A 13C pulse chase experiment and isotopic analysis on biochemical fractions of fungal sporocarps was used to investigate carbon dynamics in two ectomycorrhizal species. Fungal structural compounds and bulk sporocarps reached similar labeling levels as phloem sap (Fig. 1b), suggesting plant C as the primary source for these fungi. In contrast, the maximum 13C labeling of the amino acid pool reached only ~ 60% of that of the structural carbohydrate pool (Fig. 3, Table 2) and the δ13C signature of the amino acid pool declined more slowly than the structural pool after the 13C pulse. A month after labeling, the amino acid pool had retained more 13C label than the structural carbohydrate pool (Fig. 2) despite its lower maximum labeling level. From these data, it appears that roughly 60% of amino acid carbon was derived from recent plant photosynthesis and 40% from carbon of longer residence time (either plant-derived or soil-derived).
A partitioning of amino acid carbon into distinct pools that differ in their turnover rate can be plausibly explained by examining the biosynthetic pathways of different amino acids. Amino acids that are derived from glycolytic or citric acid cycle intermediates in only a few steps can be frequently resynthesized and have a higher turnover rate compared to amino acids with carbon skeletons not directly derived from glycolytic or citric acid cycle intermediates (Hobbie and Hobbie 2012). For example, amino acids such as glutamic acid (synthesized in one enzymatic step from the citric acid cycle intermediate 2-oxoglutarate), aspartic acid (synthesized in one enzymatic step from the citric acid cycle intermediate oxaloacetate), and glycine (synthesized in four steps from the glycolytic intermediate 3-phosphoglycerate), have more rapid turnover rates in microbes than those with more complex biosynthethic pathways (Hobbie and Hobbie 2012), such as lysine (eight enzymatic steps from the citric acid cycle intermediate 2-oxoglutarate), leucine (four enzymatic steps after valine synthesis), and isoleucine (five enzymatic steps after threonine synthesis) (Barton et al. 2010). Assuming that C in ‘fast’ amino acid pool has the same origin of recent plant photosynthates as the structural carbon pool, the regression analysis indicates that the 13C of this ‘slow’ pool of amino acids peaked later (30 days post-labeling at 18‰) and then gradually declined to a δ13C value of -12‰ by day 93 (Fig. 4). Estimated peak labeling levels of this ‘slow’ amino acid pool are only ~ 18% of the 13C labeling in structural carbon, with an estimated 41‰ enrichment relative to control values, compared to maximum 13C enrichment of structural carbon of the ectomycorrhizal fungi of 222‰ higher than controls.
To further assess the plausibility of two pools of amino acids which differ in turnover time, information on the typical amino acid composition of ectomycorrhizal mushrooms was used to examine the possible identity of the amino acids in these two different pools. In Table 3, information on the complexity of the biosynthesis of different amino acids is listed (Barton et al. 2010), together with relative contributions of different amino acids to sporocarps of four commercially cultivated fungal species (Mattila et al. 2002). The nine amino acids with four or less enzymatic steps to final synthesis (average, 2.0 steps) comprised 47% of sporocarp amino acid carbon, whereas the eleven amino acids with six or more enzymatic steps (average, 9.0 steps) comprised 53% of amino acid carbon in sporocarps. Limited data on catabolic metabolism of these amino acids in microbes agreed with less turnover (and greater retention) of more biosynthetically complex amino acids (Hobbie and Hobbie 2012). Turnover (assessed as conversion to CO2) averaged 55% for glutamic acid and aspartic acid (one enzymatic step to synthesis), averaged 30% for glycine (four enzymatic steps), and averaged 18% for leucine, lysine, and isoleucine (respectively, seven, ten, and eleven enzymatic steps to final synthesis).
The results from this labeling experiment cannot reveak if some of the ‘slow’ amino acids are derived from current-year photosynthates or from uptake from soil organic nitrogen. However, in prior work with radiocarbon, both Cortinarius and Lactarius amino acids were partially derived from soil organic N uptake. Cortinarius amino acids from northern Finland were significantly older than Lactarius amino acids (Hobbie et al. 2013). Given the increase in age and δ13C signatures from surface litter horizons to deeper horizons, this plausibly accounts for the 0.9‰ higher δ13C signatures of Cortinarius amino acids than Lactarius amino acids (Table 1). For example, δ13C signatures of bulk soil increased by about 2‰ from the S to the H horizon at two nearby research sites of Norrliden and Rosinedalsheden (Hobbie & Hasselquist, 2018).
A hypothesis of how amino acids and carbohydrates cycle in this ectomycorrhizal system is summarized in Fig. 5. After assimilation of added 13CO2 by Pinus sylvestris into plant sugars, these are transported belowground within a few days and converted into fungal sugars and amino acids. These sugars and amino acids are used to construct the developing sporocarp. Easily synthesized amino acids (such as aspartic acid or glutamic acid) turn over quickly (‘fast’ amino acids) and will have essentially the same 13C labeling level as fungal sugars used to construct the structural carbohydrates that make up most of sporocarp biomass. Amino acids with longer biosynthetic pathways also turn over more slowly (‘slow’ amino acids) and will lag the 13C labeling of fungal carbohydrates and the ‘fast’ amino acid pool. Because of the slow turnover of some amino acids, they are more likely than the ‘fast’ amino acids to be preserved after mobilization from soil organic nitrogen or after biosynthesis from plant-derived sugars. These ‘slow’ amino acids in sporocarps accounted for the lower δ13C signatures of fungal amino acids than fungal structural carbon during the first few weeks after the pulse-chase addition of 13CO2, and presumably accounted for the higher δ13C of fungal amino acids four weeks and 70–100 days after 13CO2 addition.
The 13C tracer data cannot be used by itself in this study to indicate amino acid uptake by ectomycorrhizal fungi. However, the natural abundance δ13C data on amino acids, with Cortinarius higher in δ13C than Lactarius (Table 1), does suggest that soil-derived amino acids differing in δ13C are taken up by these two taxa. If ‘fast’ and ‘slow’ amino acids are taken up from the soil in similar proportions to their abundance in soil protein but turnover rates in fungal biomass differ as suggested, then mass balance considerations would dictate that a greater proportion of ‘slow’ amino acids than ‘fast’ amino acids must be derived from soil amino acids (Fig. 5).
Studies of amino acid uptake by plants in soil have primarily focused on metabolically simple amino acids, such as glycine, aspartic acid, alanine, and glutamic acid (Chapin et al. 1993; Gallet-Budynek et al. 2009; Kielland 1994; Näsholm et al. 1998; Näsholm et al. 2000). One study investigated the uptake and catabolism of 13C-labeled glycine, arginine, and a mixture of extracted peptides from cyanobacteria (Persson et al. 2003). In that study, respiratory losses of 13CO2 were intermediate for the peptide mix, highest in the glycine treatment, and lowest in the arginine treatment. This corresponded to the relative rates of predicted respiration rates based on the division of amino acids into ‘fast’ cycling (e.g., glycine), ‘slow` cycling (e.g., arginine) amino acid pools, and a combination of these two amino acid types in peptides. By comparing the dynamics of the structural carbon pool and the amino acid pool here, the presence of ‘fast’ and ‘slow’ cycling pools of amino acids were inferred; these pools conformed to literature information on the turnover rates and biosynthetic pathways of amino acids (Barton et al., 2010; Hobbie and Hobbie, 2012). The higher energetic cost of synthesizing these complex amino acids should make both soil uptake and retention of these amino acids a favored strategy for many fungi. The proposed dichotomy in amino acid behavior suggests that additional effort should be devoted to comparing the turnover and uptake patterns of amino acids with their biosynthetic and catabolic pathways.