferation, cytokinin negatively regulate root meristem activity (exogenously supplied cytokinin inhibits root growth, and plant with a lower cytokinin content grow faster as shown by Werner et al (2003) in plant overexpressing cytokinin oxidase (CKX) genes (The CKX proteins degrade irreversibly cytokinin).

Role of cytokinin in the regulation of root gravitropism

The models explaining root gravitropism propose that the growth response of plants to gravity is regulated by asymmetric distribution of auxin (indole-3-acetic acid, IAA). Since cytokinin has a negative regulatory role in root growth, Aloni et al (2004) suspected that it might function as an inhibitor of tropic root elongation during gravity response. Therefore, they examined the free-bioactive-cytokinin-dependent ARR5::GUS expression pattern in root tips of transformants of Arabidopsis thaliana (L.) Heynh., visualized high cytokinin concentrations in the root cap with specific monoclonal antibodies, and complemented the analyses by external application of cytokinin. Their findings show that mainly the statocytes of the cap produce cytokinin, which may contribute to the regulation of root gravitropism. The homogenous symmetric expression of the cytokinin-responsive promoter in vertical root caps rapidly changed within less than 30 min of gravistimulation into an asymmetrical activation pattern, visualized as a lateral, distinctly stained, concentrated spot on the new lower root side of the cap cells. This asymmetric cytokinin distribution obviously caused initiation of a downward curvature near the root apex during the early rapid phase of gravity response, by inhibiting elongation at the lower side and promoting growth at the upper side of the distal elongation zone closely behind the root cap. Exogenous cytokinin applied to vertical roots induced root bending towards the application site, confirming the suspected inhibitory effect of cytokinin in root gravitropism. Our results suggest that the early root graviresponse is controlled by cytokinin. They conclude that both cytokinin and auxin are key hormones that regulate root gravitropism.

Cytokinin and root response to nutrients

Cytokinin and nitrate, phosphate nutrition. Regulation of a set of phosphate starvation genes by cytokinins. Increase in cytokinin content in presence of nitrate. Regulation of Some AtIPT genes by nitrate (Miyawaki et al 2004; Takei et al 2004).

From Franco-Zorilla 2005. Cytokinins control key processes during plant growth and development, and cytokinin receptors CYTOKININ RESPONSE 1/WOODEN LEG/ARABIDOPSIS HISTIDINE KINASE 4 (CRE1/WOL/AHK4), AHK2, and AHK3 have been shown to play a crucial role in this control. The involvement of cytokinins in signaling the status of several nutrients, such as sugar, nitrogen, sulfur, and phosphate (Pi), has also been highlighted, although the full physiological relevance of this role remains unclear. To gain further insights into this aspect of cytokinin action, we characterized a mutant with reduced sensitivity to cytokinin repression of a Pi starvation-responsive reporter gene and show it corresponds to AHK3. As expected, ahk3 displayed reduced responsiveness to cytokinin in callus proliferation and plant growth assays. In addition, ahk3 showed reduced cytokinin repression of several Pi starvation-responsive genes and increased sucrose sensitivity. These effects of the ahk3 mutation were especially evident in combination with the cre1 mutation, indicating partial functional redundancy between these receptors.We examined the effect of these mutations on Pi-starvation responses and found that the double mutant is not significantly affected in long-distance systemic repression of these responses. Remarkably, we found that expression of many Pi-responsive genes is stimulated by sucrose in shoots and to a lesser extent in roots, and the sugar effect in shoots of Pi-starved plants was particularly enhanced in the cre1 ahk3 double mutant. Altogether, these results indicate the existence of multidirectional cross regulation between cytokinin, sugar, and Pi-starvation signaling, thus underlining the role of cytokinin signaling in nutrient sensing and the relative importance of Pi-starvation signaling in the control of plant metabolism and development.

Auxin and cytokinin have overlapping patterns in the Arabidopsis root

Both have maximum concentrations in stele of mature root, columella and vascular initials, both relocalised in the lateral rootcap upon gravi-stimulation.

Cell division control in the RAM

Role of the quiescent center and the root cap in the maintenance of cell division and meristem size (Werner et al, 2003; Ponce et al, 2005) The mitotic cells can be visualized using genetic markers such as cyclin B1-GUS or GFP fusion reporters genes. The cyclin B is mostly expressed during the G2/M phase of the cell cycle.

Auxin dependent maintenance of an oxidative state of the quiescent center in maize.

Role of Argonaute in meristem formation. The ARGONAUTE gene family is involved in the regulation of gene expression via the RNAi Silencing Complex (RISC). microRNA (miRNA) are 20-22bp RNAs that direct RISC to target genes (Kidner and Martienssen, 2005).

Transcriptional framework for root patterning

http://www.bio.uu.nl/mg/pd/research/transcrframe.html

Root hairs

Root hairs are filamentous structures produced by a subset of cells on the root epidermis. The cells producing root hairs are called trichoblast cells whereas cells which don't are called atrichoblast cells They function to increase the surface area of the root, allowing more efficient water and nutrient uptake. The root epidermis consists of elongated cells of 2 cell types: hair cells and hairless cells. Positional cues specify the pattern of root hair development. Arabidopsis roots are particularly well suited for studying cell patterning. The root is small, with a 1 cell layered cortex, and a very regular pattern of cell division in the RAM that allows the lineage of each cell to be known. The cortex consists of an invariant ring of 8 cells. The epidermis consists of a variable number of cells but there is invariably 8 files of root hair cells. Root hairs always form on epidermal cells positioned over the radial cell wall between cortical cells. The control of cell pattern will be considered in more detail in a later lecture.

Root hair cells can be cytologically distinguished by their dense cytoplasms (delayed vacuolation) early in development, near the onset of elongation; thus cell fate has been specified prior to elongation. The root hair first becomes visible as a swelling at the apical end of the elongate epidermal cell. The protuberance then grows into a root hair by tip growth, similar to pollen tubes. It requires a Ca2+ gradient, golgi vesicle transport and vesicle fusion with the plasma membrane.

Root architecture

Root branching

Root hairs are responsible for nutrient and water uptake and provide physical support to the plant. Most of the root system is made of lateral roots that originate postembryonically. Lateral root development is controlled by different factors including nutrient concentration in the plant and the soil (Lopez-Bucio et al., 2003; Malamy, 2005, for a review). This plasticity allows adaptation of the root system to the soil, a very heterogeneous and changing environment, and is consequently very important for the survival of the plant (Grime et al., 1986; Hodge, 2004).

Like shoot branching, root branching can be either terminal or lateral, with the terminal mode being more common in lower plants and lateral much more common in angiosperms. Terminal branching involves the division of the RAM into 2 with the subsequent production of 2 roots. Lateral branching is different in roots than in shoots. Lateral roots initiate from internal cells of the pericycle. Initiation occurs in the late cell elongation/early cell differentiation zone, in pericycle cells that are partially to fully differentiated. Thus there is no detached meristem. A small group of pericycle cells reorient their axis of polarity to the radial dimension and begin growing and dividing to form a mound of cells. With continued growth and division, the mound of cells becomes organized into a RAM with root cap, while still within the tissues of the main root. Continued growth allows the lateral root to penetrate the endodermis, cortex and epidermis, finally reaching the exterior of the parent root.

Auxin and cytokinins are key factors in determining lateral root initiation. Exogenous auxin induces a extra lateral roots in a dose dependent manner (ie. more auxin, more lateral roots). Conversely, many auxin mutants show a paucity of lateral roots. Environment also influences lateral root development. Lateral roots in nitrogen deficient environments will respond to applied NO3- by elongating. This requires the function of a myb transcription factor called ANR1. Lateral roots of antisense ANR1 plants do not respond to nitrate. This also demonstrates that this is a specific response to the nitrate stimulus and not simply a nutritional response because the antisense plants are otherwise healthy.

Lateral root development

In Arabidopsis thaliana, lateral roots originate postembryonically from a small number of differentiated cells situated in the root pericycle in front of xylem poles called pericycle founder cells (Casimiro et al., 2001; Dubrovsky et al., 2001). As they exit from the root meristem area, pericycle cells in front of xylem poles undergo a round of cell cycle but stop in the G2 phase and remain in this state (Beeckman et al., 2001). Later in the development of the root, these cells can undergo a defined program of oriented cell divisions and expansion to form a lateral root primordium (Malamy and Benfey, 1997; Dubrovsky et al., 2001; Casimiro et al., 2003). Unlike primary root formation that occurs during embryogenesis, lateral root formation is easily accessible to observation and experimentation. Moreover, lateral root formation can be initiated by the application of the plant hormone auxin. Nevertheless, molecular mechanisms of root branching are still poorly understood.

Himanen et al (2002) have shown cyclin-dependent kinase inhibitor KRP2 plays a significant role in the regulation of lateral root initiation. KRP2 appears to regulate early lateral root initiation by blocking the G1-to-S transition and to be regulated transcriptionally by auxin.

Pericycle cell-cycle marking studies (Beeckman et al., 2001; Himanen et al., 2002) and cell type specific marking (Malamy and Benfey, 1997), each utilizing GUS staining, showed the earliest indications of lateral root initiation. In addition, Malamy and Benfey (1997) produced a cell lineage map of the median longitudinal view of the emerging lateral root primordium (LRP). The relative contribution of different pericycle cell files to the LRP is not clear, however various studies placed the number of LRP founder cells at 6–11 (Casimiro et al., 2003; Dubrovsky et al., 2000; Laskowski et al., 1995). Pericycle cells in longitudinal files in contact with the xylem differentiate as LRP founder cells. Generally, two files of pericycle cells are in direct contact with a xylem file but a minimum of three files of the pericycle are usually involved in LRP formation (Casimiro et al., 2003). Xylem-contacting pericycle cells show the first signs of differentiation as LRP founders but adjacent files also contribute to the LRP. Kurup et al (2005) used a lineage marking system combined with a live-cell outlining marker to demonstrate uneven contribution of the LRP pericycle founder files to the emerging lateral root.

Nodule development is thought to be a modified lateral root development

Nutrient Availability Influences Root Development

Several soil nutrients can alter root hair development (Reviewed in López-Bucio et al, 2003). Fe or P deficiencies both induce more epidermal cells to differentiate as root hairs, and both induce root hairs to elongate more than normal. These two pathways appear independent. The Fe pathway appears to function through the ethylene and auxin pathways because ethylene and auxin mutants show altered responses to Fe deficiency. P appears to be ethylene and auxin independent. The increased root hairs increases the surface area of roots, increasing their capacity to absorb limited nutrients.

Several nutrients can also alter root architecture by altering lateral root formation or growth, or by altering primary root growth. High nitrate inhibits lateral root elongation if the root system is uniformly exposed. However if only a portion of the root system experiences high nitrate while the rest experiences deficiency, the section with high nitrate will show elongated lateral roots. A MADS box transcription factor, ANR1 is induced by local high nitrate and is required for the root architecture response.

Nitrate supply increase cytokinin content in the root. Takei et al (2004) have shown the adenosine phosphates-isopentenyltransferase AtIPT3 is a key determinant of nitrate-dependent cytokinin biosynthesis in Arabidopsis.

Phosphate deficiency induces the formation of lateral roots and inhibits root elongation. The result is a dense, highly branched root system. This is compounded by the effect on root hairs. In addition, expression of phosphate transporter genes and other physiological changes result in a root system highly adapted for efficient uptake of P. The effects on root growth are brought on by inhibition of the cell cycle and by low auxin concentrations in the root apical meristems.

Sulfate deficiency also increases lateral root density. The NIT3 (nitrilase3) gene is induced and thought to increase auxin synthesis.

Moreover the cytokinin receptor CRE1/WOL/AHK4 has been shown to be involved in the interaction between phosphate-starvation, sugar, and cytokinin signaling in arabidopsis (Franco-Zorilla et al., 2005)

Several lines of evidence suggest the nutrient ions may act directly as signaling molecules. Mutants in nitrate metabolism still show the normal response to Nitrate. Root systems on plants with adequate phosphate show the classic P starvation phenotype in localized areas of deficiency. Thus the changes in root architecture are not secondary effects of altered metabolism, but appear to be primary effects regulated by the ions themselves.