Review

A plant’s diet, surviving in a variable nutrient environment

See allHide authors and affiliations

Science  03 Apr 2020:
Vol. 368, Issue 6486, eaba0196
DOI: 10.1126/science.aba0196

Root growth regulation by requirement

Plant productivity depends on the elemental nutrients nitrogen and phosphorus, which are drawn from the soil. Oldroyd and Leyser review how root growth patterns adjust according to the physiological needs of the above-ground plant. Systemic signals, including small peptide signals, mediate communication between the shoot's needs and the root's supply.

Science, this issue p. eaba0196

Structured Abstract

BACKGROUND

Although plants are dependent on the capture of a number of elemental nutrients from the soil, the principal nutrients that limit plant productivity are nitrogen (N) and phosphorus (P). Acquisition of these nutrients is essential for crop performance, but levels of these nutrients in most agricultural soils limit productivity. Therefore, these nutrients are typically applied at high concentrations in the form of inorganic fertilizers to support food production. However, overuse of fertilizers allows environmental nutrient release, which reduces biodiversity and contributes to climate change. Many farmers around the world lack the financial resources to access fertilizers, and their crop productivity suffers as a consequence. A more sustainable and equitable agriculture will be one that is less dependent on inorganic fertilizers.

ADVANCES

Accessibility of N and P in the soil is affected by many factors that create a variable spatiotemporal landscape of their availability, both at the local and global scale. Plants optimize uptake of the N and P available by modifications to their growth and development and through engagement with microorganisms that facilitate their capture. Where N and P are ample, the ratio of root:shoot biomass allocation can be low, with minimal root systems capturing sufficient nutrients. Typically, vegetative growth is extended, allowing resource accumulation and investment in seed production. In environments where these nutrients are limiting, overall growth is reduced but root systems are expanded and colonization by microorganisms is encouraged to facilitate nutrient capture. Plants can recognize a patchwork of nutrient availability and activate root growth within this patchwork to optimize nutrient capture.

Plants are able to measure multiple facets of nutrient availability: local sensing of nutrients in the soil, roots experiencing nutrient deprivation, roots experiencing high nutrient availability, and the total nutrient requirements of the plant. Such sensing involves an integration of root and shoot signaling, with a variety of hormones moving between the root and the shoot to both signal nutrient availability and coordinate plant development. Such root-shoot-root signaling is essential to allow plants to make use of local nutrient patches, but to do so only when there is sufficient need for that nutrient.

Some microorganisms have capabilities for the capture of N and P from the environment. For instance, N-fixing bacteria can access nitrogen from the atmosphere, something that plants are unable to do. Arbuscular mycorrhizal fungi can access insoluble forms of phosphate in the soil that are mostly inaccessible to plants. Under situations where plants are unable to access N and P from their immediate environment, they turn to these microorganisms to find new sources of these limiting nutrients. Many of the processes that coordinate the plants’ developmental response to nutrient availability also regulate the plants’ interaction with microorganisms. These processes regulate the plants’ receptiveness to their microbial communities, promoting symbiotic associations and restricting immunogenic processes.

OUTLOOK

Although our understanding of how plants engage with nutrients has advanced, there are few examples of how such knowledge has affected plant performance, perhaps because much of our understanding derives from studies in model, not crop, plants. Years of breeding crops for success under high-nutrient environments have left us with some crop varieties that are poor at optimizing use of limited nutrients. Nonetheless, many processes exist in plants to ensure productivity under poor nutrient conditions, some of which are already accessible in the diversity of crop species and wild near-relatives. We are poised to use the knowledge generated in model systems to optimize the performance of crop plants under nutrient limitation.

N response and signaling.

Root responses of Arabidopsis plants grown in uniform high N (NO3; dark gray, left), uniform low N (light gray, middle), and differential treatments of high and low N (right). Note how the root responses are opposite to the local treatments in uniform versus differential treatments. Underpinning these responses are C-terminally encoded peptides (CEPs) produced in roots experiencing low N, cytokinins produced in roots experiencing high N, and an N-sufficiency signal in the shoot. All regulate shoot-to-root signaling, which involves CEP DOWNSTREAM 1 (CEPD) peptides. Systemic signaling is integrated with local signaling (indicated by red) that is induced by local perception of NO3.

Abstract

As primary producers, plants rely on a large aboveground surface area to collect carbon dioxide and sunlight and a large underground surface area to collect the water and mineral nutrients needed to support their growth and development. Accessibility of the essential nutrients nitrogen (N) and phosphorus (P) in the soil is affected by many factors that create a variable spatiotemporal landscape of their availability both at the local and global scale. Plants optimize uptake of the N and P available through modifications to their growth and development and engagement with microorganisms that facilitate their capture. The sensing of these nutrients, as well as the perception of overall nutrient status, shapes the plant’s response to its nutrient environment, coordinating its development with microbial engagement to optimize N and P capture and regulate overall plant growth.

Although plants are dependent on the capture of a number of elemental nutrients from the soil, the principal nutrients that limit plant productivity are nitrogen (N) and phosphorus (P). Acquisition of these nutrients is essential for crop performance, but low levels of these nutrients in most agricultural soils limit productivity. Therefore, these nutrients are typically applied at high concentrations in the form of inorganic fertilizers to support global food production. However, the global supply of P, which is sourced from rock mining, is finite and the manufacture of reactive N for fertilizer is currently dependent on fossil fuels to drive the energy-intensive Haber-Bosch process (1). Where farmers can afford fertilizers, their use is often profligate, and although this ensures crop productivity, it also creates problems because of the environmental release of these nutrients (2), which reduces biodiversity and contributes to climate change (3, 4). The opposing problem exists for small-holder farmers in developing nations, who generally lack the financial resources to use inorganic fertilizers, with their crop yields suffering as a consequence (5). A more sustainable and equitable agriculture will be one that is less dependent on inorganic fertilizers.

Because of the central importance of N and P availability in both natural and agricultural ecosystems, we focus this review on plant perception, acquisition, and response to these two nutrients. Where these nutrients are ample, the ratio of root:shoot biomass allocation can be low, with minimal root systems capturing sufficient nutrients. Typically, vegetative growth is extended, allowing resource accumulation and investment in seed production. In environments where these nutrients are limited, overall growth is reduced but root systems are expanded and colonization by microorganisms is encouraged to facilitate nutrient capture. We now have a detailed knowledge of the processes that control the plant’s response to N and P, and in this review we attempt to summarize this understanding. Virtually all of the results we describe are derived from model plant systems, and although there are ample opportunities to use this knowledge to improve sustainable productivity in agriculture, this requires extensive research in crops.

Root developmental adaptations to N availability

Plants are reliant on reactive forms of N, of which nitrate (NO3) is the most prevalent in soils, although ammonium and amino acids can also be present (6, 7). N limitation results in biomass allocation to the root at the expense of the shoot (Fig. 1). Where N supply is uniformly low, lateral root growth is promoted, and when N is uniformly high, lateral root growth is suppressed (Fig. 1) (8). However, when N supply is uneven, lateral roots proliferate into local patches of high N (9) and this N-foraging response has been well studied for NO3 (10). These effects are particularly clear in split-root experiments (11), in which the root system is divided at an early stage in development with different nutrient treatments supplied to the two halves of the root system feeding one plant (12). Lateral root growth is promoted when both compartments have low N but is suppressed in the low-N compartment and promoted in the high-N compartment in differentially treated plants (Fig. 1). The shoot is essential for mediating these root responses in split-root systems because the root response is lost after decapitation (12). These results contribute to a growing body of evidence suggesting that the root response to N involves four signaling processes: (i) local signaling in the root associated with perception of local N; (ii) root-shoot-root signaling indicating the presence of roots experiencing low [N]; (iii) root-shoot-root signaling indicating the presence of roots experiencing locally high [N]; and (iv) a systemic inhibitory signal that suppresses root N-foraging activities when shoot [N] is sufficient.

Fig. 1 N response and signaling.

Shown is a schematic of the root responses of Arabidopsis plants grown in uniform high N (NO3; dark gray), uniform low N (light gray), and differential treatments of high and low N, based on knowledge generated from split-root experiments. Note how the root responses are opposite to the local treatments in uniform versus differential treatments. Underpinning these responses are CEPs produced in roots experiencing low N, cytokinins produced in roots experiencing high N, and an N sufficiency in the shoot (inhibition indicated in left panel), all of which likely regulate shoot-to-root signaling—at least that which involves CEPD. Systemic signaling is integrated with local signaling in the root (indicated by red), which is induced by local perception of NO3 (see Fig. 2) to regulate the degree of root growth.

Mechanisms of N sensing and response

Signal 1: Local perception and uptake of NO3

Because NO3 is an important N source, plants have developed processes for NO3 uptake, as well as the ability to sense NO3 directly at the plant surface (13). One protein, Nitrate Transporter 1.1 (NRT1.1), a plasma membrane transceptor, contributes to both high- and low-affinity NO3 uptake and acts as a receptor for the perception of NO3 (1416). NRT1.1 controls responses to different NO3 concentrations (17). A switch in the NRT1.1 mode of action, which is controlled by phosphorylation (18), involves a suite of calcium-regulated kinases (14, 19) that dictate the dimerization state of the protein (20, 21). This switches this transceptor from a low- to a high-affinity NO3 transporter (18) and affects the modality of signaling (14, 17).

At least three additional classes of NO3 transporters contribute to low- and high-affinity uptake (22, 23). NRT1.1 controls the expression of these additional transporters, especially NRT2.1 (24), through a NO3-induced calcium influx into the cytosol and nucleus (Fig. 2) (25, 26). The calcium transient has two primary modes of action: further regulation of NRT1.1 (14, 27, 28) and induction of calcium-sensor protein kinases (Fig. 2) that phosphorylate the transcription factor NIN-like protein 7 (NLP7) (26). NLP7 is excluded from the nucleus, but NO3-induced phosphorylation drives NLP7 into the nucleus (26, 29), where it up-regulates the transcription factor Arabidopsis Nitrate Response 1 (ANR1), which promotes lateral root proliferation into NO3-rich patches (8) and NO3 transporters to drive further uptake of NO3 (2931).

Fig. 2 Local NO3 perception and signaling.

NO3 is both perceived and transported by the transceptor NRT1.1. Phosphorylation of NRT1.1 affects both transport and signaling and is controlled by a range of calcium-induced protein kinases (CIPKs). NRT1.1 signaling results in a calcium (Ca2+)–induced transient across the plasma membrane, which induces feedback regulation of NRT1.1 (red arrows) and downstream signaling (green arrows) involving a suite of calcium-dependent protein kinases (CPKs) that phosphorylate NLP7, allowing its transport into the nucleus. NLP7 interacts with a second transcription factor, TCP, which facilitates the integration of the local NO3 response with the systemic N status, likely through the action of CEPD. At least some of the targets of NLP7 are the transcription factor ANR, which coordinates root growth according to N availability, and high-affinity NO3 transporters such as NRT2.2.

Signal 2: Long-distance signaling to indicate local N depletion

C-terminally encoded peptides (CEPs) have emerged as regulators of systemic N signaling (32). These peptides are produced in roots experiencing N limitation and activate NO3 transporters such as NRT2.1 in roots, where NO3 is plentiful (Fig. 1) (33). CEPs travel to the shoot through the xylem, where recognition by CEP receptors (33) leads to the production of a second class of peptides, CEP DOWNSTREAM 1 (CEPD) (34). CEPDs are produced in leaves but function in roots, where they up-regulate NRT2.1 expression in locations where NO3 is ample (34). However, there is no preferential transport of CEPDs to roots expressing NRT2.1 (34), suggesting that additional signals must integrate with this systemic signal to drive the local response. We suggest that the CEPD signal likely integrates with the local NO3 response (signal 1; Fig. 2) to coordinate root growth into NO3-rich compartments. The transcription factor TCP-domain family protein 20 (TCP20) has a function in the systemic N response (35) by interacting with NLP6 and NLP7 (36) and is therefore a candidate to integrate the systemic and local N responses (Fig. 2).

Signal 3: Activating N foraging in N-rich patches

Cytokinin biosynthesis is promoted in roots perceiving high NO3 (Fig. 1) (37) and then translocated to the shoot, where it coordinates growth (38). Deficiency in cytokinin biosynthesis blocks the N-foraging response but can be counteracted by the addition of cytokinin even in roots under low N (12). This suggests that cytokinin acts as a root-to-shoot signal from patches of high N availability and coordinates with the CEP signal, because the effect of cytokinin only occurs when there are roots in low NO3 (12, 39). CEPD production may be directly or indirectly dependent on cytokinin (Fig. 1), allowing for integration of signals indicating locally low N supply with signals indicating locally high N supply.

Signal 4: Coordinating root growth with shoot N status

Plants also assess their overall N status (40) and use the information to modulate growth and metabolism, as well as to balance carbon (C) and N acquisition. Amino acids or NO3 may act as proxies for N status (4143). Although amino acids and NO3 contribute to N-status sensing (44), there is an intrinsic problem with using these alone to signal global N status. Doing so is equivalent to assessing one’s financial situation based only on the current account balance: projected income, expenditure, and savings all add to the demand for N and must be considered when calculating responses to nutrient environments. The N-sufficiency signal (possibly relative to the C supply) must integrate with the CEP and cytokinin signals to block N foraging when shoot N is sufficient (8, 44).

Root developmental adaptations to P availability

As for N supply, plants also adjust their root system architecture in response to P (45, 46), which is primarily available in soil as phosphate (PO4). As with NO3, there are different responses to PO4 whether spatially uniform or supplied heterogeneously. Although the roots of PO4-deficient plants proliferate into patches of high P (47) in a similar way to N (48), their responses to uniformly low P are significantly different. Under uniformly low PO4, primary root growth is repressed, lateral root growth promoted, and root hairs elongated, but the total root system length remains quite constant (47). These differences are consistent with PO4 having limited mobility in soil and the greatest availability in topsoil, whereas NO3 is freely mobile (49). Dense foraging in topsoil improves PO4 capture, whereas low-density exploration of a large soil volume improves NO3 capture. Such dynamic root responses are important for P capture: Suppressing lateral root production does not impede competitiveness under N limitation but does so under P limitation (50). PO4 has systemic effects on root development (47), but gene expression in split roots suggests that root growth is primarily regulated by local PO4, whereas systemic P signaling regulates genes associated with PO4 uptake and assimilation (51).

Mechanisms of P sensing and response

Responses to local sources of PO4

P regulation of root system architecture is driven by the local perception of PO4 at the root tip (52) and involves changes in multiple plant hormones (Fig. 3). Primary among these are complex patterns of auxin redistribution: increased auxin levels in primary root and young lateral root tips but decreased auxin levels in older lateral root tips (53, 54). These auxin dynamics function alongside other P-starvation signals to coordinate root development: strigolactones (5557); the peptide hormones Root Growth Factor 1 (RGF1) and RGF2 (58); DELLA-domain–containing proteins that accumulate under the declining concentrations of gibberellins (59); and ethylene, which controls root hair elongation (6063). These hormonal changes must together coordinate the root developmental responses to P availability, but the mechanisms of integration need to be established.

Fig. 3 Signaling P availability.

Shown is a summary of the signaling processes activated under P starvation (light gray) versus ample P (dark gray). Multiple hormones are affected by P availability, as indicated. Auxin levels are also affected in complex ways not shown in the figure. Some systemic signals associated with P availability are strigolactones, which regulate shoot growth, and microRNAs that control the levels of phosphate importers: Phosphate Transporter 1 (PHT1) and Phosphate 1 (PHO1). Under P sufficiency, PHR1 action is blocked by SPX-domain proteins through an interaction with phytate (IP6), and this suppression is released upon P starvation.

Systemic signaling under P limitation

Systemic signaling between the root and the shoot is implicated in P regulation of shoot growth and the regulation of PO4 transporters (45). PO4 itself is implicated as a systemic signaling molecule (64), but as outlined for N, additional signals are required for the plant to ascertain its full P status and needs. Strigolactones act as systemic P-availability signals (56), coordinating shoot growth (65), whereas cytokinins are down-regulated under P starvation (66) to remove a transcriptional repression (67, 68). The systemic P-starvation transcriptional response is induced by the transcription factor Phosphate Starvation Response 1 (PHR1) and its homologs (69, 70), which are regulated by PO4 through the availability of phytate (7173) (Fig. 3). PHR1 contributes to the systemic regulation of PO4 transporters in roots (7476) through the shoot-derived microRNAs (77) miR399 (7880) and miR827 (81, 82), which can be regulated by microRNA mimics (83).

Microbes to the rescue under nutrient deprivation

Plant engagement with microorganisms can improve nutrient acquisition: Mutualistic arbuscular mycorrhizal associations increase the surface area for PO4 and NO3 capture, whereas symbiotic N-fixing bacteria convert N2 into NH4+. The arbuscular mycorrhizal association dates to the earliest land plants (84). In these early land plants, the fungal association likely provided the primary interface for nutrient capture, facilitating the transition from an aquatic to a terrestrial lifestyle (84). Because of the early origin for this symbiosis, the arbuscular mycorrhizal association is widespread in the plant kingdom and intricately networked with the physiology of plant nutrient capture. The evolutionary processes that underpinned the emergence of arbuscular mycorrhizal associations also facilitated the evolution of other intracellular symbioses (85). However, independent losses of these symbioses, which are associated with loss of the underpinning genetic networks (8689), suggest a cost for symbiosis, with negative selection in environments or plant lifestyles where microbial nutrient services are no longer advantageous. Supporting these microbes’ energy needs (90), as well as the plants’ need to suppress immunity to facilitate symbiont colonization (91), may drive selection against these associations when they are no longer of value. The molecular components that guide plant developmental responses to the nutrient environment also coordinate engagement with microorganisms.

Regulation of N-fixing associations

Mutualistic bacterial associations facilitate N capture through the fixation of N2 by the bacterial enzyme nitrogenase. Many such N-fixing bacterial associations have evolved across the plant kingdom, ranging from intracellular colonization of specialized root organs known as nodules (92) to bacteria existing outside the root in polysaccharide matrices (93). The most effective associations, demonstrated by legumes, involve intimate engagement between intracellular N-fixing bacteria housed inside membrane-bound compartments within root nodule cells (92). Nodules allow a gaseous environment to be created that optimizes N capture and facilitates the integration of plant and bacterial metabolism, ensuring that ammonia is released to the plant (94).

Because of the high energetic costs of bacterial N fixation, this process generally only benefits the plant when it is starved for N but has ample C (95, 96); thus, the association is regulated according to both N and C availability (97). Nodulation is controlled by the CEPs and their receptors (98100), suggesting that as plants evolved N-fixing capabilities, they linked the decision to engage with these symbionts to existing low-N signals. In addition, plants evolved a second systemic signaling process to regulate the total level of N fixation, so-called autoregulation of nodulation (101) (Fig. 4). This system restricts further engagement with N-fixing bacteria after a first round of colonization but also regulates this process according to the availability of NO3 (101). Here, Clavata3-like (CLE) peptides are produced in the root after recognition of N-fixing bacteria and then transported to the shoot, where receptor recognition leads to the suppression of miR2111 expression, which acts as a shoot-to-root signal to inhibit a negative regulator of nodulation (101, 102). CEP peptides have been shown to positively regulate the levels of miR2111 (103). High levels of miR2111 create a symbiotically permissive state, whereas inhibition of miR2111 expression switches the plant to a symbiotically restrictive state (Fig. 4). This systemic signaling process is linked to local NO3 perception through NIN-like proteins that also activate the expression of CLE peptides in the root and can block aspects of nodulation signaling (104, 105). Thus, regulation of N fixation has integrated both local and systemic processes that control plant responses to N availability.

Fig. 4 Regulation of symbiotic associations.

Plants that enter intracellular symbioses have a symbiotically permissive state under low-N or low-P environments that is regulated by CEPs, the absence of gibberellins (GA), and the induction of miR2111. The target for miR2111 is Too much love (TML), a negative regulator of nodulation. Plants that are permissive to symbiosis perceive lipochitooligosaccharide (LCO) elicitors produced by arbuscular mycorrhizal fungi and N-fixing rhizobial bacteria. Strigolactones, which are produced under P starvation, and flavonoids act as plant signals to mycorrhizal fungi or rhizobial bacteria, respectively. Symbiotically restrictive conditions are generated as a result of either high nutrients or previous colonization by symbiotic microorganisms, leading to the accumulation of CLE peptides that suppress miR2111 expression through the Supernummery (SUNN) receptor.

Regulation of arbuscular mycorrhizal associations

Arbuscular mycorrhizal fungi are dependent on their host plant for C (90), making PO4 uptake through this association costly and thus regulated according to P availability (106). Evidence is emerging that the mechanisms controlling systemic autoregulation of nodulation also regulate mycorrhization (101, 107), implicating miR2111. Indeed, miR2111 is induced under P starvation (82). The role of miR2111 in regulating N fixation may be another manifestation of the recruitment of preexisting plant-mycorrhizal signaling to nodulation (85). Additional P starvation–induced components also regulate mycorrhizal colonization: DELLA-domain–containing proteins, which accumulate in the absence of gibberellins (the situation under low P), are positive regulators of mycorrhization (108), and strigolactones act as plant signals to mycorrhizal fungi (109), promoting the fungal processes necessary for the interaction. A signaling process related to that of strigolactones, but involving karrikin-like molecules, is essential for mycorrhizal colonization (110), but there is no evidence yet that karrikin-like signals are associated with plant nutrient status. It appears that multiple apocarotenoid molecules, including strigolactones and perhaps karrikin-like molecules, coordinate plant development and microbial associations (111), and there may be other unknown signaling molecules in this class of small metabolites. Once established, the level of PO4 supply by the fungus is monitored and if insufficient PO4 is forthcoming, then the plant blocks the fungal association (112). Similarly, N supply from N-fixing associations is monitored (113). Such constant substantiation of the delivery of nutrients helps to limit the emergence of microbial “cheaters” that might colonize the root to gain benefits without providing nutrient services.

Coordinating the root microbiome with plant P status

Plants engage with a diversity of microorganisms in the rhizosphere. Some of these commensal associations can facilitate nutrient acquisition. The demonstration that a species of Colletotrichum can confer plant fitness (114) highlights the delicate balance between mutualist and pathogenic lifestyles because many closely related species of Colletotrichum are plant pathogens. Facilitation of plant acquisition of PO4 can be through direct delivery of PO4, as in the case of C. tofeldiae (114), or may be due to the reprogramming of the plant’s PO4 uptake machinery upon fungal colonization, as proposed for Piriformospora indica (115). These commensal associations are also regulated by components that control the plants’ broader response to their nutrient status: C. tofeldiae infection and rhizosphere bacterial communities are regulated by P status in a manner dependent on PHR1 (114, 116). P starvation is associated with the suppression of immunity in Arabidopsis (116), implying that plants are willing to take risks under nutrient starvation to facilitate colonization by associative microorganisms to help in the capture of nutrients.

Integration

A focus on single-nutrient treatments is a key limitation of many of the studies that have led to our current understanding of the plant response to its nutrient environment. N and P availability are not alone in directing the degree of plant performance; other factors such as light, water, other nutrients, and the prevalence of pests and pathogens all dictate the way in which the plant partitions resources to optimize performance. Plants coordinate their developmental and microbial responses to N and P availability with their capacity to fix C. For example, light intensity can be relayed by the phytochrome B–responsive transcription factor Elongated Hypocotyl 5 (HY5), which acts as a mobile shoot-to-root signal that regulates N and P acquisition (117, 118). HY5 has functions in both root-developmental adaptations and the regulation of symbiotic N fixation (117, 119). Similarly, plant responses to water availability are modulated by the availability of N (120), and the water stress–signaling system that functions through abscisic acid directly affects the local root response to NO3 (19). The plant is reliant and responsive to a wide array of nutrients (121), and these must be measured in combination to optimize plant performance. N and P responses are integrated (122) and affected by the availability of a range of additional nutrients (123), but the underlying mechanisms for such nutrient coordination are only just beginning to emerge (124126).

Conclusions

Plants have adapted to survive spatiotemporally variable environments, including the variable landscape of nutrient accessibility. A combination of local nutrient sensing in the root with systemic signaling integrated in the shoot creates the response to the nutrient landscape, whether low or high in nutrient content or patchy or continuous. Under nutrient deprivation, plants turn to microorganisms to help facilitate nutrient capture through processes linked to developmental adaptations to nutrient availability. N and P are critical for global food production. Opportunities exist to adapt crop performance such that N and P use is improved even as environmental sustainability is enhanced (127, 128). Although our understanding of how plants engage with their nutrients has advanced, few examples exist where such knowledge has affected plant performance and this is almost certainly related to the fact that virtually all of our understanding, as described in this review, derives from studies in model, not crop, plants. Years of plant breeding under high-nutrient environments have left us with some crop varieties that are poor at optimizing use of limited nutrients (128). Nonetheless, many processes exist in plants to ensure productivity under poor nutrient conditions, some of which are already accessible in the diversity of crop species (127129). It is time we applied the deep understanding gained in model plants to impart improvements of crops to wean the global population from its dependency on inorganic fertilizers.

References and Notes

Acknowledgments: We thank H. Kool and Equinox Graphics for help in preparing the figures. Funding: The authors are supported by the Gatsby Foundation (GAT3272C and GAT3395/GLH) and by the Bill and Melinda Gates Foundation’s Engineering the Nitrogen Symbiosis for Africa (OPP1172165). Competing interests: O.L. acts on the board of the John Innes Institute and the council of the BBSRC and is a Gatsby Plant Science Advisor.

Stay Connected to Science

Navigate This Article