Differential C Isotope Discrimination by Fungi during Decomposition of C₃- and C₄-Derived Sucrose

Abstract

Stable isotope analysis is a major tool used in ecosystem studies to establish pathways and rates of C exchange between various ecosystem components. Little is known about isotopic effects of many such components, especially microbes. Here we report on the discovery of an unexpected pattern of C isotopic discrimination by basidiomycete fungi with far-reaching consequences for our understanding of isotopic processing in ecosystems where these microbes mediate material transfers across trophic levels. We measured fractionation effects on three ecologically relevant basidiomycete species under controlled laboratory conditions. Sucrose derived from C₃ and C₄ plants is fractionated differentially by these microbes in a taxon-specific manner. The differentiation between mycorrhizal and saprotrophic fungi observed in the field by others is not explained by intrinsic discrimination patterns. Fractionation occurs during sugar uptake and is sensitive to the nonrandom distribution of stable isotopes in the sucrose molecule. The balance between respiratory physiology and fermentative physiology modulates the degree of fractionation. These discoveries disprove the assumption that fungal C processing does not significantly alter the distribution of stable C isotopes and provide the basis for a reevaluation of ecosystem models based on isotopic evidence that involve C transfer across microbial interfaces. We provide a mechanism to account for the observed differential discrimination effects.

In the last decade, the analysis of stable C isotopes has emerged as a major tool to trace and quantify C transfers across trophic levels in a variety of ecosystems (10, 21, 28). Two major premises concerning stable C isotopes in the environment are frequently assumed. First, it is known that the mechanism of CO₂ uptake from the atmosphere and incorporation of CO₂ into plant organic matter through photosynthesis result in characteristic isotopic ratios, which distinguish C₃ plants from C₄ plants (37). Second, it is frequently assumed that the effects of other biological transformations on the natural distribution of stable C isotopes are relatively insignificant compared to the photosynthesis-determined isotopic discrimination (17, 18, 41, 42). These two assumptions are at the core of isotope-based models of ecosystem function and global nutrient cycling.

While much refinement in knowledge has occurred with respect to photosynthetic pathways under various ecological conditions (11, 37), other equally important processes, such as decomposition, have remained mostly unexplored with respect to their isotopic discrimination effects, although they are critical for understanding ecosystem C flows. Previous studies have shown that microbial metabolic processes can be associated with characteristic fractionation patterns at the subcellular scale (1, 7), but the discrimination effects are often masked at the whole-organism level, resulting in the assumption that, overall, C isotopic discrimination due to microbial processing is not significant (<1.0‰). Circumstantial evidence supporting this assumption is derived from the observation that, in general, the isotopic ratios in ecosystem-respired CO₂ roughly match those of the dominant vegetation (12, 13). Similarly, when ecosystems are transformed from C₃-dominant vegetation to C₄-dominant vegetation, the relative abundance of stable C isotopes in the soil is roughly maintained and is similar to that in the original vegetation, suggesting that only minor changes due to microbial transformations occur (5, 43). Nevertheless, more detailed measurements of C isotopic distributions have revealed patterns that suggest that significant discrimination by microbes occurs during soil formation; these patterns include the consistent enrichment of ¹³C often observed with increasing depth in soil profiles (34) and the relative enrichment observed in the CO₂ produced from soil respiration compared with canopy measurements (12, 13).

More directly, recent studies of fungi indicate that significant isotopic effects can be apparent when fungal tissues are compared to their presumed plant substrates (20, 22, 24, 25, 27, 44, 45), and a consistent difference in isotopic fractionation between mycorrhizal and saprotrophic basidiomycetes has also been found in the field (22, 25, 27). Given the importance of fungi in terrestrial ecosystems (15, 35, 40), the implications of isotopic discrimination associated with fungal C processing are of great consequence for isotope-based nutrient cycling models. The question remains as to whether fractionation patterns observed in the field result from intrinsic fungal processing or are due to substrate effects (20, 44, 45).

In this work we documented intrinsically determined isotopic fractionation in fungi and studied the basis for a newly discovered isotopic discrimination mechanism that is sensitive to differences in C₃- and C₄-derived sugars during their catabolism. We concentrated on three basidiomycete species chosen for their contrasting fractionation patterns and ecophysiological roles in pine-dominated ecosystems.

MATERIALS AND METHODS

Cultures.

Live mycelia were isolated from fresh fungal sporocarps of Cryptoporus volvatus (a wood decayer), Marasmius androsaceus (a litter decomposer), and Suillus granulatus (an ectomycorrhizal organism) on solid modified Melin-Norkrans medium (MMN) (30). These three fungal species were chosen from a larger collection of higher fungi associated with Monterey pines in California because of their relative importance in the ecosystems as well as their contrasting ecological roles. Fewer than three subculturings were performed between the initial isolation from sporocarp tissues and the experiments. Liquid MMN was further modified to make sucrose the dominant C source (1.17% [wt/vol] sucrose, 0.13% [wt/vol] malt). A malt concentration of 0.13% (wt/vol) was experimentally determined to be the minimum concentration necessary for growth of the three fungal species (data not shown). Four sources of sucrose were used: pure beet sucrose (98% pure as assayed by the Beet Sugar Foundation), impure maple sucrose (Andronico's Market), pure cane sucrose (C&H Sugar Company), and impure cane sucrose (Cumberland Packing Corporation). N was supplied as ammonium phosphate, and the medium was autoclaved for 15 min at 120°C. The initial medium pH was 5.9 to 6.4. The growth vessels (open system) were 125-ml Erlenmeyer flasks that contained 75 ml of MMN and were capped with foam and aluminum foil. Two small mycelial plugs were obtained from the edge of fast-growing colonies on solid medium and used to inoculate liquid MMN vessels. The inoculum plugs were 1 mm in diameter and contained less than 0.5 μg of C; thus, carryover of C from stock cultures was minimized. Cultures were incubated at 25°C on a shaker at 175 rpm and were grown on the experimental C source until the inoculated fungal biomass grew by 4 to 5 orders of magnitude. Blanks contained no fungus but were otherwise treated like the inoculated vessels. The final pH of the filtered medium was 6.25 ± 0.02 (mean ± standard deviation) for blanks and 2.90 to 3.60 for fungally processed medium. The low pH of the medium after growth prevented reabsorption of respired CO₂ into the medium as carbonate (pH 10.33). At harvesting, samples were plated on solid MMN to check for contamination and to check the identify of the fungus. Culture identity was confirmed by morphology and also by matching rDNA internal transcribed spacer (ITS)-restriction fragment length polymorphism patterns of cultures and the original sporocarps used for isolation (19). PCR amplification of DNA was performed with primers ITS1F and ITS4, and restriction fragment length polymorphism patterns were obtained by restriction of the amplicons with AluI and HinfIII. The numbers of replicate flasks used for each treatment were 14, 16, 18, and 15 for the blank, C. volvatus, M. androsaceus, and S. granulatus treatments on C₃ sucrose, respectively, and 8, 3, 8, and 8 for the blank, C. volvatus, M. androsaceus, and S. granulatus treatments on C₄ sucrose, respectively. Except for the experiments in which three replicates were used, all experiments were performed on at least two separate occasions and the results of the runs were pooled.

Mycelia were harvested by filtration through Fisher G8 fiberglass filters. Fungal biomass was placed in 2.5-ml Nalgene cryovials. For each replicate, biomass and a 12-ml aliquot of the filtered medium were frozen at −80°C and then lyophilized. Two clean stainless steel ball bearings were added to cryovials to pulverize the biomass to a fine powder with a Wiggle-L-Bug (Crescent Dental) for 1 min before samples were weighed for isotopic analysis.

CO₂ collection.

The CO₂ collection vessels used (closed system) were 250-ml Erlenmeyer flasks containing 75 ml of liquid MMN with pure beet sucrose (pH 6.31) or pure cane sucrose (pH 6.45) as the C source. The vessels were each capped with a rubber stopper perforated to hold two 5-mm-diameter glass tubes plugged by rubber septa. One of the tubes extended into the medium. After autoclaving, the flasks were inoculated with either C. volvatus or M. androsaceus. We chose these species because of their relatively fast growth rates and contrasting isotopic discrimination characteristics. The flasks were purged of CO₂ by removing the rubber septa and pumping air through two Ascarite II (Thomas Scientific)-packed 10-ml VOST traps (Supelco), a fiberglass prefilter, and a 0.2-μm-pore-size filter (Millipore). Sterile, CO₂-free air was bubbled through the medium at a rate of 200 ml/s for 1 min. The septa were repositioned, and the vessels were incubated as described above for 18 days. On day 17, a 5-ml sample was extracted from the headspace with an air-tight syringe (Air-Tite) and immediately analyzed for CO₂ with a mass spectrometer. One day after CO₂ collection, cultures were harvested and used for mass spectrometry of fungal biomass and media as described above. The CO₂ concentrations in blanks after 17 days were 0.17% ± 0.06% (mean ± standard deviation). Each system was replicated three times.

Mass balance.

Mass balance calculations were obtained directly for the closed-system cultures and were estimated for the open-system cultures. Between 89.5 and 102.6% of the original C provided in the growth medium was recovered in the closed-system cultures (Table 1). Variations in C recovery were attributable to losses during mycelial filtration, because medium entrapped by mycelia was not forcefully recovered to avoid contamination of media with intracellular fungal products. C. volvatus and M. androsaceus had utilized between 6.8 and 10.0% of the original C supplied for growth by the time of harvest (Table 1). No significant growth rate differences (Mann-Whitney U test; P > 0.20) between C. volvatus and M. androsaceus were found when the organisms were grown on either of the sugar types provided as a C source (data not shown). The amounts of C utilized by fungal species during growth on C₃-derived sucrose and growth on C₄-derived sucrose did not vary significantly in the closed- or open-system flasks (Tables 1 and 2). C utilization could only be estimated in the open-system cultures since CO₂ could not be collected quantitatively (Table 2). For this estimation, the weight of the fungal biomass was determined at the time of harvest and the amount of CO₂ respired was calculated from the average CO₂/fungal biomass ratio obtained from the closed-culture measurements (Table 1). Less than 12% of the C originally supplied was utilized by C. volvatus or S. granulatus in the open-system cultures according to these estimates; for M. androsaceus the estimated values were 25 and 24% (Table 2). We believe that latter finding was an anomaly perhaps related to specific efficiencies in C use by this fast-growing fungus. Otherwise, the small proportion of the total C utilized over the experimental period ensured that measurements were obtained when the substrate was not limiting, reducing potential artifacts related to differential growth rates and saturation dynamics. Under these conditions, isotopic discrimination is expected to be at a maximum (29). Furthermore, preliminary experiments with S. granulatus (a slow grower) and M. androsaceus (a fast grower) indicated that the isotopic values of whole-cell preparations were constant for each species, i.e., independent of incubation time between 25 and 58 days. For these reasons, the isotopic values reported in this paper are not corrected for the amount of C utilized.

TABLE 1.

C mass balances for closed-system incubations for two fungal species on C₃- and C₄-derived sucrose

Sugar	Species	C mass balance (g)				% of original C recovered	% of original C used	Expected medium δ13C	Observed medium δ13C
		Blank medium	Medium	Fungusa	Respired CO2
C3-derived sucrose	C. volvatus	0.440 ± 0.017 (16)b	0.421 ± 0.160 (3)	0.018 ± 0.002 (3)	0.012 ± 0.002 (3)	102.6 ± 3.3	6.8 ± 0.6	−22.011 ± 0.044	−22.829 ± 0.096
	M. androsaceus	0.440 ± 0.017 (16)	0.383 ± 0.007 (3)	0.021 ± 0.003 (3)	0.018 ± 0.001 (3)	96.3 ± 2.7	9.1 ± 0.8	−22.712 ± 0.029	−22.539 ± 0.060
C4-derived sucrose	C. volvatus	0.438 ± 0.015 (10)	0.383 ± 0.019 (3)	0.023 ± 0.003 (3)	0.014 ± 0.000 (3)	96.0 ± 3.6	8.6 ± 0.7	−11.132 ± 0.030	−11.045 ± 0.013
	M. androsaceus	0.438 ± 0.015 (10)	0.348 ± 0.020 (3)	0.025 ± 0.005 (3)	0.018 ± 0.004 (3)	89.5 ± 5.5	10.0 ± 0.7	−11.135 ± 0.002	−11.007 ± 0.025

^aThe original fungal inoculum contained less than 0.5 × 10⁻⁶ g of C. 

^bThe numbers in parentheses are numbers of replicates. 

TABLE 2.

Estimated C mass balances for open-system incubations for three fungal species on C₃- and C₄-derived sucrose

Sugar	Species	No. of replicates	C mass balance (g)			% of original C used
			Blank medium	Fungusa	Respired CO2b
C3-derived sucrose	C. volvatus	16	0.440 ± 0.017	0.020 ± 0.006	0.015 ± 0.005	7.97 ± 2.42
	M. androsaceus	18	0.440 ± 0.017	0.065 ± 0.013	0.047 ± 0.009	25.38 ± 4.89
	S. granulatus	15	0.440 ± 0.017	0.015 ± 0.009	0.011 ± 0.007	5.83 ± 3.82
C4-derived sucrose	C. volvatus	3	0.438 ± 0.015	0.030 ± 0.010	0.022 ± 0.006	11.66 ± 3.70
	M. androsaceus	8	0.438 ± 0.015	0.061 ± 0.016	0.044 ± 0.012	23.66 ± 6.47
	S. granulatus	8	0.438 ± 0.015	0.017 ± 0.009	0.012 ± 0.007	6.44 ± 3.87

^aThe original fungal inoculum contained less than 0.5 × 10⁻⁶ g of C. 

^bThe amount of respired CO₂ was estimated by using the average ratio of CO₂ to fungal biomass obtained in closed systems (0.722) (Table 1). 

Isotopic analysis.

C isotopic composition was determined with a 20/20 stable-isotope-ratio mass spectrometer (Europa Scientific). For biomass, isotopic signatures represented total C and were not differentiated based on cellular fractions. Values are reported in the standard notation (δ¹³C; per mille) relative to the international standard Pee-Dee Belemnite using NIST Peach Leaves no. 1547 as a standard, where δ¹³C = [(R_sample/R_standard) − 1] × 1,000, and R is the ¹³C/¹²C molar ratio (29). Isotopic discrimination—the change in δ¹³C from the medium to the fungal biomass or the biomass to the CO₂—is reported as Δ = δ¹³C_final - δ¹³C_initial. Subscripts of Δ designate the reaction under consideration, so that Δ_b→f represents the discrimination observed from C uptake into fungal biomass (f) compared to the blank growth medium (b) and Δ_f→CO2 is the discrimination that results from fungal respiration. The blanks were controls containing uninoculated media that were treated like other experimental flasks. The instrumental precision values, estimated by using standard ground peach leaves (NIST no. 1547) for solid samples and pure CO₂ (Matson) for CO₂, were 0.02 to 0.06 and 0.04‰ (standard deviations for individual runs), respectively. Each sample was run twice, and values were averaged; the duplicate values were always within <0.07‰ of each other for solid C and within <0.05‰ for CO₂ samples. The values were corrected for linearity relative to the beam area of the standard. The values reported are averages from all experimental runs.

An “expected” δ¹³C value for media after fungal growth was estimated by calculating the molar amount of ¹³C involved in the shift from blank medium to fungal products, taking into account the total C in the biomass and its corresponding CO₂. This molar amount was subtracted from the equivalent value for blank medium to produce the expected δ¹³C value for the medium after fungal growth (Table 1).

Statistical analysis.

Statistical analyses were performed by using parametric and nonparametric statistical methods in JMP 3.2.1 (SAS Institute) according to the distributions and variances of the data. Since no difference was found between pure and impure sugars for either C₃ or C₄ sucrose (on C₃ sucrose, the t test values for M. androsaceus and S. granulatus were P = 0.495 and P = 0.949, respectively; on C₄ sucrose, the corresponding values were P > 0.696 and P > 0.125, respectively), the data reported correspond to data for all relevant runs performed with impure and pure sucrose. The results for C. volvatus represent pooled data obtained with two genetically different isolates.

RESULTS AND DISCUSSION

On C₃-derived sucrose, M. androsaceus, C. volvatus, and S. granulatus each had a characteristic isotopic enrichment in ¹³C relative to the substrate, and the Δ_b→f values were 0.67‰ ± 0.05‰, 4.87‰ ± 0.42‰, and 4.91‰ ± 0.47‰ (mean ± standard error), respectively (Fig. 1). On C₃-derived sucrose, significant taxon-specific isotope effects were suggested by the low variability exhibited by the two C. volvatus genotypes studied (standard error of Δ_b→f = 1.04‰) compared to the variability among different species (P < 0.01), although only one isolate of M. androsaceus and one isolate of S. granulatus were tested. This fractionation was maintained independent of incubation time or C₃ plant source (beet or maple). However, when organisms were cultured on C₄-derived sucrose (cane sucrose), isotopic fractionation was always reduced to negligible levels (Fig. 1). We concluded from these results that (i) there are significant isotopic effects caused by sucrose utilization by different basidiomycete species, (ii) these effects are determined by species-specific factors, and (iii) the isotopic effects can be markedly different depending on the origin of the sucrose available for fungal growth.

FIG. 1.

Isotopic fractionation by three basidiomycete species grown under ambient atmosphere conditions in liquid cultures on either C₃- or C₄-derived sucrose. The species used were C. volvatus (Cv), M. androsaceus (Ma), and S. granulatus (Sg). Isotopic discrimination (Δ_b→f) was calculated from δ¹³C values for the blank medium (Blank) and the total cell fungal biomass (cross-hatched bars). The standard error of the mean is also indicated for each bar. For each species, the solid bars represent the growth medium after fungal growth. The Δ_b→f values are significantly different from a hypothesized value of 0.00– for all C₃ treatments (t test; P < 0.001) and for C. volvatus (P = 0.04) and M. androsaccus (P = 0.007) on C₄ sucrose.

It has been suggested recently that isotopic fractionation by fungi is determined by the ecological role of fungi (27). Recent studies in various temperate forest ecosystems have shown that a distinction can be made between ectomycorrhizal and saprotrophic basidiomycetes, with the former showing depletion of ¹³C compared to the latter (22, 25, 27). This difference has been proposed as a diagnostic tool for these two major basidiomycete functional groups (22, 25, 27). Our results suggest that such a distinction must stem from a mechanism other than intrinsic discrimination by fungi. Ecological determinants, such as substrate effects, should be responsible for the patterns observed by other authors in the field, since we were unable to find a correlation between ecological role and intrinsic isotopic discrimination effects in the three species which we studied.

Our discovery of differential fractionation of C₃- and C₄-derived sucrose by different fungi not only reinforces the emerging view that microbial processing can involve significant species-specific discrimination of stable C isotopes but also signals that there is fine-scale regulation of such isotopic discrimination. Since the sucrose molecules in both C₃- and C₄-derived media are identical in chemical structure, differences in intramolecular atomic distributions must account for the observed differences in fractionation of each sugar type by basidiomycetes. Rossman et al. (38) have shown that ¹³C is not randomly distributed within the glucose molecule and also that the distributions in the glucose molecules produced by a C₃ plant (beet) and a C₄ plant (maize) are different. This differentiation is the only plausible point at which whole-cell differential discrimination could occur during fungal growth on either C₃- or C₄-derived sugars. There are two possible mechanisms mediating the selective accumulation of ¹³C-enriched products within the fungal cell, and steric enzymatic effects are logically ruled out due to the negligible difference between C₃- and C₄-derived glucose or sucrose in terms of the ¹³C/¹²C molar isotopic ratios in relation to total molecular weight (16). We therefore favor the alternative, that chemical species derived from C₃ sucrose having different isotopic ratios are routed through specific biochemical pathways at different kinetic rates, resulting in the observed total cellular isotopic discrimination.

To shed light on the physiological roots of this problem, we concentrated on the processing of C₃-synthesized sucrose by three basidiomycetes chosen because of their contrasting fractionation patterns and ecophysiological roles in pine-dominated ecosystems (Fig. 1). The direction of the fractionation that occurred on C₃-derived sucrose regardless of incubation conditions was always consistent, resulting in enrichment for ¹³C in the fungal biomass (Fig. 1 and 2). This enrichment was apparently balanced stoichiometrically by depletion of the same isotope in the medium (Fig. 1 and 2; Table 1). This stoichiometric balance was strongly suggested by the consistent direction of a small shift towards depletion in the medium, although the relatively large volume of medium used made the measured shift in δ¹³C too small to be statistically significant in some cases (Fig. 1 and 2; Table 1). The expected fractionation in medium values calculated from the amount of C contained in fungal biomass and its corresponding CO₂ is undistinguishable from the observed values (Table 1).

FIG. 2.

Isotopic discrimination by C. volvatus (Cv) and M. androsaceus (Ma) grown in closed liquid cultures on either C₃- or C₄-derived sucrose. Solid bars, medium; diagonally cross-hatched bars, fungal biomass; horizontally cross-hatched bars, CO₂. See the text for the definition of Δ. See Table 1 for mass balance data. The difference between biomass and CO₂ was significant only for C. volvatus grown on C₄ sucrose (t test; P = 0.032). The Δ_b→f values are significantly different from a hypothesized value of 0.00‰ only for C₃ treatments (t test; P < 0.01).

Since in these experiments isotopic measurements were obtained by using the whole organism (total biomass), enrichment of ¹³C in the fungal biomass could occur only from selective uptake of ¹³C-enriched C or from selective loss of ¹²C from the fungal biomass, presumably through release of depleted CO₂ from respiration. It is generally known that discriminating biochemical reactions normally result in apparent depletion of the heavier ¹³C isotope in the products compared to the reagents (17, 20), and Gleixner et al. (20) suggested that depleted CO₂ could account for their observation that there was significant enrichment of fungal tissues in the field. To test this “depleted-CO₂ hypothesis,” we produced closed-atmosphere culture conditions that allowed us to collect CO₂ from the headspace in incubation vessels. The results of these experiments contradict the depleted-CO₂ hypothesis, since the isotopic values for CO₂ were virtually identical to those for the fungal biomass and were not depleted as the hypothesis requires (Fig. 2). The Δ_f→CO2 values obtained for all fungi for all sugar types (C₃ or C₄) are not significantly different (P > 0.25) (Fig. 2), indicating that the observed difference in fractionation of C₃ and C₄ sugars cannot be attributed to differential isotopic effects on the production of respired CO₂. We concluded from these observations that the only explanation which can account for the observed enrichment for ¹³C in the fungal biomass must be found in C uptake mechanisms.

From these observations, we propose a model for C uptake in fungi involving two alternative routes leading to differential isotopic discrimination (Fig. 3). In the first, nonfractionating route, hexose molecules are brought into the cell without catabolism at a rate, r₁, so that discrimination between nonrandomly distributed ¹³C and ¹²C isotopes in these molecules cannot occur. In the second, fractionating uptake route, hexoses are broken down into triose fragments extracellularly before they are transported into the cell through two separate routes having different kinetic rates (r₂ and r₃, with r₂ > r₃). If a triose containing a proportionally enriched part of the C₃-derived hexose molecule is transported at rate r₂, enrichment of the cellular fraction would occur (Fig. 3). We refer to this model as the “dual-uptake hypothesis” of fungal isotopic discrimination. The fact that no cumulative fractionation effects were observed as cultures grew also supports the dual-uptake hypothesis rather than the depleted-CO₂ hypothesis, since if the latter hypothesis were true, a cumulative effect should have been observed as increasing amounts of depleted CO₂ were produced.

FIG. 3.

Proposed mechanism of differential isotopic enrichment in fungi grown on C₃ sucrose based on the known asymmetrical distribution of ¹³C and ¹²C in glucose (38). Extracellular digestion of glucose followed by uptake with different apparent uptake rates (r₂ and r₃) for the resulting trioses leads to comparative enrichment of the total cellular biomass. In contrast, direct uptake of glucose does not result in total cellular fractionation. The specific points of enrichment or depletion of glucose molecules depicted are illustrative and are not known for the specific sugars used in this study. An alternative fractionation mechanism could involve isotope asymmetry between fructose and glucose moieties in sucrose. See the text for details.

Rossman et al. (38) observed isotopic asymmetry in glucose which is much sharper in the C₃-derived sugar than in the C₄-derived sugar. For instance, the Δ values comparing the average isotopic composition of the whole molecule and those of individual atoms in the glucose ring analyzed after chemical degradation were as follows: for C atom number 3, C₃ glucose had enrichment in ¹³C with a Δ_{average→C-3} of +2.2‰, while C₄ glucose had an equivalent Δ_{average→C-3} of +1.2‰; for C atom number 6, C₃ glucose was depleted with respect to the average molecular value by a Δ_{average→C-6} of −5.2‰, whereas C₄ had a Δ_{average→C-3} of −4.0‰ (38). Other differences between C₃ and C₄ glucose were also shown by Rossman et al. (38), and while it would be tempting to use the numerical values of these authors to provide details of the fractionation mechanism proposed here, this would be inappropriate, given that the chemical and fermentation procedures used by Rossman et al. produced numerically different results and given the fact that we used sugars with different origins than the origins of the sugars used by these authors (H. L. Schmidt, personal communication, 1999). Nevertheless, the results of Rossman et al. clearly identified (i) the existence of an asymmetrical, nonrandom distribution of stable isotopes within the glucose molecule and (ii) a difference between C₃- and C₄-derived glucose in the δ¹³C values of individual C atoms, which provides the only currently plausible explanation for our observations of differential fungal fractionation of C₃ and C₄ sucrose.

A dual-uptake mechanism can also be proposed on the basis of our results showing that isotopic asymmetry could be found between the fructose and glucose moieties in C₃-derived sucrose. Direct sucrose uptake would not result in total cellular enrichment, whereas extracellular cleavage of the disaccharide into its monosaccharide components, followed by uptake with different uptake rates for fructose and glucose, could generate the observed enrichment. Although this mechanism is conceptually equivalent to the one illustrated in Fig. 3, there is no experimental evidence which suggests that direct sucrose uptake occurs in fungi or that there should be differential C isotopic distribution between glucose and fructose. We were unable to obtain a reliable source of C₃-derived glucose to eliminate this possibility.

Earlier work at the subcellular level by Monson and Hayes with Escherichia coli (32) and Saccharomyces cerevisiae (31) also supports the hypothesis that there is an isotope effect in microbial metabolism derived from the asymmetrical distribution of C isotopes in glucose molecules and intermediate metabolites. These authors found that unsaturated fatty acids derived from different C atoms of acetyl coenzyme A in E. coli differed in their isotopic signatures, with the moieties derived from the C-1 position showing depletion in ¹³C relative to the source glucose, while the moieties derived from the acetyl coenzyme A C-2 position displayed a signature similar to that of the source glucose (32).

In spite of evidence provided here, the dual-uptake mechanism must remain hypothetical in the absence of quantitative analytical values for the specific sugars and fungi utilized in our experiments. Nevertheless, the dual-uptake hypothesis is compatible with known differences in sugar usage during fermentative and respiratory processing (26). It is also known that glucose can be taken up directly through specific transporter proteins for respiratory catabolism (6, 26). Fungi may catabolize glucose and/or fructose via an extracellular aldolase into trioses, but conclusive evidence that this occurs is not available; however, several kinases, as well as glucose oxidase, modify glucose extracellularly (6, 14, 26, 33), and cell-free fermentation is well documented (14, 26). Significantly, we observed a shift towards greater enrichment of fungal biomass grown on C₃ sucrose as incubation conditions were changed from fully aerated to the closed-atmosphere conditions in our CO₂ capture experiments (Fig. 2). When M. androsaceus was grown under reduced O₂ tension, significant fractionation occurred, although discrimination effects for this species were negligible when it was grown in a fully aerated culture. M. androsaceus was enriched in ¹³C by 1.01% under reduced O₂ tension compared to the ¹³C level in the aerated culture (150% increase). A similar shift (1.91‰ or a 40% increase) was observed for C. volvatus. Our results therefore suggest that isotopic discrimination in basidiomycetes can occur when C₃-synthesized hexoses are processed through biochemical pathways that involve extracellular cleavage of the substrate sugar and that the degree of discrimination might be correlated with environmental conditions that favor extracellular sugar processing, such as microenvironments with reduced [O₂].

It could be suggested that the differential fractionation observed here could be the result of differentially depleted metabolites, such as alcohol and organic acids released into the medium by the fungus. We find this alternative explanation untenable given that (i) the CO₂ yields in our closed cultures (up to 0.85 times the amount of C in fungal biomass [Table 1]) suggest that proton acceptors other than oxygen were not predominant in terms of mass; (ii) our wild cultures and nonoptimized culture conditions would be unlikely to produce metabolites predominantly derived from a single biochemical pathway in all fractionating species; and (iii) differentiation of C₃- and C₄-derived sugars after the accumulation of stochastic effects due to multiple enzymatic reactions and intracellular recycling would be highly unlikely. den Hollander et al. (8) have shown that metabolic intermediaries corresponding to glucose carbons 1, 2, and 3 and 4, 5, and 6 are scrambled by aldolase and triose phosphate isomerase during glycolysis in S. cerevisiae. The intermediate sugars [1- and [6-¹³C]fructose 1,6-bisphosphate and the metabolic end products [1- or [3-¹³C]glycerol and [2-¹³C]ethanol result when S. cerevisiae is grown on [1-¹³C]glucose (8). Similar results were obtained for growth on [6-¹³C]glucose (8). The CO₂ values reported here (Fig. 2) are compatible with such an atomic scrambling effect, since they cannot be differentiated from values obtained for the cell as a whole. The consistency of our results, as shown by the small error values (Fig. 1 and 2), therefore suggests that differentiation between C₃ and C₄ sugars must occur during early processing stages, such as uptake, while the isotopic differences in the sugars are still maintained by their stereochemical configurations.

We identified two major factors, substrate atomic composition and microenvironmental conditions, which drive isotopic discrimination (or the lack thereof) by fungi. Our results establish the need to explicitly address the assumption that no isotopic fractionation occurs during transfers between trophic levels whenever material exchanges involve a fungal interface. Furthermore, we identified the need to determine the relative influence of physiological and substrate effects in each specific ecological context. Nonfractionation by fungi cannot be invoked without independent evidence, but conversely, fractionation cannot be assumed to happen in fungally mediated ecosystem processing unless it is independently supported or at least not ruled out by substrate quality and specific microenvironmental conditions. Careful in situ measurements correlated with laboratory simulations will be exceedingly useful in deciding whether fractionating or nonfractionating catabolic physiology is dominant in a given ecological situation.

Our results also provide a mechanistic hypothesis to account for the patterns of C isotope distribution in the field, which have remained generally unexplained. For example, it is generally known that there is a pattern of ¹³C enrichment with depth in well-aerated soil profiles in various ecosystems (5, 9, 34, 36), but there has been no conclusive explanation for this pattern. Two competing hypotheses have been proposed to explain this major isotopic distribution pattern: (i) the incorporation in more recent soil layers of plant materials derived from atmospheric CO₂ that is known to have experienced a historical trend towards depletion of ¹³C (12) and (ii) the cumulative effect of decomposition processes, particularly the selective preservation of presumably enriched plant components in the soil organic matter (2, 3, 5, 34). While the first of these hypotheses has not been critically tested, chemical analyses of plant components indicate that the selective preservation of lignin and its derivatives should result in depletion of ¹³C in deeper soil layers, which contradicts empirical observations (4, 34, 39). Focusing on microbial biomass, our results suggest that the observed increase in ¹³C abundance in deeper soil layers can be the direct result of isotopic discrimination by fungi during decomposition, since we show here that intrinsic metabolic processing makes fungal biomass enriched in ¹³C when fungi are presented with C₃-derived sucrose (Fig. 1 and 2). Compounds derived from microbial biomass can account for a large percentage of the C contained in soil organic matter, and the accumulation over time of such compounds in the form of recalcitrant organic matter in deeper soil layers would therefore result in enrichment for ¹³C with soil depth, as empirically observed.

Our results showing differential fractionation between C₄- and C₃-derived substrates can seriously affect statements based on isotopic evidence, particularly when land use change between C₃- and C₄-dominated ecosystems is studied. Field data supporting the importance of recognizing differential fractionation between C₃ and C₄ substrates has been provided by a recent comparative analysis of Brachiaria humidicola (a C₃ legume) and Desmodium ovalifolium (a C₄ grass), which showed that plant materials became enriched in ¹³C during the decomposition process for the C₃ plant but not for the C₄ legume under identical experimental conditions (39). Therefore, it can be proposed that the combination of ecosystem-specific photosynthetic and decomposition physiologies is paramount in determining the natural distribution and processing of stable C isotopes in terrestrial ecosystems.

The increasing reliance on stable isotope analysis to understand ecosystem processing must take into account important fractionation effects mediated by fungal interfaces. Although some of these effects might conveniently mask each other, it is important to recognize that species-specific fractionation of stable C isotopes by fungi does occur and that such fractionation is finely dependent on the interaction of specific physiological processing, substrate effects, and microenvironmental conditions.