Biological approaches are considered eco-friendly as they are natural processes that do not produce any harmful by-products. There are several approaches for biological ammonia production including nitrogen fixation, nitrification, nitrate/nitrite reduction, urea hydrolysis, metabolic engineering of microorganisms, and in vitro ruminal microbial fermentation of protein biomass, but the most reported methods are biological nitrogen fixation (BNF) and metabolic engineering of microorganism methods. Biological ammonia production by rumen bacteria fermentation of protein biomass experimented on in this review is a relatively new approach and has shown the potential to complement ammonia bioproduction.
Biological nitrogen fixation (BNF) is a natural process that converts atmospheric molecular nitrogen (N
2) to ammonia (NH
3). BNF, an ATP-dependent reduction reaction catalyzed by nitrogenase enzyme, is responsible for approximately half of the bioavailable nitrogen that supports all life forms [
31]. Relative to the Haber-Bosch process which requires high temperature and pressure conditions to break down molecular nitrogen, nitrogen-fixing microorganisms produce ammonia at ambient temperature and pressure. Nitrogen-fixing microbes are robust and have been explored to produce biofertilizers in commercial quantities [
32,
33]. Researchers are actively making attempts to mimic the natural process of BNF by isolating nitrogen-fixing (
Figure 3) bacteria and nitrogenase for synthetic ammonia production. The major challenge with this research effort is that nitrogenase catalysis is highly energy dependent, making its reaction rate slower than most enzymes in nature [
34].
The main microorganisms that possess nitrogenase and carry out nitrogen fixation are the genus
Rhizobia which colonizes the root of legumes, and species in the genera
Azotobacter and
Klebsiella that can fix nitrogen without parasitizing plant roots. The latter group is the main focus of research for synthetic BNF [
35,
36]. Nitrogenase requires up to eight molecules of ATP to produce a molecule of ammonia in an anoxic condition. Although the reaction mechanism of nitrogenase is unclear due to its interrelated multiple subunits, scientists have attempted to construct a heterologous expression system for
Klebsiella nitrogenase subunits in
E. coli [
37]. Similarly, an heterologous expression of the Klebsiella nitrogenase gene cluster has been constructed in E. coli and yeast to understand the mechanism by which nitrogenase functions without oxygen as well as to increase its activity [
38,
39]. Various studies have also investigated how nitrogen-fixing bacteria can function under aerobic conditions without inactivating nitrogenase. Such researches involve the use of polysaccharide membrane to protect nitrogenase from oxygen exposure [
39,
40].
4.2. Cell and Metabolic Engineering for Ammonia Production
Various biomass including food waste, microbial biomass, and protein-rich crop residues can be fermented by engineered microorganisms whose metabolisms are well understood for ammonia bioproduction. In a metabolic engineering study on the conversion of protein wastes into biofuels and ammonia using microbes, codY gene (a transcriptional regulator), in Bacillus subtilis was knocked out. codY gene regulates the activity of several other genes involved in different processes, such as producing branched-chain amino acids (ilvABHCD and leuABCD), removing amino groups from other molecules (ybgE, ald, yhdC, appBC, and dppBC), and inhibiting the expression of genes that cause protein break down and uptake (yhdG, appBC, and dppBC). In bacteria, proteins are encoded for amino acid biosynthesis by the ilv-leu operon. The deletion of codY gene removed regulatory constraints on this operon causing a significant increase in the production and uptake of branched-chain amino acids (BCAA) due to the derepression of ilv-leu operon and subsequent upregulation of genes responsible for BCAA synthesis.
In addition to the deletion of the
codY gene, the
BkdB gene in Bacillus subtilis was also knocked out.
BkdB is a lipoamide acyltransferase enzyme that helps in the biosynthesis of branched-chain fatty acids by converting branched-chain keto acids into their acyl-CoA derivatives. This conversion inhibits the production of biofuels and ammonia.
BkdB gene knockout had a significant impact on the production of branched-chain fatty acids in Bacillus subtilis. Obstruction of the production resulted in increased availability of metabolic precursors for the production of biofuels and NH
3. To completely transform
B. subtilis to favor ammonia synthesis, an alcohol dehydrogenase gene,
leuDH, and two-keto-acid decarboxylase were overexpressed. LeuDH is an alcohol dehydrogenase gene that plays an important role in the conversion of amino acids to alpha-keto acids while two-keto-acid decarboxylase is an enzyme that catalyzes the decarboxylation of alpha-keto acids, which are important metabolic intermediates in amino acid biosynthesis. Overexpression of LeuDH increased the rate of amino-acid nitrogen reflux which helped to increase the efficiency of protein conversion. Similarly, overexpressing two-keto-acid decarboxylase led to the increased availability of metabolic precursors such as alpha-ketoisocaproate (
KIC) and alpha ketoglutarate (
AKG) for the production of ammonia. The resulting final strain of
B. subtilis was employed in the fermentation of protein biomass obtained from
E.coli cells. This process produced ammonia with a theoretical yield of about 50% [
45].
A similar study on ammonia production from amino acid-based biomass-like sources using engineered
E. coli has been reported [
46]. Since E.coli assimilates ammonia intracellularly [
47], the two genes involved in the ammonia assimilation pathway,
glnA and
gdhA which are both glutamine assimilation genes, were knocked out to enhance ammonia production.
glnA encodes for enzyme glutamine synthetase (GS) and catalyzes the conversion of glutamate and ammonia to glutamine while
gdhA encodes for enzyme glutamate dehydrogenase (GDH) and catalyzes the reversible conversion of glutamate and ammonia to alpha-ketoglutarate. Deletion of
glnA promotes the extracellular leaching of ammonia while the deletion of
gdhA increases ammonia flux to produce more glutamate, a known precursor of ammonia. In this study, deleting the two genes redirected the nitrogen assimilation pathways in
E. coli toward ammonia production, resulting in a peak titer yield of 458 mg/L equivalent to an overall yield of 47.8% [
46].
Further studies on the metabolic engineering of
E. coli for ammonia production converted different food wastes including soy sauce cake,
mirin cake, and tomato peel to ammonia. Using metabolic profiling to assess the correlation between substances in the media (amino acids, sugars, and organic acids) and ammonia production, glucose was implicated as an inhibitor of ammonia production. When glucose was added to the amino acid-containing medium at different concentrations, a negative correlation with ammonia production was obtained. Thus,
E coli was engineered to hinder the inhibitory effect of glucose by knocking out the transporter gene,
ptsG, and the phosphotransferase system which transports glucose and other sugars. Briefly, the polymerase chain reaction (PCR) technique was used to amplify and copy specific fragments of genes that encoded resistance to
pts'G-Kim and
glnA-Km (amplified from pKD13) using primers
ptsGF and
ptsGR. The amplified DNA fragments were then transferred into
E. coli cells through electroporation. Following the transfer,
E. coli cells were grown on LB agar containing specific antibiotics – ampicillin and kanamycin. This allowed only the cells that had taken up the amplified DNA fragments to survive and grow, while the others died off. By repeating this process with different combinations of DNA fragments and antibiotics, more varieties of
E. coli strains with different genetic modifications, such as
AptsG and
AglnA were created. To ensure that the modified DNA fragments had been inserted into the correct location in the
E. coli genome, PCR was used to amplify and sequence the insertion region using insertion-checking primers. The resulting
E. coli strain succeeded in producing ammonia in a glucose-containing amino acid medium, with up to 73% yield [
48]. In the studies described above, ammonia was, however, produced intracellularly. As a result, the produced ammonia can still be used up by these microbes for growth [
47]. Therefore, a system that can produce ammonia extracellularly without impeding microbial growth may improve productivity.
Studies on yeast for extracellular ammonia production have been attempted. Prominent among such studies is the use of yeast cell surface engineering (YCSE) systems to avoid ammonia toxicity and assimilation. In YCSE, the protein to be converted to ammonia is displayed on the cell surface usually by the attachment of a secretory signal to the N-terminus of the target protein and a signal sequence, an α-agglutin containing a glycosylphosphatidylinositol anchor, on its C-terminus. Briefly, the plasmid for yeast cell surface display of L-amino acid oxidase was constructed by synthesizing and inserting the codon-optimized sequence of the HcLAAO (L-amino acid oxidase) into pULDl, resulting in a plasmid named pULDl-HcLAAO. A strep-tag negative control plasmid called pULDl-s was also constructed. The yeast strain Saccharomyces cerevisiae BY4741/sedlA was utilized to display HcLAAO on the cell surface. The constructed plasmid was then introduced into the yeast strain. Yeast cells were then transformed and grown in a synthetic dextrose medium and cultured in SDC buffer at pH 7.0. Using this approach, up to10
6 target proteins could be displayed on the yeast cell surface which are then used as biocatalysts for enzyme immobilization [
47,
49,
50].
Ammonia production from soybean residues has been successful with the YCSE technique [
51]. Amino acid catabolic enzymes that produce ammonia from amino acid precursors such as ammonia lyases have attracted interest for their efficiency in being displayed on the yeast cell surface because their catalysis does not require cofactors, unlike nitrogenases. With yeast cells displaying glutamine ammonia-lyases, ammonia was produced from glutamine solution reaching a titer of up to 3.34g/L and efficiency of 83.2% [
51]. The limitation of this approach is that only glutamine of the 20 amino acids can be utilized. Interestingly, L-amino acid oxidase with a broad substrate specificity can be displayed for ammonia production from several amino acids [
52,
53]. These are lab-scale studies that may be difficult to transition to an industrial scale for eco-friendly biological ammonia production.
Table 1 shows a summary of the metabolic engineering route for biological ammonia production.
4.3. Hyper ammonia-producing bacteria route
The digestive compartment of ruminant animals, the rumen, is a biorefinery for ammonia production. Ruminal microorganisms can break down plant materials containing carbohydrates and proteins in their feeds for energy. The products of protein degradation including peptides and amino acids are metabolized to protein and/or ammonia. The microbial protein thus formed is required for animal products, but the ammonia is absorbed from the rumen, metabolized, and excreted in the urine. This is an inefficient use of dietary proteins with devastating consequences on the environment through environmental nitrogen pollution [
54].
Several studies in the animal sciences have sought strategies to promote microbial protein synthesis and regulate ammonia production. These studies revealed the identity of a certain group of bacteria whose rate of ammonia production is much higher than can be used up by the ruminal microbes for other functions including microbial protein synthesis [
55,
56,
57]. This group of bacteria, known as the hyper ammonia-producing bacteria (HAB) can effectively convert dietary protein to surplus ammonia [
58,
59]. This type of natural ammonia is produced when the digestive system of man and animal undergoes a biochemical reaction leading to the breakdown of nitrogen-containing amine (NH2) of proteins into ammonia or the ionic form (ammonium) is referred to as biological ammonia.
The first step towards the degradation of amino acids is deamination which is the removal of an amine group to convert it to ammonia. It has been reported that amino acid deamination in the rumen produces more ammonia than can be utilized by the bacteria [
60]. Deamination may occur through oxidation, reduction, hydrolysis, or removal of elements. It helps to free the carbon skeleton by removing the amine group from the amino acid. Furthermore, deamination could be carried out on either a single amino acid, pairs of amino acids as in the case of Stickland reaction, or a combination of amino acids and a non-nitrogenous compound with all resulting into ammonia and keto-acids as major products [
61].
The next biochemical reaction is called ammonification which is the second stage of mineralization [
62]. Useful energy can also be derived metabolically by bacteria and related microorganisms through ammonification. Ammonium (NH4+) is thus produced by microorganisms and if in excess, it is excreted into the environment as nutrients for uptake by plants or as feedstock for further nitrification [
62]. HABs have been implicated in converting ~50% of ruminal dietary protein to ammonia [
63,
64].
HABs are found in cattle rumen or swine manure stored in the pit [
65,
66,
67]. Additionally, HABs thrive in the rumen of hay-fed cattle compared to grain-fed cattle [
68] because the pH of hay-fed cattle rumen environment is relatively neutral, thus providing a favorable condition for their growth compared to the slightly acidic pH (<6.0) observed in grain-fed cattle [
69]. HABs are capable of producing up to 40 mM (0.6812 mg/L) of ammonia in peptone-amino acid medium depending on energy and carbon source [
66,
70]. HABs can operate in both anaerobic and aerobic environments, but anaerobic-HAB are more prominent and of major concern because they convert a large percentage of dietary protein in the rumen to ammonia. Although HABs are detrimental to ruminant metabolism due to excess ammonia generation causing toxicity to rumen microbes and hyperammonemia in farm animals [
71], they can be harnessed as a sustainable source for large-scale ammonia production with low energy requirements and zero emissions.
There are several strains of hyper-ammonia-producing bacteria (HAB) with different biological ammonia-production capacities.
Selenomonas ruminantium, Peptostreptococcus elsdenii, and
Bacteroides ruminicola are HAB strains that are capable of producing at least 1µM of biological ammonia on a lab scale through deamination.
S. ruminantium catabolizes cysteine hydrolysate, while
P. elsdenii breaks down casein hydrolysate and specific amino acids (L-serine, L-threonine, and L-cysteine) to produce biological ammonia [
72,
73]. Depending on HAB strain and environmental conditions, it is also possible to produce much higher concentrations of biological ammonia (> 24 mM) [
70]. HAB can operate in both anaerobic and aerobic environments, but anaerobic-HAB are more prominent and of major concern because they convert a large percentage of dietary protein in the rumen to ammonia [
65].