Skip to main content
SearchLoginLogin or Signup

De novo nitrogenase to facilitate engineered nitrogen fixation in plants

Published onNov 16, 2023
De novo nitrogenase to facilitate engineered nitrogen fixation in plants

This problem statement is a community effort, and we invite comments, criticisms, and suggested edits.

Origin: This problem statement is based on ideas presented by Steven Singer, Program Director at ARPA-E, at a workshop held by Homeworld Collective on Oct 10, 2023. Speakers presented ideas on how protein engineering could be applied for climate and sustainability.

Contributors: Steven Singer, Paul Reginato, Ariana Caiati (and you?)

Lead contact: Paul Reginato ([email protected])


Nitrogen is naturally fixed only by diazotrophic (nitrogen-fixing) bacteria and archaea. A minority of plants, e.g. legumes, can obtain nitrogen through mutualism with diazotrophs. Nitrogenous fertilizers are therefore required for cultivation of most food and energy crops. Use of synthetic nitrogenous fertilizer contributes substantially to global GHG emissions, with ~0.45 GtCO2e/yr emitted through manufacturing and ~0.66 GtCO2e/yr emitted post application to fields, mostly as N2O (Menegat, 2022). While methods are under development to decarbonize the Haber Bosch process, for example using electrochemistry, those methods still have high energy demands, cannot resolve post-application emissions, and also cannot resolve the harms of fertilizer pollution via runoff and soil degradation. 


In principle, it is expected to be possible for plants to be engineered to fix their own nitrogen through expression of a functional microbial nitrogenase complex (Guo, 2023), thereby obviating exogenous fertilizer and completely mitigating its harmful impacts. In practice, however, efforts to engineer diazotrophy in plants have failed due to the metabolic complexity of assembling nitrogenase, which involves numerous (>=10) proteins and multiple distinct biosynthetic pathways to synthesize two separate cofactors (Guo, 2023). The oxygen-sensitivity of nitrogenase further complicates efforts to engineer plant diazotrophy, although several strategies may offer oxygen protection (Bennett, 2023), including localization to organelles such as mitochondria (Oldroyd, 2014, Bùren, 2017, Guo, 2023), which contain low oxygen due to respiration;  or chloroplasts (Guo, 2023), which possess mechanisms for protecting oxygen-sensitive proteins (Oldroyd, 2014,). Simplification of the nitrogenase assembly pathway to involve fewer distinct components and steps could enable more straightforward engineering of nitrogen fixation capabilities in plants.


Rapid ongoing progress in de novo protein design (Verkuil, 2022, Goverde, 2023, Chalkley, 2021, Watson, 2023) may soon provide a robust set of tools for forward engineering of complex protein functions as an alternative to transgenic expression of metabolic pathways. For example, de novo design has shown remarkable progress toward engineering a simplified photosynthetic system, which involves electron transfer via redox of metal clusters and a multi-cofactor protein, similar to nitrogenase (Ennist, 2022). Work should begin to engineer a de novo nitrogenase assembly pathway that involves fewer distinct components and pathways than natural ones. While this approach challenges the present frontiers of de novo protein engineering, success would provide a streamlined way to engineer plant diazotrophy.

Additional information

Open Questions

The significance could be made more clear by answering:

  • What evidence is there that nitrogenase could be protected from oxygen by localization to the chloroplast? Oldroyd, 2014 is the citation given in recent nitrogenase reviews for suggesting chloroplast localization, based on the fact that chloroplasts include some oxygen-protected proteins. However, chloroplasts also generate oxygen. Are there more recent or detailed descriptions of why chloroplast localization could provide oxygen protection?

  • What are the anticipated relative advantages of engineering diazotrophic plants vs engineering symbiotic relationships between non-legume plants and microbial diazotrophs?

The goals could be made more granular by answering:

  • What are some expected sub-problems within  de novo nitrogenase engineering? For example, how advanced is de novo synthesis of the nitrogenase metal cluster cofactors compared to the cofactors engineered for other metalloproteins?

  • Is there more granular guidance as to how these efforts might begin? For example, would it be ideal to conduct initial engineering work in yeast mitochondria?


  • It is assumed that a sufficiently low-oxygen environment can be found or constructed within plants in which nitrogenase could function.

  • For nitrogen fixed within organelles, it is assumed that fixed nitrogen can diffuse or be transported from those organelles to its site of use.

  • It is assumed that nitrogen fixation does not have unintended negative impacts on plant growth or organelle function.

Other information

  • Symbiosomes are compartments housing endosymbiotic nitrogen-fixing bacteria (de Faria, 2022). Perhaps synthetic symbiosomes could house nitrogenase in plants.

  • Work is also underway to engineer nitrogen fixation for non-legumes via symbiotic relationships between non-legume plants and microbial diazotrophs (Guo, 2023, Bennett, 2023)

No comments here
Why not start the discussion?