Rumen bacteria 1,12 and protozoa (Nasrollahi et al., 2020) fractionate N isotopes thus contributing to the natural 15N enrichment of ruminant proteins over the diet and explains to some extent the biological basis relating Δ15Nanimal−diet to NUE in ruminants 6,7. Results from the present study confirm that the higher Δ15Nanimal−diet usually observed in ruminants fed high vs low N diets 13 and related to low NUE 6 would arise to some extent from the rumen bacterial activity. However, while this finding seems to be true when bacterial strains (Exp #1) were grown on organic N substrates such as tryptone, it was not always the case when ammonia was the main N source. Furthermore, although no significant interaction was observed between the N level and the N source in the batch rumen mixed culture trial (Exp#2; P = 0.17), numerical trends pointed to a greater N isotopic depletion (lower Δ15Nbacteria−substrate) when the N source in mixed culture was an inorganic vs organic substrate (-1.15‰ vs -0.25‰ difference between the highest and lowest N level, respectively). Our results agree with previous studies reporting different N isotopic signatures in rumen bacteria when shifting from organic to non-organic N substrates at similar N levels 1. We highlight a clear opposite trend in bacterial N isotopic fractionation in relation to the bacterial ability to use ammonia-N, with 15N depletions only occurring when ammonia-N is taken up by rumen bacteria. Indeed, when using a cellulolytic bacterial strain such as Fibrobacter succinogenes S85 (Exp #1) or alternatively a mixed cultured batch enriched for a cellulolytic bacterial population (Exp #2) greater ammonia-N levels promoted lower isotopic N discrimination. The conceptual model developed by Dijkstra et al. 14 pointed to the balance between N assimilation (uptake) and dissimilation (release) as a key determinant of the N isotopic signatures of soil microbial biomass. Our results may further indicate that N isotopic signature in rumen bacteria may also change according to the balance between synthesis of microbial protein from ammonia versus non-ammonia N sources. In this regard, Collins et al. (2007) found that Escherichia coli grown on organic nitrogen sources released NH4+ and were enriched in 15N (relative to the nitrogen source) while it was highly depleted in 15N when grown on an inorganic nitrogen source. The ability of rumen bacteria to use peptides or ammonia under different nutritional contexts seems thus to determine the extent of N isotopic fractionation in the rumen.
Several metabolic pathways could be involved in the N isotopic fractionation occurring in the rumen. Although N isotopic fractionation by bacteria could theoretically originate from transport, assimilation, transfer, synthesis and excretion of N compounds 15, current knowledge suggest that the main fractionation pathways are related to bacterial ammonia assimilation and release (Collins et al., 2007; Dijkstra et al., 2008), the latter as a result of peptide degradation and amino acid deamination. Some of the bacterial enzymes involved in ammonia assimilation (in ruminants the glutamine synthetase [GS], glutamate synthase [GOGAT], glutamate dehydrogenase [GDH], and alanine dehydrogenase [ADH]; 16) have been demonstrated to preferentially use 14NH4 vs 15NH4 17 and explains somehow why the 15N depletion of rumen 1; present study), soil 14 and marine 8 bacteria increases as ammonia assimilation is improved. Wang and co-workers 16 suggested that the amidation of glutamate to produce glutamine and the subsequent conversion of glutamine plus α-ketoglutarate to glutamate is the dominant pathway for ammonia assimilation in most rumen bacteria when ammonia N is low and is governed by the coupled GS-GOGAT enzymes. An alternative pathway for ammonia assimilation in some bacteria is through glutamate synthesis performed by GDH via the amination of α-ketoglutarate when ammonia-N is high 18. However recent results indicate that both pathways for ammonia assimilation may operate concurrently within a bacterium under non-limiting ammonia-N conditions 19–21. Less is known about N isotope use in transamination reactions involved in the catabolism as well as synthesis of amino acids when bacteria take up and assimilate peptides for growth, although the 14NH2 appears to incorporate faster than the 15NH2- 10. Transamination reactions involved in the assimilation of amino acids into protein appear to be important for the Prevotella genus when organic sources of nitrogen are abundant in the rumen 20. In our study, 15N was more depleted when X. ruminicola (previously named Prevotella ruminicola 22) was grown on ammonia versus peptides which may indicate that the final isotopic signature of this bacterium in the rumen will reflect the net contribution of both of these N assimilation pathways. Interestingly, a recent study in dairy cows found a negative rather than positive correlation between the relative abundance of Prevotella taxa in the rumen and natural abundance of 15N in plasma 23, which may suggest that in those particular conditions Prevotella species could be a net ammonia user rather than ammonia producer. Indeed, Prevotella is a very diverse genus, based on genetic diversity and also phenotypic diversity of the cultured rumen isolates 24–26. In addition, the relative abundance of the various Prevotella taxa may vary according to the diet 24.
According to results by Wang et al. 16 there exists a correlation between the Prevotella ruminicola and Fibrobacter succinogenes populations and GDH and ADH activities, respectively. These two enzymes have been shown to be discriminant against 15N in the bovine liver and in Bacillus subtilis cultures, respectively 27. This may explain why in our study at similar N levels the 15N natural abundances in Prevotella ruminicola and Fibrobacter succinogenes (Exp #1) were lower with ammonia chloride vs tryptone sources. Interestingly, this pattern was not observed with the third bacterium, Eubacterium limosum, having the greater ammonia production in our conditions even when ammonia chloride was the N source. Eubacterium limosum is one of the few acetogens that can produce butyrate 28 as observed in our study from their higher butyrate molar proportion compared to the other two strains. According to Genthner et al. 29 this rumen strain is very versatile in utilization of N sources, such as peptides, ammonia and single amino acids. However, the greater 15N enrichment of E. limosum when ammonium chloride replaced tryptone (a finding not observed with the other strains) might indicate its poor ability to use inorganic N sources.
Hoch et al. 8 reviewed some studies reporting N isotope fractionation in several microorganisms using ammonia for growth, including bacteria, and found that microbial biomass was always 15N depleted compared to ammonia substrate. The enzymes GDH and GS as well as membrane NH4 transport were believed to be the main fractionation pathways involved in ammonia assimilation by marine bacteria 8,30. Fibrobacter succinogenes is one of the major fibrolytic rumen bacteria and it is known that the strain S85 of this species utilizes ammonia as its sole nitrogen source 31,32. However, Atasoglu 33 demonstrated that another strain of Fibrobacter succinogenes (strain BL2) was able to utilise non-ammonia nitrogen at high concentrations of peptides and amino acids. Furthermore, the main pathways of ammonia assimilation (GDH, ADH and GS) for Fibrobacter succinogenes S85 have been identified using nuclear magnetic resonance 34 and genome sequencing 35. These findings support the 15N depletion observed with Fibrobacter succinogenes S85 in the current study, and indicates that this rumen microorganism might preferentially use 14NH4 vs 15NH4.
The 16S rDNA amplicon sequencing analysis (Experiment #2) revealed ASVs strongly correlated to Δ15N for both nitrogen sources. Several ASVs classified to the RC9_gut_group showed a negative relationship with Δ15N for both nitrogen sources. Svartström et al., 36 indicated that RC9 possess fibrolytic activity in the rumen and identified genes encoding saccharolytic activity against mannan, xylan and pectin, however to the best of our knowledge the nitrogen metabolism of this bacterial group has not been studied. In experiment #1 another fibrolytic bacterium, Fibrobacter succinogenes, promoted lower isotopic N discrimination which indicates that fibrolytic bacteria may be involved in 15N depletion and more efficient N utilization at the rumen level. On the other hand, F082 group appeared to be associated with a less efficient nitrogen utilization at the rumen level when inorganic nitrogen was provided. Interestingly, RC9 and F082 have been found to increase in the rumen of cattle grazing in dry tropical rangelands during the dry season 37, when nitrogen sources are scarce and inorganic nitrogen is being supplemented to animals. Further research needs to be done to determine the role of these novel rumen bacterial populations in nitrogen utilization and fractionation in animals under different feeding conditions and nitrogen sources.
In conclusion, our results showed that N isotopic fractionation by rumen bacteria may change according to the balance between synthesis of microbial protein from ammonia versus non-ammonia N sources and suggest that the extent of ammonia uptake and release strongly affects the N isotopic signature of rumen bacteria. Interestingly, a stronger 15N depletion occurred when a rumen fibrolytic bacterium specialised in ammonia assimilation was grown on non-organic N sources compared to other bacteria. These results confirm the key role of rumen bacteria as modulators of N isotopic fractionation and the subsequent link with N use efficiency in ruminants. Also, results strongly suggest that this contribution may be dependent on the type of diet used and the predominant rumen bacteria populations. Further studies should determine whether natural 15N abundances in rumen bacteria or ruminant body proteins could also reflect the efficiency of N utilization for bacterial protein synthesis in the rumen.