Scripting a new dialogue between diazotrophs and crops

Abstract

Diazotrophs are bacteria and archaea that can reduce atmospheric dinitrogen into ammonium. Plant-diazotroph interactions have been explored for over a century as a nitrogen source for crops to improve agricultural productivity and sustainability. This scientific quest has generated much information about the molecular mechanisms underlying the nitrogenase function, assembly and regulation, ammonium assimilation, and plant-diazotroph interactions. This review presents various approaches to manipulating nitrogen fixation activity, ammonium release by diazotrophs, and plant-diazotroph interactions. We discuss the research avenues explored in this area, propose potential future routes, emphasizing engineering at the metabolic level via biorthogonal signaling, and conclude by highlighting the importance of biocontrol measures and public acceptance.

Addressing the nitrogen conundrum

Nitrogen (N) is an essential plant nutrient. Despite dinitrogen gas (N₂) comprising about 78% of the Earth’s atmosphere, it is not directly available to plants, and N is often a limiting nutrient in agricultural systems [1–3]. Low N availability affects many plant physiological processes, resulting in stunted growth, a decline in plant health, and an overall reduction in crop yields. N fixation is the reduction of N₂ into ammonium (NH₄⁺), and this chemical reaction requires large energy input to break the triple bounds of the N₂ molecule. In industrial N fixation, also known as the Haber-Bosch process, N fertilizers are synthesized using energy from natural gas. Since the Green Revolution, these synthetic fertilizers have been used widely in agriculture to overcome N limitation. Synthetic fertilizers are highly costly to growers, especially in developing countries, and relying on fossil fuels to produce them is unsustainable. Their intensive use also has many detrimental environmental effects, including the eutrophication of rivers, degradation of coastal zones, and greenhouse gas emissions [2]. For these reasons, cheap, sustainable alternatives to synthetic fertilizers are desired [2,4]. In biological N fixation, bacteria and archaea (henceforth diazotrophs) use the nitrogenase enzyme complex and ATP as an energy source to perform N fixation [5]. However, the nitrogenase enzyme is inhibited by oxygen and requires large amounts of energy in the form of ATP. Diverse diazotrophs live in the soil; some associate with plants to obtain photosynthesis-derived carbon. In some cases, the plant can even provide a low-oxygen niche that protects the nitrogenase enzyme from oxygen. In return, some diazotrophs can provide significant amounts of N to host plants.

Such efficient plant-diazotroph associations include root nodule symbioses, particularly rhizobium-legume associations. These symbioses lead to the development of new organs, the root nodules that provide a proper niche for N fixation and in which plant cells are infected intracellularly by alpha- or beta-proteobacteria called rhizobia or by Frankia actinobacteria. Some rhizobia-legume associations are extremely efficient as these crop hosts require little to no N inputs and even add N to the soil for subsequent crops through crop rotations. However, these associations only occur in some plants from the monophyletic “N-fixing” clade, including the Fabale, Fagale, Cucurbitale, and Rosale (FaFaCuRo) plant families.

Multiple approaches are being explored to increase the economic and environmental benefits of biological N fixation on crops, especially cereals requiring significant N inputs to achieve maximal yields. One exciting approach aims at expressing the nitrogenase enzyme directly in plants. Progress was made in expressing different components of the nitrogenase enzyme complex in plant organelles, with the chloroplasts and mitochondria regarded as the most suitable sites for engineering, due to their ability to provide the necessary energy and reducing power for N fixation. However, many challenges remain, such as the undesired instability of the proteins [6,7] (Fig. 1). Another approach is to engineer root nodule symbioses in plants outside the FaFaCuRo clade. Despite the progress made over the last decades in understanding the plant genes controlling early rhizobia-legume recognition, nodule organogenesis, and evolution of this symbiosis, the molecular mechanisms, such as those governing intracellular infection, remain poorly understood [3,8]. A third approach is identifying better non-leguminous hosts for N fixation in natural plant diversity. Maize landraces that develop aerial roots and produce abundant mucilage can obtain about half of their N from the atmosphere by hosting diazotrophs in this mucilage. This plant trait can be transferred to conventional maize accessions by classical breeding without transgenic approaches [9–11] (Fig. 1). A fourth approach to improve nodule-independent associations between cereals and diazotrophs involves engineering the plants to secrete a specific metabolite, which would selectively enhance the competitiveness of selected diazotrophs in the rhizosphere, for instance, by acting as an energy source. The rhizosphere is the zone of chemical, biological, and physical influence generated by root growth and activity [12]. This approach aims to make the plants better hosts for N fixation. Since free-living or associative diazotrophs do not altruistically share their fixed N with plants, they need to be manipulated to release the fixed N so the plants can access it [13–15]. Engineering ammonia excretion in diazotrophs can be achieved by disrupting some of the tight regulatory circuits required for the biosynthesis and activity of the nitrogenase enzyme (Fig. 1).

Figure 1: Strategies to enhance biological nitrogen fixation in non-leguminous crops and their challenges.

These approaches include expressing the nitrogenase enzyme in plant organelles, engineering root nodule symbioses in plants outside the N-fixing clade, exploring plant natural diversity to identify better host genotypes for diazotrophs, and engineering plants to secrete specific metabolites that can enhance selected diazotrophs in the rhizosphere.

Engineering nitrogen-fixing associations between crops and diazotrophs: candidate diazotrophs and desired traits

The interface between plants and microbes represents a region of a bidirectional flow of metabolites. In the soil, plant root exudates select specific microbes from the soil in the rhizosphere by providing them with carbon sources, mostly carbohydrates, dicarboxylic acids, and amino acids. The composition of plant exudate varies between plant species and even accessions and depends on a plant’s developmental and physiological status. While carbohydrates, dicarboxylic acids, and amino acids are generally observed in high amounts in the exudates, the levels of other metabolites such as auxins, coumaric acids, and salicylic acid are highly variable, dependent on the plant genotype, and change with the environmental context [16–18]. Specific metabolites, such as flavonoids, are observed in high amounts in some legumes but in low abundance in other plants. Plants actively take N from the soil, making the rhizosphere a low N environment that favors diazotrophs. The soil microbiota influences the composition of root exudates through direct consumption and affects plant growth, development, and defense reactions [19]. Flavonoids can be sensed by diazotrophs such as Bacillus subtilis and Herbaspirillum seropedicae [20,21]. Some diazotrophs, such as Azospirillum brasilense, have chemoreceptors that allow them to perceive various metabolites from root exudates [22]. A mutual influence via metabolites provides several avenues to engineer plant-diazotroph interactions. Physical proximity also matters as it provides diazotrophs with a high carbon (root exudates), low oxygen (plant respiration), and low N (plant uptake) environment. Proximity to a host plant also facilitates N transfer to the plant preventing loss of fixed N in the environment and provides some level of biocontrol (see below). Associative diazotrophs can also be epiphytes on plant surfaces, but some can be endophytes and live in plant tissues. Plant exudates elicit a more robust chemotactic response from endophytic bacteria than other bacteria in the rhizosphere [23]. For instance, Azospirillum brasilense, Azoarcus sp. strain BH72, and Herbaspirillum seropedicae associate with plant roots and express nif genes during their association with plants [24–26]. Azoarcus sp. strain BH72 was found in the stele, suggesting it can colonize the host systemically[27]. A comparison between Azoarcus sp. strain BH72 wild-type and mutant nifK, which lacked the nitrogenase enzyme, revealed that plants associating with the wild-type strain resulted in enhanced growth and increased N accumulation than the mutant. Azoarcus sp. strain BH72 seems to rely significantly on plants for survival, making biocontainment measures relatively simple [27]. Some associative diazotrophs have been described as intracellular, but these results are highly debated [28,29]. Diazotrophs are also present on and in plant leaves, but the amount of N transfer to the plant via this mechanism needs to be quantified more precisely [30]. Some free-living diazotrophs are excellent N-fixers and can thrive on soil carbon without depending on plants. This independence limits the ability of the plants to benefit from the high N fixation rate and could enhance biocontainment challenges.

In addition to N benefits, many plant-associated microbes also promote plant growth through phytohormone production (e.g., auxin, cytokinin, and gibberellic acid), degradation of the ethylene precursor 1-aminocyclopropane-1-carboxylate (ACC), phosphorous solubilization, iron acquisition, or protection against pathogens. Due to the low N environment, these Plant Growth Promoting Rhizobacteria (PGPR) are often diazotrophs. The plant growth-promoting effects of PGPRs often act synergistically with N fixation through increasing root biomass, which facilitates the uptake of N from diazotrophs [31].

How can we engineer better N-fixing associations? There are several factors which, ideally, should be collectively considered. Epi- or endophytes would be preferred targets due to the better transfer of fixed N to the plant. Particular attention should be given to the composition of plant exudates, as these diazotrophs may prefer specific plant metabolites. Root exudate metabolites can be a carbon source, chemoattractant, or protective compound. Flavonoids and betaines have been studied extensively in this context during root nodule symbiosis [32–34]. Strigolactones, important for mycorrhizal associations, influence rhizobial motility and likely shape the rhizosphere more broadly [35,36]. The chemical composition and consistency of plant mucilages above- and underground can enrich diazotrophic bacteria [9]. Plants could be engineered, or their natural diversity exploited, to identify the ones that preferentially produce higher amounts of selected metabolites. Similarly, the bacteria could be engineered to respond to a selected plant metabolite. Ideally, such metabolites should not elicit a generic response in bacteria; only the engineered ones should be able to respond.

Metabolites rare in the soil and present in root exudates are most desirable to engineer specificity in plant-microbe interactions. For instance, barley was engineered to synthesize the rhizopine scyllo-inosamine in the rhizosphere, which is naturally produced in nodules but not observed in the rhizosphere outside of this symbiotic context [37]. A range of bacteria have been engineered to sense specific metabolites, providing a means to improve plant-microbe associations. [16,38]. Several of these bacteria can fix N, naturally or through engineering, and associate with plants to various degrees [39–49]. Fig. 2 ranks the engineered bacteria by their potential to improve N acquisition by plants. At the same time, bacterial engineering need not be restricted to a single species. Engineered consortia of bacteria could be developed where individual taxa occupy different spatial or temporal niches. Together, they would work cooperatively to fix N for nearby plants to access. Some of these bacteria can synergistically enhance plant growth as described previously.

Figure 2: Bacterial biosensors that respond to specific metabolites.

Metabolites, host range, and qualitative levels in the root exudates are shown in green. The list of biosensors was obtained from previous publications [16,38]. * refers to cases where nif gene expression was engineered in Escherichia coli.

Candidate diazotrophs must be amenable to genetic transformations as well as be culturable under laboratory conditions. The engineered diazotrophs must also adhere to industry and safety regulations to gain traction and broader acceptance. This ultimately means using non-pathogenic strains with limited natural antibiotic resistance to prevent safety or biohazard concerns. These strains must also thrive in large-scale bioreactors to meet production demands for broader agricultural usage. Additionally, sporulating bacteria may be particularly effective due to the increased shelf life for bioactive products [50]. Therefore Gamma-Proteobacteria such as Klebsiella, Kosakonia, and Pseudomonas seem prime targets for such engineering. The development of new technologies for high-throughput engineering of other Proteobacteria such as Azospirillum and Herbaspirillum but also Firmicutes such as Paenibacillus and Bacillus offers exciting opportunities too.

Improving the fitness of engineered diazotrophs through bioorthogonal signaling.

N fixation is hugely energy-intensive for diazotrophs, requiring 16 molecules of ATP per molecule of ammonia produced. Therefore, the expression of nif genes is tightly regulated by N and oxygen in the environment [51]. In several gamma-proteobacteria, the key negative regulator of nitrogenase biosynthesis (nifL) precedes the nitrogenase activator (nifA) in an operon. One common strategy to increase N fixation is ergo the replacement of the nifL gene with a constitutive promoter. This genetic change removes the nitrogenase negative regulator while simultaneously increasing the abundance of the nitrogenase activator in N-rich environments [52–54]. Despite some successful use cases, constitutive overexpression of nifA can often have lethal drawbacks due to the high energy demands and the toxicity of ammonium accumulation [55]. In short, these modifications to the nifLA operon increase nitrogenase production by decoupling the nitrogenase activator from current cellular N levels (Fig. 3). Additionally, since the nifLA operon is present in many N-fixing gamma-proteobacteria, this modification can be transferred to many symbionts [56]. Ammonia-secreting strains have also been engineered through a point mutation or posttranslational inactivation of glutamine synthetase (glnA), both of which slow the incorporation of ammonium into amino acids [57,58]. Affecting ammonium incorporation rate can also extend to modifying P_II family proteins for measuring N levels in the cell (e.g., glnB, glnK) to prevent signal transduction that shuts off N fixation. However, these regulatory systems often overlap other facets of central metabolism and have adverse consequences on bacterial fitness [59,60]. A similar result may also be achieved by dysregulation of the draT/draG regulatory mechanism, removing the cell’s ability to sense current intracellular ammonia levels and switch off N fixation post-translationally [61]. Finally, deleting ammonium importer amtB ensures that these N-fixing bacteria do not recapture the ammonium that diffuses out of the cells [53,62]. The amtB deletion reduces futile cycling, improving fixation efficiency [63].

Figure 3: Strategies used to enhance nitrogen fixation in gamma-proteobacteria.

The number of PII homologs, which play a crucial role in N metabolism, can vary among proteobacteria. These PII proteins undergo modifications by the GlnD enzyme. This diagram shows a representative PII protein and highlights some of its regulatory functions. Under N-sufficient conditions (red arrows), the uridylyl-removing (UR) activity of the GlnD enzyme is active. The non-uridylylated form of the PII protein interacts with the ammonium transporter AmtB, inhibiting N intake. It also stimulates the adenylylation of the glutamine synthetase (GS), reducing ammonia assimilation. Recipocally, during N-limited conditions (green arrows), GlnD uridylylates the PII proteins (UTase). These uridylylated PII proteins are unable to interact with AmtB. They also stimulate the de-adenylylation of the GS. NtrB phosphorylates NtrC, activating the expression of σ⁵⁴ genes crucial for N metabolism. Additionally, PII proteins interact with NifL to activate NifA. Some common strategies used in engineering diazotrophs to enhance N fixation are (1) deleting nifL and over-expressing activator nifA. (2) mutating or posttranslationally inactivating the glutamine synthetase or decreasing its expression, and (3) deleting the amtB ammonium importer.

Models for N-fixing regulation have been developed in gram-positive bacteria, as well. For example, in the genus Paenibacillus, N fixation is activated in N-limited conditions by glnR. Under N-rich conditions, glnR is inhibited by the feedback-inhibited conformation of glutamine synthetase to prevent transcription of nif genes [64]. Upregulation of fixation in N-rich conditions can nonetheless be achieved in some species of Paenibacillus by the overproduction of alanine, which inhibits glutamine synthetase and mimics the intracellular environment of N-starved conditions [65]. However, no matter the regulation scheme, work with N-fixing bacteria focused solely on increasing N-fixation and ammonia secretion often decreases fitness [66]. Metabolic models for diazotrophs such as A. vinelandii DJ and Klebsiella oxytoca are promising for predicting fitness costs and N benefits delivered by engineered diazotrophs with various plant hosts and under different environments [60,66,67].

More competitive N secretion strains can be generated using additional regulatory circuits dependent on plant-microbe signaling. For example, research for engineering N fixation in native cereal microbes resulted in inducible signaling systems for nifA overexpression that depends on secreted metabolites from the plant root or fixation dependent on specific carbon sources [68,69]. Relying on signaling from plant-dependent small molecules would ensure that N is only fixed when the engineered strain is proximal to the desired crop species rather than constitutively [68]. A similar strategy could be applied to glnA inactivation, creating temporary, inducible shifts in ammonia incorporation speed using engineered unidirectional adenylyltransferase (uAT) [70,71]. The uAT removes the adenylyl-removing domain of the native adenylyltransferase and prevents activation of glnA. These metabolic modifications would limit the harmful effects caused by permanent overexpression; in these systems, cells perform energy-intensive fixation only when most beneficial to the crop.

With conditional induction in mind, known signaling systems can be engineered for high specificity and even provide a fitness benefit to engineered microbes. To this end, several plant-microbe signaling systems have been characterized and engineered. For example, elegant work with rhizopines has allowed for specific, synthetic induction of N fixation in Azorhizobium caulinodans [37,72,73]. If produced in sufficient amounts, these molecules could also be a potentially orthogonal carbon source in the cereal rhizosphere for bacteria specifically engineered to catabolize them [74,75]. Access to a privileged carbon source could confer a unique growth benefit on engineered strains to outcompete other soil microbes in the rhizosphere, which could help offset the high energy costs of N fixation [76,77]. Besides rhizopines, root exudates also contain carbohydrates (e.g., sucrose, rhamnose, arabinose) with well-characterized bacterial activators that can also be used as a carbon source and to induce fixation [23,78–81]. Several soil bacteria catabolize these carbohydrates, making them unlikely to provide a specific carbon niche, but they offer an alternative approach for difficult-to-engineer crop species [82–84]. Other small-molecule inducers produced in plants elicit well-characterized responses in bacteria, including flavonoids such as naringenin, luteolin, and apigenin or phenols such as vanillic and cuminic acid [85–87]. Unfortunately, in some cases the natural production levels for these molecules in root exudates is too low to be used as carbon sources and even possibly for detection by bacterial sensors. To solve this abundance problem, crops may need to be engineered to overproduce these signals constitutively, under N limitation, or in response to the perception of diazotrophs themselves. Nonetheless, bacterial induction’s flexibility offers many signals and signaling pathways to control plant-diazotroph interactions.

In addition to engineering bacterial responses, improving and titrating crop responses to engineered bacteria could allow for tightly controlled feedback loops between plants and engineered microbes. These systems can use the plants’ native ability to recognize microbe-associated molecular patterns (MAMPs) such as flagellar structures (flg22 epitope) or exopolysaccharides (EPS) through their pattern-triggered immune (PTI) system since these MAMPs vary between bacterial species. Modifying these MAMPs can help engineered strains avoid stringent defense responses [88,89]. Similarly, bacteria can be engineered to produce plant hormones such as ethylene or salicylic acid to activate native defense pathways against other pathogenic organisms [90,91]. However, while the reception and activation cascade for these molecules is well-understood in plants, the ability to modulate the defense responses for the expression of custom genes will require more investigation through plant synthetic biology [92–94]. Engineering strategies could then rely on harnessing and modifying native signaling systems in plants, which could be used to express genes to provide a synthetic niche for engineered microbes, or for specific induction of microbial N fixation (Fig. 4).

Figure 4: Common chemical signaling mechanisms used for synthetic associations between bacteria with engineered crops.

Metabolites such as carbohydrates, phenols, flavonoids, and rhizopines in root exudates can be used as signals to induce responses in engineered bacteria. Reciprocally, plants can recognize engineered bacteria through microbe-associated molecular patterns (MAMPs) or respond to phytohormones produced by bacteria to drive responses of interest. Some of these responses can be signal production towards bacteria to create positive feedback loops.

Once specificity in engineered diazotroph-crop interactions has been achieved, future research should focus on the ability to stack beneficial traits in the same microenvironment. However, this field has seen relatively little attention to date. For example, engineered strains can promote heat tolerance, drought tolerance, and uptake of scarce metals or phosphorus in crops by utilizing pathways already characterized in plant growth-promoting bacteria [95–99]. These circuits for plant induction in bacteria could be transferred to amenable hosts providing additional, titratable growth benefits to crops in suboptimal soil. Providing several benefits with engineered microbes would be desirable. However, improving plant resilience to other stressors often relies on bacterial production of beneficial small molecules, which would be another significant fitness burden on N-fixing cells. Developing microbial communities may be necessary to provide additional benefits, as it would spread the production load among several strains.

Biocontainment and public acceptance

When dealing with genetically modified organisms in complex open systems such as agricultural fields, it is essential to adopt effective biocontainment measures to limit the transfer of transgenic material to native microorganisms in the ecosystem. Currently, the risk assessment metric evaluates the engineered organisms’ escape frequency, and a rate below 1 in 10⁸ is considered safe by the United States Institutes of Health guidelines [100]. Still, containment at a field scale and under variable environmental conditions using non-model organisms has yet to be fully addressed [101].

Three major strategies for designing microorganisms with genetic safeguards have been used in laboratory settings: Trophic containment, induced lethality, and gene flow barriers (Fig. 5). Metabolic auxotrophy, or trophic containment, involves deleting essential genes responsible for producing crucial metabolites, making genetically modified organisms reliant on external sources to survive and grow. Nevertheless, this strategy may not be adequate in field conditions since these metabolites are commonly available in natural environments or can be produced by other organisms. To enhance the efficiency of this approach, the use of non-canonical amino acids in synthetic auxotrophy could provide an additional layer of protection by making the genetic coding incompatible with natural organisms and cutting off the horizontal gene transfer between them; this is referred to as “semantic containment”. However, the effect of introducing these in living systems is largely unknown at this stage [101–103]. The second approach involves using genetic circuits to control microorganisms by activating “killer” or “suicide” genes in response to specific environmental signals. This response is triggered by allosteric transcription factors that regulate gene expression [102,104,105]. Another type of gene circuit uses a two-layered approach to regulate gene expression. The first regulatory element represses the expression of the second regulator responsible for suppressing the production of a toxic output that triggers cell death. To survive, microorganisms need the first regulatory element to repress the expression of the toxic product [102,106]. While these systems are effective for a specific duration, spontaneous mutations pose a significant challenge unless there are multiple redundant killing mechanisms or a combination of different approaches. Finally, biocontainment strategies often involve the implementation of gene flow barriers to prevent horizontal gene transfer. However, these gene flow barrier strategies are not applicable in field conditions as they rely on plasmids. While each biocontainment strategy has limitations, implementing multiple layers of genetic safeguards may improve their effectiveness, especially if the goal is to use them in the field, with variable conditions, and using non-model organisms.

Figure 5: Common biocontainment strategies used.

These strategies might also be used in combination with one another. In the metabolic auxotrophy approach, specific genes responsible for producing essential metabolites are deleted from bacteria, making them dependent on environmental sources to grow. In this case, the bacterium cannot survive without metabolites produced by the plant. In the suicide gene approach, genes are designed to trigger the organism’s death in response to specific environmental signals.

Biocontainment critically contributes to the regulatory constraints and public acceptance of synthetic microbial communities and their use in agriculture. Public acceptance is a complex, ever-evolving area, and a detailed discussion is beyond the scope of this review. Its essential components are general public awareness, legal, political, economic, ecological, and ethical implications. Currently, there is no global standardization in regulating genetically modified microorganisms, which can differ significantly from one country to another. For example, in the United States, three federal agencies regulate agricultural biotechnology, and microbial products that use cis-genic diazotrophs are commercialized and used extensively by growers [107]. In contrast, in the European Union, all genetically modified microorganisms are regulated under the same directive, and they have strict regulations that prevent their commercialization and use in the field [108]. There needs to be transparent communication between scientists, breeders, growers, and consumers about the risks and benefits of these emerging technologies. Engineered bacterial consortia will require multiple rounds of evaluation for optimal performance. The design-build-test-learn cycle is foundational to the public acceptance of synthetic microbial communities in sustainable agriculture.

Concluding remarks

Engineering associative diazotrophs to provide N to crops is a promising and relatively quickly realizable solution to the high cost and sustainability issues associated with synthetic N fertilizers. Many of these bacteria, in addition to being diazotrophs, can provide various benefits to plants, such as plant growth promotion and stress tolerance. While a noble outlook, the practical use of plant-microbe interactions and their lab-to-land transition are still challenging due to the high variability of biotic and abiotic environmental factors and their impact on plants, microbes, and their interactions. Trials in highly controlled environments such as greenhouses often translate poorly to field conditions, and we propose that engineered strains should be tested more readily under highly replicated field trials. At the same time, current regulations make testing these genetically engineered microbes under field conditions challenging. We believe that extensive agronomic assessment of the economic and environmental benefits provided by these microbial technologies compared to synthetic N fertilizers, as well as biocontainment strategies to script the dialogue between crops and engineered diazotrophs, will be essential to the wider acceptance of these technologies.

Highlights:

• The availability of nutrients such as nitrogen affects crop productivity. Unfortunately, the intensive use of synthetic nitrogen fertilizers has detrimental economic and environmental consequences.

• Engineered associative diazotrophs hold immense potential to help many crops acquire nitrogen from the air via biological nitrogen fixation.

• The mechanistic understanding of plant-diazotroph interactions forms the basis of engineering these associations to increase the transfer of fixed nitrogen to the plants.

• Biorthogonal signaling is a promising approach to engineering specific, efficient, eco-friendly plant-diazotroph interactions.

• The composition of root exudates, the physical proximity of diazotrophs to the plants, their non-toxic nature, their efficiency in fixing nitrogen, and their amenability to genetic manipulations are essential for engineering associations between diazotrophs and crops for sustainable agriculture.

Outstanding Questions:

• How sensitive are the engineered bacterial consortia to environmental fluctuations and the physicochemical characteristics of the soil?

• How can the amount and composition of root exudates be manipulated without a significant trade-off with crop yield?

• In the engineered bacterial consortia, how can the fixed nitrogen be selectively transferred to the plant and not used up by the other bacteria?

• What interactions could be expected between the endophytic, engineered diazotrophs and the native endophytic communities?

• How will the engineered bacterial consortia influence other native microbial communities in the soil, particularly those in the nitrogen cycle?

• Can mycorrhizal fungi enhance the transfer of fixed nitrogen from the engineered diazotrophs to the plants?

• How can the lab-to-land transition more efficiently use engineered bacterial consortia for enhanced nitrogen fixation? What would be an ideal timeline? What are the biological, technological, technical, legal, and ethical constraints?

• Is it possible to develop worldwide guidelines for using engineered bacterial consortia in the field?