Regulation of nitrogen isotopic ratios of cellular components

Abstract

This paper is an attempt to generalize the nitrogen isotope information of the molecules that make up the cell from our point of view. Nitrogen in the cell exists as 20 proteinaceous amino acids, nucleobases, hemes, chlorophylls, and others, and their composition is similar among organisms. Based on a physiologically simple autonomous system that maintains a balance between inputs and outputs, it is theoretically predicted that ¹⁵N is distributed to each cellular compound with a certain regularity, and thus a specific relationship in nitrogen isotopic ratios among compounds. Previous studies essentially confirm this. The nitrogen isotopic ratio of compound i constituting the cell can be generalized as δ¹⁵N_i = δ¹⁵N_plant + Δ_i (TP − 1) + γ_i, where TP represents the trophic position, and Δ_i and γ_i are the trophic discrimination factor and the intracellular ¹⁵N distribution of compound i, respectively. Knowing reliable values of Δ_i and γ_i will help us to better understand nitrogen dynamics in the biosphere and advance our understanding of the picture of the Earth’s environment through biogeochemistry.

1 Introduction

Since nitrogen is an essential element for living organisms, its uptake and metabolism are fundamental information for understanding living organisms. On the other hand, while stable nitrogen isotopic ratio, a fundamental tool in geochemistry, has long been employed to study interactions among organisms in food webs and biogeochemical cycles in natural environment (Wada 1980; DeNiro and Epstein 1981; Minagawa and Wada 1984; Altabet and Francois 1994; Sigman et al. 2009), they have not been used as much to understand the organisms themselves and the cells they comprise. However, recent advances in the study of nitrogen isotope ratios of individual amino acids make it clear that it is a useful tool for providing deeper insight into these topics. Compound-specific nitrogen isotopic ratio of various cellular molecules, including amino acids, which make up living organisms, is being measured for a variety of purposes and applied to a wide range of studies (e.g., Macko et al. 1987, Hare et al. 1991, Petzke et al. 2005, Sachs et al. 1999, Chikaraishi et al 2009, Sherwood et al. 2014, Ohkouchi and Takano 2014, McMahon and McCarthy 2015). In particular, the nitrogen isotopic ratio of individual amino acids is essential information for understanding the bulk nitrogen isotopic ratio of living organism, because nitrogen derived from amino acids constitutes most of the nitrogen that makes up cells in terms of quantity (Ohkouchi 2023).

Cells, or organisms as aggregates of cells, have mechanisms that autonomously control the amount of nitrogen and other substances that flow through them. Organisms are designed as autonomous systems that maintain a balance between inputs and outputs, which is inherently different from systems controlled by first-order kinetics often found in geologic topics. At the same time, this has advantages from an isotopic point of view. That is, one can predict the existence of certain rules in the isotopic composition of molecules that govern biological processes. The results of many studies conducted to date certainly suggest this (McClelland and Montoya 2002; Chikaraishi et al. 2007; McCarthy et al. 2007; Ohkouchi et al. 2017; Popp et al. 2007). This paper outlines the system in which this autonomous regulation operates and attempts to further develop Ohkouchi (2023) and generalize the isotopic composition of nitrogen components in the cell.

2 Enrichment of heavy isotopes of bioelements by predation

It has long been known that the enrichment of heavy isotopes of bioelements (e.g., carbon-13, nitrogen-15, and sulfur-34) occurs in the predation process (Miyake and Wada 1967; Peterson and Fry 1987; Fry 2006). Such isotopic phenomenon is generalized by the following equation:

$$ \delta_{{{\text{TP}} + {1}}} =\delta_{{{\text{TP}}}} + \varDelta$$

(1)

where δ and Δ represent the isotope ratios and their enrichment (‰) associated with the predation, and the subscript TP represents trophic position. The trophic position, or trophic level, is 1 for plants, 2 for herbivores, 3 for animals that eat herbivores, and so on. Depending on the combination of diets, decimals are possible (e.g., an omnivore that consumes equal parts of plants and herbivores would have TP = 2.5). The trophic position serves as both an indicator of the position in the food chain and a numerical representation of the energy flow order in nature. In Eq. (1), the term Δ is sometimes referred to as the ‘trophic discrimination factor’ or ‘trophic enrichment factor.’ This simple and fundamental equation for deciphering food chains is widely applied (e.g., Wada et al. 1987; Post 2002Newsome et al. 2010) and has given rise to the field of ‘stable isotope ecology’ (e.g., Fry 2006).

A cell or organism (a collection of cells functioning as a single entity with mass M and isotope ratio δ) can be conceptualized as a single box controlled by first-order dynamics and represented as in Fig. 1a. Mass balance and isotope balance in this system can be expressed as the following two simple differential equations under the condition that ¹⁴N >> ¹⁵N:

$$\frac{{{\text{d}}M}}{{{\text{d}}t}} = f_{{{\text{in}}}} - f_{{{\text{out}}}}$$

(2)

$$\frac{{{\text{d}}\left[ {M\delta } \right]}}{{{\text{d}}t}} = f_{{{\text{in}}}} \delta_{{{\text{in}}}} - f_{{{\text{out}}}} \delta_{{{\text{out}}}}$$

(3)

Fig. 1

Two approaches to understanding the discrimination of nitrogen isotopes in organisms. a Simplified cellular model as a single box regulated by first-order kinetics. b Autonomously regulated cellular model based on physiological knowledge of nitrogen metabolism (Ohkouchi 2023). ‘AA’ denotes amino acid pool in vivo. Nitrogen isotopic ratios of diet (d), temporal pool of amino acids (AA), protein reservoir (p), and excrement (ex) are denoted by δ_d, δ_AA, δ_p, and δ_ex, respectively. Flows of nitrogen denoted by f and isotope effects by ε. The isotopic ratios of nitrogen transferred the indicated pathways are given by δ_d–ε_o, δ_p–ε₁⁻, and δ_AA–ε₂

By setting the isotope ratio (δ_out) of the output of this simple model as (δ–Δ) and assuming that both mass balance and isotope balance hold true, solving these equations yields:

$$ \delta = \delta_{{{\text{in}}}} + \varDelta$$

(4)

Equation (4) represents the same concept as Eq. (1) and can explain the heavy isotope enrichment associated with predation mentioned above. In other words, a cell or an organism can be thought of as a system that discriminates isotopic ratio by Δ.

A mature cell or organism, as dynamic system, is fairly well balanced, especially in terms of mass. Otherwise, the body size of the organism would be variable, expanding and contracting. Even in an organism that is not yet fully mature, input is often only slightly greater than output. Organisms are highly self-regulating systems, and the control functions powerfully to maintain a state of dynamic equilibrium.

3 Nitrogen in the cell

In nature, cells have a great variety of shapes and sizes, generally ranging from 10 to several hundred micrometers for eukaryotic cells. Some eukaryotic cells are exceptionally large, on the order of millimeters, while prokaryotic cells are usually ~1 µm in diameter (e.g., Alberts et al. 2019). Archaea are even smaller, many of which lack some functions and often compensate for functional deficits by parasitism, symbiosis with other organisms, or consortium formation with other microbes.

From a microscopic viewpoint, eukaryotic cells are surrounded by a cell membrane and are composed of various organelles, including the nucleus, mitochondria, Golgi apparatus, endoplasmic reticulum, and cytoplasm. From a chemical perspective, however, the cell can be viewed as a container filled with a variety of compounds, including proteins and nucleic acids, among others. Focusing particularly on nitrogen, proteins and nucleic acids, which control the metabolism of living organisms, are not only functionally important but also quantitatively important. Furthermore, hemes and chlorophylls are potentially important as nitrogenous compounds in cells.

Proteins are large molecules composed of amino acid polymers and typically have molecular weights ranging from 10³ to 10⁶ Da, serving as catalysts for biochemical reactions such as enzymes. In individual cells, thousands to tens of thousands of different proteins are packed at concentrations of 2–4 × 10⁶ per μm³ cell (Milo 2013). Assuming a spherical eukaryotic cell with an average cell size of 30 μm in diameter and an average protein mass of 4 × 10⁴ Da (Fig. 2, Brocchieri and Karlin 2005), the total amount of proteins within this cell would be 1.9–3.7 ng. Assuming a cell density of 1 g cm⁻³, that corresponds to 13–27% of the cell weight. Calculated from the average composition of amino acids widely observed in organisms, there would be 0.28–0.56 ng of protein-derived nitrogen in the above cells.

Fig. 2

Comparison of the nitrogen amount derived from each intracellular component (protein, nucleic acid, heme, chlorophyll, and creatine) of eukaryotic cell, assuming a cell diameter of 30 μm. See details in text

The compositional balance of the 20 α-amino acids that make up proteins is remarkably similar across a wide range of organisms, from simple bacteria to complex animals like humans. Figure 3 illustrates the weight percent compositions of hydrolyzable amino acids from four oceanic organisms (Cowie and Hedges 1992; Wilson and Cowey 1985). Glutamic acid, aspartic acid, and lysine are the most abundant amino acids, often individually accounting for more than 10% of the total amino acids. Additionally, non-protein amino acids such as ornithine, β-alanine, γ-aminobutyric acid, and nitrogen-containing compounds like taurine, while present in cells, are generally found in lower quantities compared to the proteinaceous amino acids (e.g., Cowie and Hedges 1992).

Fig. 3

Hydrolyzable amino acid compositions as weight % of four types of oceanic organisms drawn from the data in Cowie and Hedges (1992) and Wilson and Cowey (1985). a Natural phytoplankton (n = 4); three from Dabob Bay and one from Saanich Inlet, b natural zooplankton (n = 3); two from Dabob Bay and one from Saanich Inlet, c cultured bacteria (n = 3; Vibrio alginolyticus, Bacillus subtilis, and Pseudomonas fluorescens), and d Atlantic salmon (Salmo salar; n = 2). Glx and Asx denote the sums of glutamate and glutamine, and aspartate and asparagine, respectively. n.d. denotes ‘no data’

Nucleic acids collectively refer to DNA and RNA, and chemically, they are polymers of nucleotides (nucleic acid base + sugar + phosphate). The composition of the five nucleic acid bases (adenine, guanine, cytosine, thymine, uracil), which are the building blocks of nucleotides, has known to be similar across all forms of life since the mid-twentieth century (Elson and Chargaff 1955). The total amount of nucleic acids per single cell is linearly related to cell volume (Gillooly et al. 2015). Assuming, like above, a spherical eukaryotic cell with a diameter of 30 μm, the content of nucleic acids is 50–80 pg in humans and rats, and even relatively small organisms such as planktonic crustaceans Daphnia have ~ 20 pg. Since average nitrogen weight ratio in nucleic acids is 9.5%, the nitrogen content per cell derived from nucleic acids is 0.002–0.08 ng. In a typical eukaryotic cell, the amount of nitrogen derived from nucleic acids is more than an order of magnitude less than that derived from proteins. In smaller prokaryotic cells, however, the relative amount of nucleic acids is much higher, in some cases reaching 7% by weight (Alberts et al. 2019). Various derivatives of nucleobases such as ATP and coenzyme A (CoA) are known, but they are present in very small amounts.

In addition to amino acids and nucleobases, cells contain a variety of nitrogen-containing compounds. Representative examples include chlorophyll, which is involved in energy harvesting during photosynthesis; heme, which involved in oxygen transport, storage, and electron transfer and shares the same tetrapyrrole structure; spermine, a polyamine, hexosamines such as glucosamine; and creatine, which is found in vertebrate muscles. For example, it is known that carbon derived from chlorophyll present in plant cells (algae) in the ocean averages ~ 2% of cell carbon (e.g., Cloern et al. 1995). Chlorophyll a (chemical formula: C₅₅H₇₂O₅N₄Mg) contains 6.3% nitrogen by weight, so in a cell of the above size, the average amount of nitrogen derived from chlorophyll is only ~ 0.002 ng, which is nearly equivalent to the amount of nitrogen derived from nucleic acids.

Heme is found in over 2300 different proteins (Smith et al. 2010) and serves as cofactors catalyzing various biochemical reactions. However, even heme B, the most quantitatively important heme, is two orders of magnitude less abundant than chlorophyll by weight in plant cells (Isaji et al. 2020). Bilirubin and other heme metabolites are also not worth considering in terms of quantity because they are several orders of magnitude less abundant than amino acids. As an exception, the amount of phycobilins in cyanobacteria is known to be in range comparable to that of chlorophylls (Beale 1993).

Creatine, found in vertebrate muscles, contains ~5 g of nitrogen per kg of muscle (Hultman et al. 1996). This amount is roughly equivalent to some amino acids such as proline and serine. Thus, creatine is a component that must be considered when strictly discussing the nitrogen dynamics of vertebrates.

In summary, more than 90% of the nitrogen in eukaryotic cells is derived from the α-amino acids that make up proteins, while nitrogen in prokaryotic cells is derived from α-amino acids and nucleobases (Fig. 2). From this point on, we will focus primarily on eukaryotic cells.

4 A physiological model of nitrogen isotope fractionation associated with metabolism

Considering that most of the nitrogen in cells originates from amino acids, the dynamics of nitrogen in the cell should be modeled from the perspective of amino acid metabolism. In physiology and biochemistry, the dynamics of amino acids have long been viewed not as in Fig. 1a but as in Fig. 1b (e.g., Koch 1962; Bender 2012; Ohkouchi 2023). That is, proteins taken up through predation are broken down into individual amino acids in the gut, and initially enter the amino acid pool. There, they spontaneously mix with the amino acids generated by protein breakdown, and some of them go to the protein reservoir (in the case of animals, they become part of the muscle), while the rest are excreted outside the cell.

General geochemical systems are governed by first-order kinetics, whereas as noted above, biological system is fundamentally about autonomous control of its reservoir size. The differential equations for the size and isotope ratio of the amino acid pool in the metabolic system are:

$$\frac{{{\text{d}}\left[ {{\text{AA}}} \right]}}{{{\text{d}}t}} = f_{0} - f_{1}^{ + } + f_{1}^{ - } - f_{2}$$

(5)

$$\frac{{{\text{d}}\left[ {{\text{AA}}} \right]\delta_{d} }}{{{\text{d}}t}} = f_{0} (\delta_{d} - \varepsilon_{0} ) - f_{1}^{ + } (\delta_{{{\text{AA}}}} - \varepsilon_{1}^{ + } ) + f_{1}^{ - } (\delta_{p} - \varepsilon_{1}^{ - } ) - f_{2} (\delta_{{{\text{AA}}}} - \varepsilon_{2} )$$

(6)

where f terms are fluxes and (δ–ε) terms are isotopic ratios for fluxes. In mature animals, inputs and outputs are in balance, so the following equation holds.

$$f_{0} = f_{2}$$

(7)

Since we can assume that the AA pool size is also kept constant:

$$f_{1}^{ + } = f_{1}^{ - } \left( { = f_{1} } \right)$$

(8)

such a balance is indeed widely observed in healthy adult animals and humans (e.g., Overman and Parrish 2001; Bender 2012).

As is known empirically, nitrogen isotope balance is also nearly achieved (i.e., $\frac{{{\text{d}}\left[ {{\text{AA}}} \right]\delta_{d} }}{{{\text{d}}t}} = 0$). The isotopic fractionations associated with peptide hydrolysis catalyzed by protease to individual amino acids ($\varepsilon_{0}$ and $\varepsilon_{1}^{ - }$) and protein synthesis ($\varepsilon_{1}^{ + }$) are assumed to be zero for simplicity. In contrast, as some experimental studies have shown, the amino acid metabolic step involving transamination and deamination described above accompanies the significant isotopic fractionation ($\varepsilon_{2}$). The isotope balance Eq. (6) can be arranged to:

$$\frac{{f_{1}}}{{f_{0} }} = \frac{{\delta_{d} - \delta_{\text{AA}} + \varepsilon_{2} }}{{\delta_{\text{AA}} - \delta_{p} }}$$

(9)

In the cases of investigated healthy higher animals including human, $\frac{{f_{1} }}{{f_{0} }}$ ratio for the maintenance of proteins generally falls in a range 2–3 (Bender 2012; Bergström et al. 1974; Wagenmakers 1999; Munro 1970; Houlihan et al. 1993). A significant proportion of the amino acids released by protein catabolism are reused in the synthesis of new proteins. Since in most animals the size of the protein pool is 2–3 orders of magnitude larger than that of the temporal amino acid pool, δ_p can be regarded as the bulk isotope ratio of the whole organism (δ_bulk). When we apply Eq. (4) to Eq. (9), the magnitude of ¹⁵N enrichment due to predation in organisms can be expressed by the following equation:

$$ \varDelta_{{{\text{bulk}}}} = \varepsilon_{2} + \left( {1 + \frac{{f_{1} }}{{f_{0} }}} \right)\left( {\delta_{{{\text{bulk}}}} - \delta_{{{\text{AA}}}} } \right)$$

(10)

On the right-hand side of Eq. (10), the first term ε₂ can be considered a constant, while the second term is a variable representing the organism’s response to the environment. In the second term, $\left( {1 + \frac{{f_{1} }}{{f_{0} }}} \right)$ is a positive value, but (δ_bulk–δ_AA) becomes negative, resulting in the second term on the right side as a whole is negative. In other words, Δ_bulk value (i.e., 3.4‰ on average) is smaller than ε₂ and dynamically determined (Fig. 4). However, since the cell is an autonomously regulated system, the Δ_bulk value falls within a relatively narrow range and has been empirically considered as a constant in ecological applications.

Fig. 4

Schematic illustration of Eq. (10) regarding Δ_bulk shown in the upper panel. In the lower panel, observation results are also shown as a frequency distribution of Δ_bulk values (0.5‰ increment; redrawn from Post 2002). Broken line indicates mean Δ_bulk value (3.4‰)

For example, one would expect that starvation (f₀ → 0) would increase the value of the second term on the right-hand side (1 + f₁/f₀) and also increase the isotopic ratio of amino acid pool (δ_AA), which strongly reflect fractionation during amino acid degradation (ε₂), leading to a decrease in the value of Δ_bulk. However, as critically evaluated by Doi et al. (2017), observations do not necessarily replicate this prediction, and the nitrogen isotopic ratio does not change significantly even after experiencing starvation. This may be explained by a metabolic switch in response to changes in the external environment, such as starvation, that reduces f₁ and f₂ (the latter in particular has the effect of reducing δ_AA) and kept the Δ_bulk value constant.

5 Mechanism of ¹⁵N enrichment

As described earlier, since most of the nitrogen in the cell is derived from proteinogenic amino acids, the nitrogen isotopic ratio of the bulk organism should approximately correspond to the weighted average of the isotopic ratios of its 20 amino acids. Focusing on the dynamics of nitrogen (amino group) of amino acids, the main metabolic processes are aminotransfer reaction and degradation by dehydrogenase reaction (oxidative deamination) (Fig. 5).

Fig. 5

Overviews of chemical mechanisms of cellular reactions related to amino group of amino acids. a A half-reaction of transamination (deamination) of amino acids catalyzed by aminotransferase. PLP indicates pyridoxal 5′-phosphate, common cofactor of aminotransferases, and PMP pyridoxamine 5′-phosphate. b Oxidative deamination of glutamic acid catalyzed by glutamate dehydrogenase (GDH). In the cell, this reaction proceeds primarily from left to right. E is enzyme and broken lines indicate nucleophilic attacks

The aminotransfer reactions are responsible for the redistribution of ¹⁵N between each amino acid in the cell. They are reactions in which amino groups of various amino acids are transferred to 2-oxoacids (α-ketoacids) to form different α-amino acids. Pyridoxal 5'-phosphate (PLP) functions as a cofactor in the active center of various aminotransferases that catalyze amino group transfers (Eliot and Kirsch 2004; Ohkouchi et al. 2015, Fig. 5a). In other words, once the amino acid binds to PLP to form a complex, it dissociates the 2-oxoacid after several intermediates from the complex. The amino group, via a temporary binding to PLP, is soon transferred to another 2-oxoacid to form a new amino acid. Most of the amino groups are eventually transferred to 2-oxoglutaric acid to form glutamic acid.

In such amino group transfer reactions, the following two points are important when considering the nitrogen isotopic ratio. First, it is the amino acid with a deprotonated amino group (R-NH₂) in which the amino group has a lone electron pair, that forms a complex with PLP. When amino acids with deprotonated amino groups are in isotopic equilibrium with their protonated (R-NH₃⁺) counterparts in the cell, their isotopic fractionation exceeds 16‰ (Hermes et al. 1985; Rishavy and Cleland 2000; Oi et al. 2023). Given the neutral to somewhat alkaline pH of the cells, the majority of amino acid amino groups would be in a protonated state, while the δ¹⁵N values of the minor deprotonated amino acids would be much lower than the protonated ones.

Another isotope fractionation occurs at the Schiff base formation between the deprotonated amino acids and PLP (Rishavy and Cleland 2000). The association and dissociation of the deprotonated amino acids with PLP is reversible, with preferential dissociation of R-¹⁴NH₂ from the PLP. Thus, when the subsequent reaction step is rate-limiting, R-¹⁵NH₂ complexed with PLP would preferentially proceed to the final product. Histidine and glutamate carboxylases, whose reactions are rate-limited at decarboxylation step that occur immediately after the Schiff base formation, have shown ¹⁵N enrichments in the products up to 23‰ (Abell and O'Leary 1988). In case of aminotransferases, this ¹⁵N enrichment partially offsets the equilibrium ¹⁵N-depletion in the substrate R-NH₂, leading to formation of other amino acids depleted in ¹⁵N by up to ~ 8‰ (Macko et al. 1987; Rishavy and Cleland 2000; Goto et al. 2018).

In contrast, in oxidative deamination, the amino groups of each of the 20 amino acids are first assembled into glutamate by an aminotransfer reaction. The glutamate is then decomposed into 2-oxoglutarate and ammonium by glutamate dehydrogenase (GDH) in the mitochondria (Fig. 5b). This reaction is reversible, but because of the low affinity of ammonium for GDH, glutamate is usually deaminated to yield ammonium. The ammonium is converted to urea through the urea cycle and excreted from the body. Thus, the metabolic enrichment of ¹⁵N is mainly due to the oxidative deamination of glutamate, which produces ¹⁵N-depleted ammonium that is then removed from the body. If the oxidative deamination by GDH is in equilibrium, ¹⁵N is enriched by 47‰ on the glutamate (Schimerlik et al. 1975; Cleland 1982). This ¹⁵N enrichment is partially offset by the effect of the aminotransfer reaction described earlier, which produces ¹⁵N-poor glutamate.

6 Generalization of nitrogen isotopic ratios of intracellular compounds

The composition of the 20 amino acids that make up proteins is relatively similar, even among very different organisms such as microbes and higher organisms, or organisms at different trophic positions (Fig. 3, Ohkouchi 2023). This implies that the f₁/f₀ ratios of individual amino acids are comparable among different organisms, as described above, suggesting that the discussions mentioned earlier at the bulk level are essentially true at the individual amino acid level. Since most of the nitrogen contained in eukaryotic cells originates from the amino acids that make up proteins, we go one step further to discuss the nitrogen isotope ratios of individual amino acids.

Here, let us introduce the concept of ‘isotope distribution within the cell (γ)’ as a perspective to consider the isotope ratios of various compounds synthesized within the cell. The γ value is defined as the deviation of compound i from the bulk isotope ratio of the cell (γ_i≡ δ¹⁵N_i − δ¹⁵N_bulk). With this concept, we can generalize the nitrogen isotopic ratio of compound i, which constitutes the cell (i.e., naturally occurring compounds), using the following equation:

$$\delta^{15} N_{i} = \delta^{15} N_{{{\text{plant}}}} + \varDelta_{i} (TP - 1) + \gamma_{i}$$

(11)

On the right-hand side of Eq. (11), the first term is the ecosystem-base term, the second term can be called the trophic term, and the final term, γ_i, represents the intracellular isotope distribution to compound i, as described above. Table 1 and Fig. 6 summarize the current version of Δ_i and γ_i values for aquatic organisms, based primarily on our previous studies. These figures are still controversial for some compounds and should be updated in the future (e.g., McMahon and McCarthy 2016). As with the Δ_bulk value discussed above, at least the Δ_i values are inherently variable, albeit within a relatively narrow range (with some exceptions). And if metabolic flux changes, for example due to dietary imbalances over time, this should be accompanied by a change in the Δ_i value (Chikaraishi et al. 2015). With this in mind, the values shown in Table 1 need to be validated and discussed.

Table 1 Current summary of intracellular ¹⁵N distribution (γ, ‰) in the cell and trophic discrimination factor (Δ, ‰) for aquatic organisms based primarily on our previous studies. These values can be applied to Eq. (11) in the text

Fig. 6

A cross-plot of intracellular ¹⁵N distribution (γ) and trophic discrimination factor (Δ) for compounds presented in Table 1. Open symbols indicate amino acids for which the error in the γ value is unknown (i.e., asparatate, threonine, and lysine). Error bars indicate 1σ

Studies of nearly 20 nitrogenous compounds from various aquatic organisms have revealed a wide range of γ_i values from − 7 to + 11‰ for each compound constituting the cell (Table 1). In particular, chlorophyll a from freshwater cyanobacteria has the highest γ_i values, indicating the existence of an intracellular ¹⁵N enrichment mechanism (Katase and Wada 1990; Sachs et al. 1999; Higgins et al. 2011, Ohkouchi et al. 2006). In aquatic organisms, there is little difference in the γ_i value of each amino acid between organisms from different evolutionary lineages, such as algae and cyanobacteria (Chikaraishi et al. 2009), but there are clear differences in chlorophylls. This suggests that somewhere in the process of chlorophyll synthesis from precursor amino acids (i.e., glutamic acid and glycine) is the source of the difference in isotopic ratios, and that the process(es) probably emerged during evolution.

On the other hand, the Δ_i values of amino acids show a distribution of more than 15‰ from − 8 to + 8‰. Some amino acids, such as phenylalanine and methionine, show little change in nitrogen isotopic ratio with predation processes (i.e., metabolism). These amino acids are called ‘source amino acids’ and are considered to have a Δ_i value close to zero because protonation–deprotonation and aminotransfer reactions are not directly involved in their metabolism. In fact, in metabolism, phenylalanine is hydroxylated to form tyrosine, and methionine is catalyzed by ATP to form S-adenosylmethionine. The metabolisms of threonine with a negative Δ_i value and lysine with a relatively small Δ_i value are not also involved in aminotransfer reactions.

7 Estimation of trophic position

Equation (11) demonstrates that if the Δ_i and γ_i values for each compound are known in advance, the nitrogen isotope ratio of primary metabolic products synthesized by organisms will become a function of the δ¹⁵N value of the ecosystem pool to which the sample organism belongs (the ecosystem-base term) and its trophic position. Now considering compounds ‘a’ and ‘b’ within the same organism and taking the difference of Eq. (11) for them, we obtain,

$${\delta} ^{15} {\text{N}}_{a} - {\delta} ^{15} {\text{N}}_{b} = ({\varDelta}_{a} - {\varDelta}_{b} )({\text{TP}} - 1) + ({\gamma}_{a} - {\gamma}_{b} )$$

(12)

and the ecosystem-base term is eliminated. Solving this equation for TP:

$${\text{TP}} = ({\delta} ^{15} {{\text{N}}_{a} - \delta} ^{15} {\text{N}}_{b} - (\gamma_{a} - \gamma_{b} ))/({\varDelta_{a}} - {\varDelta}_{b} ) + 1$$

(13)

This yields a general formula to determine the trophic position.

Substituting two amino acids with relatively small error γ_i and Δ_i values into Eq. (13) yields a TP with relatively small error. Typical examples of such amino acids are glutamic acid (Glu) and phenylalanine (Phe), which have very different Δ_i values (Table 1). Equation (14) can be derived for aquatic organisms (Chikaraishi et al. 2009):

$${\text{TP}} = ({\delta} ^{15} {\text{N}}_{{{\text{Glu}}}} -{\delta} ^{15} {\text{N}}_{{{\text{Phe}}}} - 3.4)/7.6 + 1$$

(14)

In other words, by measuring the nitrogen isotope ratios of the two compounds, glutamic acid and phenylalanine, in a sample organism, it is possible to estimate the trophic position of that organism. To accurately estimate TP, (γ_a–γ_b) and (Δ_a–Δ_b) must not change much. In fact, evidence has been presented and discussed for various aquatic plants and prey–predator combinations as to how much these values vary (e.g., Chikaraishi et al. 2009; McCarthy et al. 2013; McMahon and McCarthy 2016; Decima et al. 2017; Whiteman et al. 2019). If TP is calculated using the combination of glutamate and phenylalanine, the estimation error is generally ~ 0.1.

While the variability of amino acid nitrogen isotope ratios has been discussed, the TP obtained by this method has provided us with a variety of information, such as the feeding habits of organisms that are difficult to observe (e.g., deep-sea organisms) and basic information on fishes useful in fisheries (Miller et al. 2012, Ohkouchi et al. 2013, Tsuchiya et al. submitted). Furthermore, the TP information has also been applied to diverse studies, including the efficiency of energy transfer through trophic pathways, the effect of eutrophication on fish diet, and paleoanthropogenic diet analysis (e.g., Ishikawa et al. 2022; Ogawa et al. 2013; Naito et al. 2016).

8 Estimation of ecosystem-base δ¹⁵N

If TP can be estimated, the other unknown in Eq. (11), δ¹⁵N_plant, or ecosystem-base δ¹⁵N can be estimated by the following equation:

$${\delta} ^{15} {\text{N}}_{{{\text{plant}}}} ={\delta} ^{15} {\text{N}}_{i} - {\varDelta}_{i} ({\text{TP}} - 1) - \gamma_{i}$$

(15)

For example, if applying phenylalanine to Eq. (15):

$${\delta} ^{15} {\text{N}}_{{{\text{plant}}}} ={\delta} ^{15} {\text{N}}_{{{\text{Phe}}}} {-}{\varDelta}_{{{\text{Phe}}}} ({\text{TP}} - 1){-}\ \gamma_{{{\text{Phe}}}}$$

(16)

Considering the very small Δ_i value of phenylalanine (Table 1), a source amino acid, it can be simplified as follows.

$${ \delta} ^{15} {\text{N}}_{{{\text{plant}}}} {\simeq\delta} ^{15} {\text{N}}_{{{\text{Phe}}}} {-}\ \gamma_{{{\text{Phe}}}}$$

(17)

Furthermore, the equation becomes more robust if applying chlorophyll a, which is synthesized only by plants (TP = 1):

$${ \delta} ^{15} {\text{N}}_{{{\text{plant}}}} ={ \delta} ^{15} {\text{N}}_{{{\text{chl}}}} {-}\ \gamma_{{{\text{chl}}}}$$

(18)

If the sample is a mixture or contains a significant amount of microorganisms, Eq. (18) may be more accurate than Eq. (17). In particular, since chlorophyll nuclei can persist as porphyrins in sediments for billions of years, ecosystem-base δ¹⁵N information is useful for reconstructing the nitrogen cycle in the surface ocean over geologic time.

The greatest advantage of applying amino acid nitrogen isotope ratios to natural biological samples is the ability to separate the two inseparable pieces of information in the bulk nitrogen isotope ratios, TP and ecosystem-base δ¹⁵N. The importance of separating these has often been overlooked. One of the important applications of ecosystem-base δ¹⁵N is its use in isoscape, which has developed remarkably in recent years (Bowen 2010). A notable example is Matsubayashi et al. (2020), who revealed the migration path of chum salmon (Oncorhynchus keta) in the North Pacific based on the nitrogen isotope ratio of phenylalanine contained in the growth ring of vertebral centra. In addition to vertebral centra, otoliths and eyeballs have been known to accumulate substances by living organisms, and the application of nitrogen isotope ratios of eyeballs, which are enriched in amino acids, to isoscape will be developed in the future (McMahon and Newsome 2019; Harada et al. 2022; Yoshikawa et al. 2024).

9 Conclusions and implications

It is often difficult to describe quantitatively the macroscopic phenomena of the Earth’s environment, where biological processes dominate. Part of this difficulty is due to the ‘biological opportunism’ with which organisms change the way they cope with the external environment (in order not to change themselves). This paper proposed a perspective that incorporates autonomous biological system a priori to solve this problem, and described the intracellular distribution of nitrogen isotopes of mainly amino acids. The nitrogen isotopic balance among the various nitrogeneous compounds in the biosphere should be essentially explained by γ_i and Δ_i values, the isotopic discrimination associated with biological nitrogen uptake and assimilation. With more robust information on γ_i and Δ_i, we will have a better understanding of the nitrogen dynamics in the biosphere and further progress in understanding the whole picture of the Earth’s environment through biogeochemistry.

The same should, of course, be true for other isotopes of bioelements such as carbon, as described earlier. Decarboxylation of amino acids, like deamination, is a PLP-dependent reaction that involves fractionation of carbon isotopic ratio (McCarthy et al. 2004; Takizawa et al. 2020; Sun et al. 2020). However, since C–C bond is stronger than C–N bond and it form the major backbone of all biological compounds, carbon isotopic ratios should not vary as much as nitrogen isotope ratios, which has been often confirmed in nature (e.g., Wada 2009; Fry 2006). Furthermore, the discussion developed in this paper is not limited to the Earth. If an autonomous system (maybe life?) exists somewhere off Earth, we would find a specific relationship between the isotopic ratios of each component. The regularity of isotope ratios predicts the system that controls them.