*Perspective* **New Breeding Techniques for Greenhouse Gas(GHG) Mitigation: Plants May Express NitrousOxideReductase**

**Jordan J. Demone 1,†, Shen Wan 1,†, Maryam Nourimand 1, Asbjörn Erik Hansen 1, Qing-yao Shu 2 and Illimar Altosaar 1,\***

1 Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, ON K1H 8M5, Canada; jdemone@uottawa.ca (J.J.D.); swa2@uottawa.ca (S.W.); mnourima@uottawa.ca(M.N.);aerik.hansen@gmail.com(A.E.H.)


† These authors contributed equally to this work.

Received: 4 August 2018; Accepted: 25 September 2018; Published: 27 September 2018

**Abstract:** Nitrous oxide (N2O) is a potent greenhouse gas (GHG). Although it comprises only 0.03% of total GHGs produced, N2O makes a marked contribution to global warming. Much of the N2O in the atmosphere issues from incomplete bacterial denitrification processes acting on high levels of nitrogen (N) in the soil due to fertilizer usage. Using less fertilizer is the obvious solution for denitrification mitigation, but there is a significant drawback (especially where not enough N is available for the crop via N deposition, irrigation water, mineral soil N, or mineralization of organic matter): some crops require high-N fertilizer to produce the yields necessary to help feed the world's increasing population. Alternatives for denitrification have considerable caveats. The long-standing promise of genetic modification for N fixation may be expanded now to enhance dissimilatory denitrification via genetic engineering. Biotechnology may solve what is thought to be a pivotal environmental challenge of the 21st century, reducing GHGs. Current approaches towards N2O mitigation are examined here, revealing an innovative solution for producing staple crops that can 'crack' N2O. The transfer of the bacterial nitrous oxide reductase gene (*nosZ*) into plants may herald the development of plants that express the nitrous oxide reductase enzyme (N2OR). This tactic would parallel the precedents of using the molecular toolkit innately offered by the soil microflora to reduce the environmental footprint of agriculture.

**Keywords:** radiative warming; atmospheric phytoremediation; N2O; nitrous oxide reductase; N2OR; *nosZ*; fertilizer; crop breeding; transgenic; GHG

#### **1. Introduction—Nitrous Oxide Continues to Bloom Unabated**

Atmospheric nitrogen (N) deposition is a pressing matter for climate change scientists concerned with the increasing danger that nitrous oxide (N2O), a noxious greenhouse gas (GHG), poses. Reactive nitrogen (Nr)—ammonia (NH3), nitrogen oxides (NOx), nitrates (NO3 −), and N2O—enters the biosphere from its original form of atmospheric N as at least three derivatives: gas, dry deposit, and precipitation (wet deposition) [1,2]. The sources of N2O are largely anthropogenic [3]. Many crops must receive N-based fertilizer to reach yield targets, which is supplied by inorganic fertilizers and animal manure [4]. In an effort to boost the yield in crop staples like wheat, corn, and soybeans, farmers apply N fertilizers at rates and times that are not always properly synchronized with crop demand [5]. While crops thrive when fertilized, experimental analysis has demonstrated that up to

40% of fertilizer N can be lost via leaching [6,7]. Other routes of N loss include soil erosion, NH3 volatilization and oxidation, and bacterial/fungal denitrification [8], although N losses through NH3 volatilization are higher than those via N leaching [9]. Around 62% of total global N2O issues from natural and agricultural soils, and the bulk of this production, mainly results from the processes of bacterial nitrification and denitrification [10].

Nr compounds enter the atmosphere through biological processes, but the invention of the Haber-Bosch process in 1908 was a critical moment for the sudden increase in Nr and GHG production globally [11]. This process of artificial N-fixation allowed for the large-scale reduction of N2 to NH3, producing massive amounts of synthetic N-based fertilizers that supported dramatic increases in high-yield farming [12]. This process now accounts for 80% of anthropogenic N-fixation (the remaining 20% resulting from combustion [13], with anthropogenic N-fixation in turn accounting for 60% of global N-fixation [14]). Haber-Bosch remains the industry standard synthetic N fertilizer today and as a result, has contributed to the ~2% increase in atmospheric levels of N2O [15,16]. This effect is also magnified by the global emissions of N2O produced by fossil fuel combustion [17] and the natural ability of legumes to fix N through symbiotic relationships with soil bacteria [18].

N2O is the third most prevalent GHG, behind carbon dioxide (CO2) and methane (CH4) [19]. The concentration of this gas in the atmosphere has been steadily increasing since the early 1900s (Figure 1), and it is 265 times more radiative than CO2 [19]. N2O also has an atmospheric lifetime of 121 years; by comparison, CH4 has an atmospheric lifetime of only 12 years, but CO2 also has a long half-life and can take anywhere from 20–200 years to be absorbed by the ocean [19], compounding the 'greenhouse gas' effect. Since chlorofluorocarbons (CFCs) were banned in 1989, N2O has become the leading cause of ozone layer depletion [20].

**Figure 1.** GHG levels since 1850. The green line represents the increase in CO2 concentration since 1850; the orange line represents the increase in CH4 concentration since 1850; lastly, the red line represents the increase in N2O since 1850 [19].

N2O emission results from the coupled oxidation and reduction of N performed by heterotrophic [21] (and some autotrophic) soil proteobacteria: (1) the nitrification pathway is catalyzed by autotrophs (*Nitrosomonas* spp. and other genera [22]) and also heterotrophs, and involves the oxidation of NH3/ammonium(NH4 +) to nitrite (NO2 −) [23] and nitric oxide (NO) [24]), which is followed by the oxidation of NO2 − to NO3 − by *Nitrobacter* spp. [25]; and (2) the denitrification pathway, whereby NO3 − is reduced to N2O and ultimately inert N2 gas [26]. As many as a third of soil bacterial species [27] lack the *nosZ* gene that reduces N2O to inert N2 [28], which leads to a sizeable amount of incomplete denitrification reactions and the subsequent buildup of N2O since it is an obligate intermediate [29]. This N2O diffuses out of the soil and into the atmosphere, contributing to the greenhouse effect, contaminating water, and leading to serious human health implications [30,31].

#### **2. Combating GHGs: Current N2O Mitigation Strategies and Limitations**

Demands for crop-borne food must be met, and so researchers must address the hazards of N-based fertilizers [32]. There are multiple N2O mitigation strategies either currently in commercial use or in development (summarized in Table 1).



BDI, biological denitrification inhibitor; BNI, biological nitrification inhibitor; EENFs, enhanced efficiency nitrogen fertilizers; SDI, synthetic denitrification inhibitor; SNI, synthetic nitrification inhibitor.


SDIs nitrapyrin [84], toluidine [54], and acetylene [44] all effectively mitigate N2O emission, albeit with toxic side-effects [55], and they do not technically inhibit nitric oxide reductase;


such as stable cultivation and control of the remediant organism and atmospheric exposure of the gas-cracking enzyme [97].

Atmospheric phytoremediation may ameliorate problems created by the other N2O mitigation strategies described. The concept here is to develop crops with the ability to "crack" N2O in both the soil and the atmosphere by incorporating the bacterial *nosZ* gene into their genomes. This gene encodes the nitrous oxide reductase enzyme (N2OR), an oxidoreductase that catalyzes the removal of N2O from the atmosphere, a process performed naturally by both denitrifying and non-denitrifying bacteria in the soil [98]. While conventional N2O mitigation strategies aim to control N2O production at earlier stages in the nitrification/denitrification pathway, this approach will target the atmospheric sum of N2O emitted by all sources (Figure 2).

**Figure 2.** Nitrification-denitrification pathway and overview of current N2O mitigation strategies. Orange arrows and lines show eight N2O mitigation strategies described in Table 1. Green arrows show nitrification and purple arrows represent denitrification reactions. BDI, biological denitrification inhibitor; BMPs, best managemen<sup>t</sup> practices; BNI, biological nitrification inhibitor; EENFs, enhanced efficiency nitrogen fertilizers; SDI, synthetic denitrification inhibitor; SNI, synthetic nitrification inhibitor; UI, urease inhibitor. **O** Encircled numbers refer to Table 1 strategies.

#### **3. Nitrous Oxide Reductase—An Orphaned Soil Protein?**

The *nosZ* gene can be categorized as either 'clade I' or 'clade II' based on sequence and *nos* operon organization, including the lack of an accessory *nosR* gene in the clade II members [99]. Clade II *nosZ* genes are also known as 'atypical' *nos* genes since they are found in non-denitrifying bacterial species. The N2OR enzyme that the clade II gene encodes catalyzes the same reaction performed by the clade I-encoded enzyme, but has a higher affinity for N2O [100], an important factor to consider when conceptualizing the development of an *nosZ*-expressing plant.

N2OR is a multi-copper protein encoded by the *nosZ* gene (which is accompanied by an operon cluster of additional genes (*nosRDFYL*) [101]) and is the only enzyme that can catalyze the conversion of N2O into N2. The first active N2OR was characterized from the soil bacterium *Pseudomonas stutzeri* and similar enzyme structures were resolved in bacterial species *Marinobacter hydrocarbonoclasticus* (formerly *Pseudomonas nautica*) (Figure 3), *Achromobacter cyclocastes*, and *Paracoccus denitrificans*. N2OR is a head-to-tail homodimer and each monomer contains two domains: an electron transferring domain (binuclear CuA centre) and a catalytic domain (tetranuclear CuZ centre) [102]. There is some variability between the species regarding CuZ bridging and cupric coordination in the catalytic centre, suggesting that N2OR substrate binding is species-specific. Regardless, the catalytic mechanism of N2O reduction in N2OR is still unclear [103].

**Figure 3.** Structure of *Marinobacter hydrocarbonoclasticus* nitrous oxide reductase (N2OR) homodimer. N2OR is organized as a head-to-tail homodimer. Monomers are coloured differently so that they can be distinguished. In both monomers, the N-terminal domain is dark-coloured. The N-terminal domain forms a seven-bladed β-propeller fold that coordinates the catalytic tetranuclear active site CuZ through seven histidine residues at its hub. The C-terminal domain forms a cupredoxin fold and binds the dinuclear mixed-valent CuA centre [104].

The proven ability of N2OR to "crack" the N2O molecule raises the question of why the protein has not ye<sup>t</sup> been incorporated into a commercially available transgenic cropping choice for environmentally motivated producers and small-plot farmers. Work has been done on this gene and its potential role in plant biotechnology since it was originally isolated in 1998 from the anaerobic soil bacterium *A. cyclocastes* [105,106], but it has ye<sup>t</sup> to be converted into a commercially valuable tool. In this sense, N2OR may be considered an "orphaned" protein, neglected among a veritable molecular toolkit of genes in the soil microflora [107,108]. Such forays into integrating soil and air sciences are demonstrative of the possibilities of what the soil microbiome offers biotechnologists [27]; it has already been discussed regarding the N-management possibilities offered by the microbiome and the current practice of 'bioprospecting' is also revealing a plethora of beneficial bacterial products, which is only accelerating thanks to whole-system approaches involving computational analyses [109].

Web of Science reports that between 1900 and 1991, there are no records binned under the combined topics "nitrous oxide reductase" and "microb\*". The scientific literature blossomed from its first occurrence of 1992 to the present day, witnessing at least 175 publications dealing with the science of this important enzyme in our total environment. The scientific community waited until 1996 to start discussing denitrification in a plant context, according to these same search terms. With the search terms "nitrous oxide reductase" and "plant", the scientific record shows that soil microbiologists have taken a growing interest in the movement of N into the atmosphere (Figure 4). It is encouraging to note that in the same time period, the linkage between N2OR and climate began its nascent phase.

3XEOLFDWLRQVE\NH\ZRUGYV QLWURXVR[LGHUHGXFWDVH 

**Figure 4.** Nitrous oxide reductase-related publications released since 1990 on Web of Science (Clarivate Analytics). Publications by key word vs "nitrous oxide reductase" from 1990 to 2018. The orange line indicates "nitrous oxide reductase" + "microb\*"; green: "nitrous oxide reductase" + "plant"; blue: "nitrous oxide reductase" + "climat\*".

#### **4. Catch Me If You Can: Can Plants Catalytically Convert N2O** *in planta***?**

Rather than a 'cat and robin redbreast' conundrum, we are confronted with an opportunity to deploy protein engineering to ensure that more N2OR molecules are attracted to the substrate binding site of the copper enzyme. Protein engineering offers ways to sidestep the challenges of expressing a complex bacterial protein in a plant [110]. There are potential issues with a recombinant metalloprotein like N2OR, such as whether the ABC transporter can assemble within a plant cell, or the plant can incorporate copper into the electron transferring and catalytic domains [111,112]. It is possible to re-engineer N2OR and produce a functional product [66], so there is precedent for designing an artificial metalloenzyme through rational protein design. This approach may be key to engineering a plant-compatible N2OR protein.

A principle challenge associated with imparting N2OR functionality to plants is that transforming the *nosZ* sequence alone may not be effective [113]; in *P. stutzeri*, the transcription of *nosZ* was dependent on the *nosDFY* genes being expressed, as they encode components of a putative ABC transporter system for the biogenesis of the CuZ centre [114]. Therefore, catalytically active N2OR may not be produced when only *nosZ* is expressed in a heterologous host [28]. Nevertheless, a model N2O-expressing plant has been engineered [64,93]. The clade I *nosZ* gene from soil bacterium *Pseudomonas stutzeri* was successfully expressed in a heterologous system—in this case, the tobacco plant (*Nicotiana tabacum*). In those proof-of-concept experiments the *nosZ*-expressing tobacco plants reduced 826 μg N2O/min/gram of leaf tissue [115]. Assuming the tobacco yield to be 0.50 tonne/ha [116], the calculated N2O-cracking ability of the *nosZ*-expressing tobacco could be as high as 600 kg of N2O/ha/day [115], or 60 tonnes/ha/year (100 day growing season). This value surpasses the calculated N2O flux of 0.05–1.98 kg N2O/ha/year [117]. In other words, if every tobacco plant in the world produced N2OR, this industrial crop (6.6 million tonnes were produced worldwide in 2016 [118]) could conceivably crack 785 Tg of N2O (1 Tg = 1 million metric tonnes) during an average growing season of 100 days, far surpassing the estimated ~30 Tg of N2O emitted per year [119]. Such catalytic capacity would give the 'Stop Smoking' campaigns a whole new flavour.

*Climate* **2018**, *6*,

Although these transgenic plants produced a functional N2OR enzyme, no gas analysis was performed to quantifiably ensure that these plants could reduce N2O to N2 using a recombinant N2OR. In the future, it is imperative that such analyses be performed to properly judge the efficacy of such a gene-stacking trait system for atmospheric phytoremediation.

An associated issue rests with *P. stutzeri* being an anaerobic species that produces enzymes that function optimally in a low-oxygen environment. While expressing *nosZ* in plants to reduce N2O appears to be an elegant solution, the N2OR enzyme was not evolutionarily engineered to be functional in the presence of oxygen. Most soil bacteria that produce N2OR do so in an anaerobic environment [102].

In the past five years, studies have identified several prokaryotic species that may express an oxygen-compatible N2OR. Aerobic N2O reducers may be undertaking an important role in mitigating the amounts of N2O emitted to the atmosphere in events of oxic-to-anoxic transitions, but these systems have not ye<sup>t</sup> been validated in plants. Here, we discuss two candidates for an oxygen-compatible *nosZ* expression system: clade II-*nosZ* member *Gemmatimonas aurantiaca gen nov.*, spp. *nov.* strain T-27, a polyphosphate-accumulating soil aerobe that is strongly represented in many oxygen-rich soil samples [120]; and *Azospira oryzae*, another clade II N-fixing bacterium originally isolated from the roots of rice (*Oryza sativa*) [121]. N2O reduction by the *G. aurantiaca* strain T-27 was observed in both the absence and presence of oxygen [120]. The inability of this organism to consume N2O in the complete absence of oxygen and the observed oxygen-induced activation of *nosZ* expression compels one to consider *in planta* overexpression, whereby the diurnal fluctuation of photosynthetic oxygen production may offer an egress for N2O accumulation. The *A. oryzae* strains I09 and I13 also show more rapid N2OR recovery rates and tolerance against oxygen inhibition than *P. stutzeri* [121] and so may be appropriate candidates for crop plant transformation and N2OR expression.

If the ideal *nosZ* sequence were to be identified and transformed into commercially important crop plants, the benefits would be numerous and profound: seed-borne GHG technology foresees the transgenic cassette passed on from generation to generation, meaning that constant application of the beneficial catalyst would not be required (as with NI application and rhizoremediation); the expression of *nosZ* in the aerial tissues of the plants allows the reducing enzyme to confront N2O much more easily than when the enzyme is expressed in the soil.

#### **5. Novel Breeding Task: "Gas Cracking" Plants**

The challenge of expressing heterologous bacterial proteins in plants necessitates codon optimization due to differences in GC content and codon bias with eukaryotes [122]. Altering the codon bias (or applying 'directed evolution' [123]) of a bacterial gene to be expressed in plants has been highly successful: *P. stutzeri nosZ* in tobacco [115], 5-enolpyruvylshikimate-3-phosphate (EPSP) synthase from *Agrobacterium tumefaciens* in Roundup Ready crops [124], and *Bacillus thuringiensis Cry* genes in maize [125] and rice [126]. Indeed, the global advance promulgating engineered crops is pillared on today's artificial intelligence-guided plant codon optimization rules offered by both large and small boutique DNA houses. However, there has been success expressing native bacterial sequences in plants, i.e., in the case of cotton expressing the native sequence of the *P. stutzeri* gene *ptxd* (*PHOSPHONATE DEHYDROGENASE*) [127,128]. One can dare to fathom how a universally-functional *nosZ* expression system could conceivably redirect some aspects of GHG mitigation research. Such a plant transformation cassette could theoretically be applied to any plant—wheat, rice, soybean, peat moss [129]—recruiting these species for the purpose of denitrification mitigation.

Even with an effective *nosZ* expression system, there are additional challenges in developing *nosZ*-expressing plant lines. There are relatively few powerful monocot-optimized expression systems available [130] (although *Bt* corn, LibertyLink wheat, and Roundup Ready wheat can attest to the effectiveness of the 35S promoter system in monocots), and there is difficulty in transforming monocots [65]. With the advent of new plant transformation technologies like the soil bacterium *Ochrobactrum haywardense* [131] and the *BABYBOOM*/*WUSCHEL2* system [132], the production of genetically modified crops with stacked or pyramided GHG genes may be expedited in the near future.

#### **6. Conclusions—Challenges to the Future Success of** *nosZ*

We must address what may be the greatest challenge of all for the modern molecular plant breeder: convincing the general public that transgenic crops may be beneficial for all the plant-planet's denizens, as modified crops that enter the food stream may appear unpopular in some boroughs. Regardless, there is a clear, urgen<sup>t</sup> need to control soil N2O losses due to the detrimental effects of this potent GHG in the atmosphere. Climate-smart crops should be given a crack at directly addressing this issue and tackling climate change. Such GHG-reducing plant lines, endowed with the ability to catalytically "crack" N2O in the air, could be vital in the battle to shift public perception towards the acceptance of "GMOs" in agricultural research.

Involvement of N2O in climate change and global warming has been the subject of increasing investigations due to its potential heat-trapping properties [3]. N2O emission from soil is primarily the result of an incomplete enzymatic reaction which is mediated by the bacterial enzyme, N2OR [98]. Therefore, in the late 1990s [105,106], the development of N2OR-positive transgenic plants was proposed as an environmental phytoremediation strategy with promise to remove N2O from soil and the atmosphere (Figure 2). However, producing a foreign protein in a plant cell is often a serious challenge. For example, different codon usage [133] and cellular properties between eukaryotic and prokaryotic cells are considered as unknown aspects of this strategy. At least two key questions need to be addressed in future studies to probe the probability for success of this green gene de-toxic tactic for accelerating the destruction of nitrous oxide via canopy catalysis: (1) Which candidate is the best source-organism to donate *nosZ* sequence for plant transformation? Activity of bacterial N2OR is associated with the anaerobic conditions in soil [101], whereas the plant cell is mostly an aerobic environment. Photosynthesis and respiration cause different levels of oxygen content in plant cells in a diurnal cycle which is not consistent with the enzymatic activity of N2OR in anaerobic soil bacteria. Therefore, selecting obligate or facultative aerobic bacteria containing active N2OR enzymes as 'the source code' would be pivotal; (2) Which plant cell compartment is the best destination for targeting N2OR accumulation? The native enzyme N2OR in bacteria is directed to the periplasm, where Cu chaperones provide enough Cu for the assembly of metal centres [134]. The absence of periplasmic space in plant cells reinforces the notion that subcellular localization of N2OR may influence its enzymatic activity *in planta*. Moreover, the important role of Cu in the functional assembly of N2OR posits whether the transformation of bacterial *nosDFY*, along with *nosZ*, is essential for a functional enzyme. Urgent exploration of how the cellular pool of metal nutrients and proteins (pseudo chaperones) in eukaryotic cells may suffice to activate N2OR *in planta* may compel the use of such climate-smart plants.

**Author Contributions:** J.J.D. constructed the initial paper draft; Q.-y.S., J.J.D., S.W., A.E.H., M.N., and I.A. contributed to study design and data analysis; J.J.D., S.W., and I.A. wrote the paper with input from A.E.H., and M.N.; M.N. constructed Figure 2; Q.-y.S. conceptualized the original approach to amplifying soil denitrification *in planta*; I.A. coordinated and supervised the project; all authors contributed to and approved the final draft.

**Funding:** This research was funded by a gran<sup>t</sup> from the Government of Ontario in 2018. Q.-Y.S. is a grateful recipient of The Rockefeller Foundation's Career Biotechnology Fellowship.

**Acknowledgments:** We thank three anonymous reviewers for taking the time to critique and strengthen this manuscript. We are grateful to Quinn Ingram for technical expertise with plant growth.

**Conflicts of Interest:** The authors declare no competing interests.
