Rapid peptide metabolism: A major component of soil nitrogen cycling?



[1] Proteinaceous and peptidic nitrogen is a potential direct nutrient source for both plants and microbes in the soil, without prior degradation to amino acids and mineralization. We used a series of five sites along an elevation gradient from 15 m a.s.l. to 710 m a.s.l. along which primary productivity decreases to investigate peptide utilization rates by soil microbes. Using 14C-labeled L-alanine, L-dialanine, and L-trialanine in a series of incubation experiments, we show that peptides are directly and rapidly assimilated by soil microbes, and that they are utilized for both biomass production and respiration. Alanine, dialanine, and trialanine were mineralized rapidly by soil microbes from the five sites along the gradient. Across all five sites, dialanine and trialanine were mineralized faster than alanine. In competition experiments, a 100-fold excess of alanine had no effect on the rate of trialanine mineralization in four of the five sites, and the same excess of trialanine had no effect on alanine mineralization. This is indicative of uptake of the intact peptide by the soil microbial community. Our findings have implications for understanding terrestrial nitrogen cycling because they point to a short-circuit whereby large peptides and proteins need only be extracellularly cleaved to short chain length peptides before direct assimilation by microbes.

1. Introduction

[2] A major goal in ecosystem science is to understand the fate of organic carbon (C) and nitrogen (N) in, and the processes regulating the turnover of, soil organic matter (SOM) [Jones et al., 2009a]. N cycling is performed predominantly by soil microbes [van der Heijden et al., 2008] and recently, debate has returned to identifying the rate-limiting steps in the microbially mediated decomposition of SOM to release plant-available N [Schimel and Bennett, 2004]. The traditional view was that only mineral N (in the form of NO3 and NH4+) can be utilized by plants [Aber and Melillo, 2001], and that the ammonification step regulates plant nutrient availability and primary productivity. However, it is now accepted that plants can directly take-up free amino acids (FAAs), bypassing the potential bottleneck of the mineralization step mediated by microbes [Schimel and Bennett, 2004; Jones et al., 2005]. There is unequivocal evidence that plants of many ecosystems are capable of taking up amino acids from soil [Kielland, 1994; Näsholm et al., 1998; Nordin et al., 2001; Bardgett et al., 2003]. However, FAAs are only a minor component of the DON pool, which is both chemically complex and varies in composition across ecosystems [Jones et al., 2005]. What is critical, but less understood, is not the concentration of various N species in soil solution, but their fluxes through the soil solution, a process which is very rapid for FAAs [Jones et al., 2009a]. Together with limitations in 15N tracer experiments [Jones et al., 2005], this rapid turnover could mean that the importance of FAA to plants and microorganisms has been underestimated.

[3] One remaining question is whether FAAs are the only significant source of organic N used by plants and microbes? Solid-state 15N NMR spectroscopy suggests that a large proportion of organic N in soil is the form of proteinaceous or peptidic moieties due to the high prevalence of amide bonds [Kögel-Knabner, 2006]. Jones et al. [2004] suggested that the primary block to N mineralization in neutral grasslands is the conversion of proteins to FAAs, and not the mineralization of FAAs to NH4+, with later experiments demonstrating that the rate of conversion of peptides to FAAs was similar to that of FAAs to NH4+ [Jones et al., 2009b].

[4] Several studies have demonstrated the ability of cultured microbes, including mycorrhizae, to assimilate peptides intact especially under N-limited conditions [Bajwa and Read, 1985; Benjdia et al., 2006; Wolfinbarger, 1980]. Further, it has been demonstrated in culture that peptides may be superior to their component amino acids as N sources for growth [Matthews and Payne, 1980]. Bacteria are known to have non-specific peptide transporters which are separate from individual amino acid transport systems [Payne, 1968; Barak and Gilvarg, 1975; Walker and Altman, 2005]. Peptides arising from extracellular protease action may therefore be a significant N source for bacteria and fungi. When direct intact uptake by plants is considered, peptide transporters have been observed in various tissues including roots, although there has been little consideration of the nutritional and ecological significance of this for plants potentially competing for N at an earlier stage of protein cleavage than FAAs [Steiner et al., 1994; Waterworth and Bray, 2006; Komarova et al., 2008; Hill et al., 2011a]. Thus, the limiting step in the soil N cycle may be the cleavage of proteins into peptides by proteases [Jones et al., 2009b].

[5] Given apoptosis in microbial and plant systems [Hoeberichts and Woltering, 2003], it is likely that a degree of protein degradation may occur prior to the deployment of microbial extracellular proteases. Thus leakage of degradation products from cells may occur, yielding free peptides in the soil solution and generating a localized soil environment akin to a ‘necrosphere’. We hypothesize that this pathway generates a ready source of free peptides that can be directly utilized by the soil microbial community, bypassing the need for peptidase-mediated hydrolysis to FAAs.

[6] There is little recent work on microbial utilization of peptides within soil. The purpose of this paper is to answer two questions: can small peptides be taken up intact by soil microbes across a wide range of environmental and soil conditions; and does the availability of peptides decrease amino acid utilization or vice versa? We investigated these questions across a well characterized elevation gradient in Snowdonia, North Wales [Farrell et al., 2011], in which primary productivity declined with increasing altitude, and DON increased in dominance of the TDN pool from a DON:DIN ratio of 0.3 in the lowest altitude, high productivity site; to a ratio of 5.1 at the highest altitude, low productivity site. We hypothesized that peptides will be rapidly taken up intact by microbes across the elevation gradient; although, as rates of N turnover are expected to be lower in low-productivity systems, we hypothesized that this rate of uptake would decrease with decreased primary productivity. In addition, we hypothesized that the uptake of peptides by soil microbes will be unaffected by the presence of excess amino acids and vice versa, due to their separate transporters and uptake pathways.

2. Materials and Methods

2.1. Site Description

[7] We used a grassland productivity gradient across a north-facing slope above Abergwyngregyn, Gwynedd, UK. Five sites were selected to reflect different soil characteristics and altitudes, with four randomly positioned replicate plots (10 m2) at each site. Soil and vegetation conditions along the elevation gradient are reported by Farrell et al. [2011] and this provided a basis for testing the hypotheses that peptides will be rapidly taken up by microbes and that this rate of uptake would decrease with decreased primary productivity due to reduced DON turnover rates. Briefly, the gradient is characterized by an increase in soil organic matter from 7.8% in the Eutric cambisol to 94.6% in the Fibric histosol, and a decrease in pH from 6.9 in the Eutric cambisol to 4.1 in the Haplic podzol and 4.2 in the Fibric histosol. We provide outline methodological information and data (Table 1) from that work to allow us to provide a context for the data presented in the current study. Aboveground standing biomass and net annual primary productivity (ANPP) were determined between April and October 2009 according to Vile et al. [2006]. We also estimated N deposition at each of the sites from NEGTAP (Transboundary air pollution: Acidification, eutrophication and ground-level ozone in the UK, 2001, http://www.freshwaters.org.uk/resources/documents/negtap_2001_final_report.pdf, accessed 04/03/11), with a 23% increase to account for organic N deposition [González Benítez et al., 2009]. Details are provided in Table 1.

Table 1. Characteristics of the Five Field Sitesa
 Eutric CambisolDystric GleysolCambic PodzolHaplic PodzolFibric Histosol
  • a

    Different letters on each row denote significant differences (P < 0.05) between sites. Soil data presented on a dry mass basis by area to 15 cm depth.

  • b

    Analyzed directly on soil solution.

Location (NGR; all SH)653729651725656716655701668680
Altitude (m)1540320530710
Estimated N deposition (kg ha−1 y−1)1212202631
NVC classificationMG7MG6U4U4M19
Aboveground Net Primary Productivity (g dw m−2 d−1)5.39 ± 0.36a2.78 ± 0.83ab0.918 ± 0.339b0.835 ± 0.047b1.448 ± 0.92b
Aboveground N (g N dw m−2)26.4 ± 2.7a8.08 ± 1.92b8.97 ± 0.68b8.31 ± 0.78b7.18 ± 1.77b
Total soil N (g N m−2)562 ± 55b597 ± 45b646 ± 104b565 ± 104b259 ± 25a
Microbial N (g N m−2)88.9 ± 12.0b41.8 ± 9.9a11.1 ± 0.7a11.3 ± 5.1a32.4 ± 7.3a
DON (mg N m−2)b169 ± 58330 ± 39301 ± 74427 ± 204313 ± 42
NH4+-N (mg N m−2)b132 ± 26116 ± 2173.4 ± 36.8179 ± 8541 ± 5
NO3N (mg N m−2)b401 ± 50b39.7 ± 36a0.604 ± 0.528a55.6 ± 19.6a20.5 ± 4.45a
FAA-N (mg N m−2)b0.54 ± 0.180.22 ± 0.060.28 ± 0.111.07 ± 0.420.34 ± 0.14
Peptidic-N (mg N m−2)b1.74 ± 0.271.29 ± 0.391.71 ± 0.640.99 ± 0.290.23 ± 0.04

2.2. Soil Characterization

[8] Bulk density was determined using 100 cm3 cores as per Rowell [1994]. Total N of both soil and vegetation was determined using a Carlo Erba NA 1500 Elemental Analyzer (Thermo Fisher Scientific, Milan, Italy). Soil solution was extracted using the centrifugal-drainage method of Giesler and Lundström [1993]. Soil solution was analyzed for soluble N using a TOC-V-TN analyzer (Shimadzu Corp., Kyoto, Japan). Microbial N was determined by chloroform fumigation-extraction according to Voroney et al. [2008]. NO3 and NH4+ were analyzed colorimetrically using the methods of Miranda et al. [2001] and Mulvaney [1996] respectively. Total FAAs were analyzed by anion-exchange chromatography using a Dionex ICS-3000 (Dionex, Sunnyvale, USA) in integrated amperometry mode using an ED40 electrochemical detector, AminoPac column maintained at 30°C, and an alkaline mobile phase of NaOH and C2H3NaO2 (see Clarke et al. [1999] for details). The sub-1 kDa peptides fraction (representing oligopeptides of up to ca. 8–12 AA units) were separated by molecular weight filtration under positive pressure of N2 (Amicon 8010 stirred cell and Millipore YM1 filter membranes: Millipore Corp., Bedford, USA) [Ford and Lock, 1985], and then hydrolyzed with 6 M HCl in an Ar atmosphere at 105°C for 16 h [Amelung et al., 2006]. HCl was removed by vacuum-centrifugal evaporation before samples were reconstituted in 100 μl 18 MΩ water and analyzed by the same procedure as FAAs outlined above. Peptidic-N was then calculated by subtraction of FAA-N.

2.3. Effect of Peptide Chain Length on Mineralization Rate

[9] Soil from each replicate experimental plot (5 g fresh weight) at each site along the elevation gradient was weighed into individual polypropylene tubes (50 cm3), and 0.5 ml (10 μM, 1.66 kBq ml−1) of a uniformly radiolabeled 14C-alanine, 14C-dialanine or 14C-trialanine (American Radiochemicals Inc., St. Louis, MO) solution added (i.e., 5 sites, n = 4) individually to separate tubes. All amino acids and peptides were L-enantiomers. To trap 14CO2 evolved, a 6 ml polypropylene vial containing 1 ml 1 M NaOH was placed within each tube above the soil, and the tube hermetically sealed and maintained at 10°C. This temperature approximates the mean annual soil temperature across the five sites. To quantify rates of respired 14CO2, traps were removed 0.25, 0.5, 1, 3, 6, 24, 48, 72, 96, 120, 144 and 168 h after 14C label addition. After removal, the amount of 14CO2 trapped in the NaOH was determined by liquid scintillation counting after mixing with ScintiSafe 3 scintillation cocktail (Fisher Scientific Ltd.) and a Wallac 1409 scintillation counter (PerkinElmer Life and Analytical Sciences Inc.). After incubating for 7 d, the soil was shaken with 25 ml 0.5 M K2SO4 for 30 min at 150 rev min−1 to recover any 14C label remaining in the solution or the exchangeable phase [Kuzyakov and Jones, 2006]. The extracts were determined by liquid scintillation counting as described above.

2.4. Preferential Uptake of Peptides and Amino Acids by Soil Microbes

[10] To investigate whether soil microbes exhibit a preference for peptide- or amino acid-N, and to determine whether peptides were cleaved extracellularly to amino acids prior to uptake, we performed two experiments using a fixed concentration of: (a) 14C-trialanine (10 μM) in the presence of an increasing concentration of unlabeled alanine (0, 1, 10, 100, and 1000 μM); and (b) 14C-alanine (10 μM) in the presence of an increasing concentration of unlabeled trialanine (0, 1, 10, 100, and 1000 μM). These experiments ran for a period of 168 h using individual replicates from the five sites as described above. The addition of an excess of alanine may be expected to reduce trialanine uptake if extracellular lysis of the peptide bonds occurred before uptake as individual alanine monomers. Conversely, in vitro studies [Matthews and Payne, 1980] have demonstrated uptake of peptides in some microbes may be favored energetically over amino acids. Therefore, an excess of trialanine may be expected to reduce uptake and subsequent mineralization of the alanine monomer; however, as far as we know, this has never been tested in soils.

2.5. Data Analysis

[11] To estimate peptide half-life in soil, a double first-order exponential decay equation was fitted to the inverse of the mineralization data using a least squares optimization routine in SigmaPlot v11.0 (Systat Software Inc., Chicago, IL) where:

equation image

in which y is the amount of 14C remaining in the soil, t is time, Yr and Yb represent the amount of 14C-peptide partitioned into microbial respiration and biomass production, respectively, and k1 and k2 are the rate constants for these two components [Boddy et al., 2008; Jones et al., 2009a]. The respiration of low molecular weight compounds in soil is well known to agree with first-order kinetics [Paul and Clark, 1996; Saggar et al., 1999], allowing a simple conceptual partitioning into the two pools. Substrate half-life in the soil solution (t1/2) and microbial C assimilation efficiency i.e., the proportion of 14C used for microbial growth (microbial yield; Yc) were calculated [Roberts et al., 2007] as

equation image


equation image

respectively. This yield equation makes the assumption that all utilized substrate not mineralized was incorporated into microbial biomass or metabolites [Thiet et al., 2006]. The low (<1% of the total) recovery of added 14C in the K2SO4 extracts strongly suggests that this assumption was correct in our investigation.

[12] The rate of microbial amino acid and peptide uptake (ϕ) at either a theoretical fixed soil pore water concentration of alanine, dialaine or trialanine (10 μM) or field-derived estimate of concentration (Table 1) was calculated as follows

equation image

where Q is the soil solution concentration (μmol kg−1). Here we assume that uptake from the soil solution occurs at a comparable rate to its subsequent mineralization within pool Yr [Jones and Hodge, 1999], and that the material mineralized in the second, slower phase (as defined in equation 1) had been taken up at a rate represented by the first phase; and

equation image

for total uptake, accounting for C allocated microbial growth. All values were normalized to molar N.

[13] All statistical analyses were carried out in SPSS v17.0 (SPSS Inc., Chicago, IL). To assess the effects of soil chemistry on peptide mineralization kinetics, Pearson's correlations were carried out against the four components of the double first-order exponential decay model, and the half-life and microbial yield derived from these. A two-way general linear model was used to assess the effects of soil type, chain length, or peptide or amino acid addition, with chain length, or amino acid or peptide addition as the primary factor, and soil as the secondary factor. Tukey's honestly significant differences post-hoc test was used to separate groups.

3. Results

3.1. Site Characteristics

[14] Comprehensive site characterization data including individual FAA and peptidic-AA concentrations are available in Farrell et al. [2011], and key information is laid out below. As discussed previously, along the gradient aboveground net primary productivity (ANPP) [Vile et al., 2006] varied sixfold, although this was largely attributed to ANPP being significantly greater at site 1 than at sites 3–5; plant diversity was greatest at site 3. Excluding the Eutric Cambisol, which regularly receives mineral fertilizer addition, DON was the dominant form of soluble N within the soil solution, contributing between 61 and 83% of the total in the four sites. While there were no significant differences between total FAA-N and total peptidic-N concentration in the soil solution at each site, these compounds comprised from 0.45% to 1.35% of total DON in the Dystric Gleysol and Eutric Cambisol, respectively, as expressed on an area basis, leaving ≥99% of the DON unaccounted for at all sites.

3.2. Preferential Uptake of Organic N Moieties by Soil Microbes

[15] The exponential coefficient k1 was used to determine the preferentiality of use in terms of microbial mineralization of 14C-alanine and increasing amounts of trialanine, and vice versa, when the substrates were supplied together (Figure 1). Across the five soils along our elevation gradient, there were significant differences between 100 and 1000 μM and 1 and 10 μM additions of trialanine (F(4, 75) = 14.2, P < 0.001), but only the addition of 1 μM trialanine significantly increased the k1 value over the control treatment of 0 μM trialanine (P = 0.001). Soil type also had a significant effect on 14C-alanine utilization (F(4, 75) = 6.39, P < 0.001) with the soils from the two highest elevation sites (Fibric Histosol and Haplic Podzol) having a significantly lower k1 value for 14C-alanine than the Eutric Cambisol (P < 0.001 and P = 0.002 respectively). There was no significant interaction between soil type and 12C-trialanine addition (F(16, 75) = 1.111, P = 0.361), indicating that the effects of soil type on 14C-alanine mineralization rate were consistent across the five 12C-trialanine concentrations.

Figure 1.

The k1 values for mineralization of (a) 14C-alanine mineralized in the presence of increasing trialanine and (b) 14C-trialanine in the presence of increasing alanine. EC, Eutric Cambisol; DG, Dystric Gleysol; CP, Cambic Podzol; HP, Haplic Podzol; FH, Fibric Histosol. Different letters denote significant differences between clusters at the P ≤ 0.05 level as a result of concentration of the added unlabelled substrate. Due to the non-significance of the interaction between the unlabelled substrate addition and soil type, these are grouped for clarity.

[16] Increasing concentrations of 12C-alanine reduced the k1 values for the respiration of 14C-trialanine (F(4, 75) = 7.456, P < 0.001), indicating that a large excess of amino acid can reduce microbial mineralization of low concentrations of peptides in soils. However, this was only significant for the addition of 1000 μM alanine (P = 0.001) (Figure 1b). Again, the two highest sites (Haplic Podzol and Fibric Histosol) in our gradient gave significantly lower k1 values (F(4, 75) = 14.90, P < 0.001) than the three lower sites, and as no significant interaction (F(16, 75) = 1.076, P = 0.393) was observed between site and 12C-alanine addition, the effects of soil type on 14C-trialanine mineralization rate were consistent across the five 12C-alanine concentrations.

3.3. Effect of Chain Length on Mineralization and Uptake of Alanine-N Moieties

[17] Alanine and the two peptides were mineralized rapidly in all five soils along the productivity gradient. In all cases, the mineralization of the 14C-labeled alanine and peptides showed a biphasic 14CO2 evolution (Figure 2); a double first order exponential decay generally described the experimental mineralization data well (r2 = 0.994 ± 0.001). The rates of mineralization differed between the compounds (F(2, 44) = 85.3, P < 0.001), and as an average across the gradient, trialanine had a half-life of 1.25 ± 0.30 h, dialanine 1.91 ± 0.15 h, and alanine 3.17 ± 0.18 h (Table 2). In this first rapid phase of respiration (attributable to pool Yr), the amount of dialanine and trialanine respired was around double that of alanine (32 ± 1, 26 ± 2%, and 15 ± 1% respectively), and this is reflected in the significant differences observed in microbial yield (Yc) between the substrates. Despite the Eutric Cambisol being the only soil to contain a significant amount of inorganic N (5.3 ± 0.8 kg N ha−1) in the soil solution, it did not differ significantly from the other soils with regard to substrate half-life (P > 0.05). However, differences were observed between the Dystric Gleysol and the two highest soils (Fibric Histosol and Haplic Podzol; P < 0.05), and the Cambic Podzol also demonstrated significantly faster 14C-substrate turnover than the Haplic Podzol. As there was no significant interaction between peptide chain length and soil type (F(8, 44) = 1.741, P = 0.116), changes in turnover rates were consistent across the three chain-lengths of 14C-alanine added between the different sites. Only low proportions of substrate added at the start of the experiment were recoverable in the K2SO4 extracts at the end (≤1% of the total 14C added) indicating that >99% of the substrates were taken up by the microbial biomass. As a result, we make the assumption that the 14C which was not respired was immobilized within the soil microbial biomass [Kuzyakov and Jones, 2006].

Figure 2.

The 14CO2 production after addition of 14C-alanine, dialanine or trialanine to the five soils in this study: (a) Eutric Cambisol, (b) Dystric Gleysol, (c) Cambic Podzol, (d) Haplic Podzol, (e) Fibric Histosol. The lines are fits of a double first-order exponential decay model (r2 of all curves fitted across the full time period was >0.99; values are means ± SEM, n = 4).

Table 2. Rates of Mineralization (Expressed as Half-Life) and Yield (Microbial C Use Efficiency) Analyzed by Two-Way ANOVA for the Substrate (Alanine and its Dipeptides and Tripeptides) and Soil Type (F Soils on a Productivity Gradient) Variables
Half-life (t1/2)Substrate85.3<0.001
Microbial yieldSubstrate87.8<0.001

[18] To estimate the relative rate of microbial N acquisition from alanine and the peptides, we calculated their uptake rate (ϕ) assuming an equal soil solution concentration of 10 μM (Table 3). Overall, the uptake of N from either di- or tri- alanine was calculated to be faster than for alanine alone, and this trend was conserved across the gradient. In particular, when this is expressed as kg N per unit land area, by correcting for soil moisture content and bulk density, we estimated a range of microbial uptake rates of trialanine ranging from 42 to 595 kg N ha−1 y−1, with microbial N uptake being greatest in the soil supporting the highest primary productivity and highest DIN availability (Eutric Cambisol; F(4) = 12.4, P < 0.001). These uptake estimates were also positively correlated with ANPP at the sites across alanine (r2 = 0.624, P = 0.010) and trialanine (r2 = 0.571, P = 0.021), and weakly correlated for Dialanine (r2 = 0.463, P = 0.071). Averaged across the five sites, N uptake rate followed the series: alanine < dialanine < trialanine (F(2, 45) = 96.0, P < 0.001). There were no significant correlations (P > 0.05) between any of the physicochemical soil characteristics and the microbial yield (Yc) and soil solution half-life of alanine and the two peptides. To relate these estimates back to measured soil solution concentrations of FAA-N and hydrolyzed peptidic-N, we used the k1 values derived from the dialaine and trialanine additions above to estimate microbial N uptake at measured FAA- and peptidic- N concentrations (Table 4).

Table 3. Estimates of Total N Uptake by Allowing for the Amount of C Mineralized in the Second Phase of Respiration, and the Rate of N Uptake Based on the First Phase of Respiration Onlya
SoilCalculated Assuming 10 μM Concentrations
SubstrateRate of Uptake of Alanine-Equivalents Allowing for C Allocated to Growth (μmol g−1 fwt soil h−1)Rate of Uptake of N (mg N m−2 d−1)
  • a

    Rates of uptake of alanine, di-, and tri- alanine are derived from the kinetics of mineralization, and calculated on the basis of the results from our 10 μM addition of 14C-labeled substrate. Different letters within soil types indicate significantly different (P < 0.05) rates between substrates in that soil type.

Eutric cambisolAlanine1.73 × 10−3 ± 2.11 × 10−4 a19.3 ± 1.2 a
Dialanine2.16 × 10−3 ± 5.52 × 10−4 a48.7 ± 8.7 a
Trialanine7.79 × 10−3 ± 7.79 × 10−4 b163 ± 15 b
Dystric gleysolAlanine1.91 × 10−3 ± 3.05 × 10−4 a12.6 ± 3.3 a
Dialanine4.04 × 10−3 ± 4.87 × 10−4 a55.4 ± 11.2 a
Trialanine1.47 × 10−2 ± 4.36 × 10−3 b133 ± 28 b
Cambic podzolAlanine1.85 × 10−3 ± 1.72 × 10−4 a8.37 ± 2.42 a
Dialanine2.64 × 10−3 ± 1.94 × 10−4 a28.8 ± 5.5 a
Trialanine8.09 × 10−3 ± 5.23 × 10−4 b74.9 ± 15.3 b
Haplic podzolAlanine1.32 × 10−3 ± 2.48 × 10−4 a5.64 ± 1.00 a
Dialanine1.74 × 10−3 ± 1.87 × 10−4 a15.8 ± 4.4 a
Trialanine1.21 × 10−2 ± 1.36 × 10−3 b58.7 ± 13.9 b
Fibric histosolAlanine1.21 × 10−3 ± 1.88 × 10−4 a1.44 ± 0.28 a
Dialanine2.56 × 10−3 ± 2.34 × 10−4 b5.66 ± 0.73 ab
Trialanine5.72 × 10−3 ± 3.30 × 10−4 c11.5 ± 3.7 b
Table 4. Rates of Uptake Calculated Using Measured FAA-N and Hydrolyzed <1 kDa Peptidic-N Concentration in Soil Solutiona
SoilRate of Uptake of N (mg N m−2 d−1)
Amino AcidDipeptideTripeptide
  • a

    Here we assume an amino acid- and peptidic-N content of 16.8 μg per μmol FAA/peptidic-AA, that for ease of representation here k1 values of dialanine and trialanine reflect those of the many different peptides likely present in the soil solution, and that all <1 kDa hydrolyzed peptidic N was originally present as di- or tri- peptidic moieties prior to hydrolysis for analysis.

Eutric cambisol4.32 ± 1.7933.1 ± 11.3107 ± 28
Dystric gleysol1.76 ± 0.6041.1 ± 14.794.6 ± 32.0
Cambic podzol2.03 ± 0.8645.6 ± 17.9121 ± 52
Haplic podzol8.57 ± 4.5517.4 ± 4.767.6 ± 17.9
Fibric Histosol1.83 ± 0.507.36 ± 0.8816.1 ± 4.9

[19] Overall, a relatively low proportion of the peptidic-C taken up soil microbes was rapidly respired (i.e., that attributed to pool Yr; Table 5), with the majority of the C contained in the three substrates being immobilized within the microbial biomass (average value for Yb across all treatments = 77 ± 1%, n = 59). The microbial yield followed the series: alanine > trialanine > dialanine (F(2, 44) = 87.8, P < 0.001). There was a significant interaction between chain length and soils for microbial yield (F(8, 44) = 2.641, P = 0.019), indicating that differences in the microbial partitioning of the added 14C to anabolic or catabolic processes between chain-lengths of added substrate were not consistent across the gradient. In particular, more trialanine was partitioned to the microbial biomass in the Haplic Podzol than in the other soils studied.

Table 5. Rate Constants and Microbial Yields From Fits of Double First-Order Exponential Decay Equations to Data on CO2 Mineralization of Alanine, Dialanine, and Trialanine
SubstrateVariableEutric CambisolDystric GleysolCambic PodzolHaplic PodzolFibric Histosol
AlanineYr (%)16.0 ± 1.013.3 ± 0.813.3 ± 1.715.6 ± 0.916.7 ± 1.9
 k1 (h−1)0.273 ± 0.0210.252 ± 0.0390.239 ± 0.0270.199 ± 0.0260.198 ± 0.022
 Yb (%)83.8 ± 1.088.6 ± 0.886.9 ± 1.684.8 ± 1.083.5 ± 1.8
 k2 (h−1)5.99 × 10−4 ± 5.70 × 10−53.82 × 10−4 ± 2.60 × 10−56.88 × 10−4 ± 3.00 × 10−56.57 × 10−4 ± 1.04 × 10−46.43 × 10−4 ± 1.22 × 10−4
 t1/2 (h)2.58 ± 0.172.96 ± 0.463.00 ± 0.323.66 ± 0.473.64 ± 0.43
 Microbial yield0.840 ± 0.0100.867 ± 0.0080.867 ± 0.0170.845 ± 0.0090.832 ± 0.019
DialanineYr (%)32.1 ± 1.128.7 ± 1.833.7 ± 1.530.5 ± 1.432.5 ± 4.7
 k1 (h−1)0.346 ± 0.0670.566 ± 0.0390.438 ± 0.0160.261 ± 0.0210.404 ± 0.062
 Yb (%)67.2 ± 1.071.5 ± 2.067.0 ± 1.470.0 ± 1.468.6 ± 4.36
 k2 (h−1)8.79 × 10−4 ± 1.43 × 10−46.80 × 10−4 ± 1.49 × 10−45.51 × 10−4 ± 5.70 × 10−55.38 × 10−4 ± 6.40 × 10−58.06 × 10−4 ± 1.62 × 10−4
 t1/2 (h)2.19 ± 0.321.24 ± 0.091.59 ± 0.062.71 ± 0.251.84 ± 0.27
 Microbial yield0.677 ± 0.0100.714 ± 0.0930.665 ± 0.0140.696 ± 0.0140.679 ± 0.046
TrialanineYr (%)29.7 ± 0.322.0 ± 3.928.8 ± 0.917.2 ± 1.733.1 ± 1.3
 k1 (h−1)0.771 ± 0.0860.922 ± 0.0760.756 ± 0.0310.663 ± 0.0350.616 ± 0.051
 Yb (%)70.7 ± 0.378.4 ± 3.773.2 ± 0.884.1 ± 1.5569.6 ± 1.1
 k2 (h−1)7.41 × 10−4 ± 1.92 × 10−46.12 × 10−4 ± 1.33 × 10−46.24 × 10−4 ± 9.10 × 10−54.82 × 10−4 ± 1.30 × 10−46.02 × 10−4 ± 1.81 × 10−4
 t1/2 (h)0.928 ± 0.0860.769 ± 0.0690.922 ± 0.0391.05 ± 0.051.14 ± 0.09
 Microbial yield0.704 ± 0.0030.782 ± 0.0390.718 ± 0.0090.830 ± 0.0160.678 ± 0.012

3.4. Internal Partitioning of Amino Acids and Peptides by Soil Microbes in the Presence of Competing Substrates

[20] As discussed above, the C assimilation efficiency was greater for alanine-C (Yc = 0.82 ± 0.01) than trialanine-C (Yc = 0.60 ± 0.01). The microbial yield was significantly decreased by the addition of 1000 μM trialanine to the 10 μM 14C-alanine (F(4, 75) = 9.65, P < 0.001; Figure 3a), although smaller additions had no effect (P > 0.05). The effect of increasing amounts of alanine on 14C-trialanine-C partitioning by soil microbes yielded significant differences between alanine additions (F(4, 75) = 14.1, P < 0.001; Figure 3b), however, there were no differences between the highest and lowest alanine additions (P > 0.05). There were significant interactions between unlabeled substrate addition rate and soil type for both 14C-alanine and 14C-trialanine carbon partitioning (F(16, 75) = 2.207, P = 0.012; and F(16, 75) = 2.875, P = 0.001 respectively). Changes in C assimilation efficiency with increased unlabeled substrate addition were not consistent across the gradient, and differed between soil types. Alanine-C assimilation efficiency was lowest in the Fibric Histosol at low trialanine concentrations, whereas it was lowest in the Eutric Cambisol in the presence of 1 mM trialanine. Similar patterns were observed for trialanine-C assimilation efficiency in the presence of increasing alanine (Figure 3).

Figure 3.

Microbial yield (as immobilization-to-mineralization ratio; equation 3) of (a) 14C-alanine in soil solution in the presence of increasing trialanine and (b) 14C-trialanine in the presence of increasing alanine. EC, Eutric Cambisol; DG, Dystric Gleysol; CP, Cambic Podzol; HP, Haplic Podzol; FH = Fibric Histosol.

4. Discussion

4.1. Direct Uptake of Peptides by the Soil Microbial Community

[21] The average half-life for mineralization of trialanine was less than half that of alanine, indicating extremely fast cycling of peptides in the soil solution relative to the amino acid monomer. This supports our hypothesis that peptides are ubiquitously rapidly utilized by the soil microbial community and potentially supports the argument that there is likely to be competition between microbes and plants for the peptide pool [Hill et al., 2011b]. While rates of amino acid utilization and turnover in soil are very high [Harrison et al., 2007; Kielland et al., 2007; Roberts et al., 2007; Jones et al., 2009a], turnover rates of peptides have not been quantified before. Here we demonstrate extremely fast mineralization of di- and tri-peptides in a broad range of soil types across an elevation gradient (Figure 2 and Table 2). Given that no significant correlations (P > 0.05 in all cases) were observed between DON, NO3, NH4+ and t1/2 or Yc of any of the substrates, we conclude that utilization rates and internal partitioning of amino acids and peptides by soil microbes are generally unaffected by the N-status of the soil. Such high turnover rates of peptidic-N could have implications for the ways in which terrestrial N cycling is considered, indicating that soil microbes may compete for peptides as well as amino acids (see below for quantification).

[22] Given this extremely rapid metabolism, uptake from the soil solution by soil microbes may be even faster. Hill et al. [2008] demonstrated that while half-life estimates for cycling of glucose through the microbial biomass derived from isotopic mineralization studies ranged from 20 min to 4 h, actual microbial removal rates from soil solution were much faster (mean residence time of ca. 30 s for glucose). From this they estimated that the pool of glucose in root-free soil thus turns over 102–103 times per day. Consequently, we acknowledge that determination of turnover times reported here may underestimate turnover times in the soil solution.

[23] Pure culture studies have demonstrated that some bacteria have the capacity to take up peptides intact, without prior cleavage to their constituent amino acids [Payne, 1980]. Naider and Becker [1975] found a strain of E. coli able to take up a variety of tripeptides intact, and argue that specificity of transporters for individual peptides is unlikely to have evolved, and instead hypothesize that a few peptide transport systems with different affinities for differing peptide classes exist. It was subsequently found that Stokes' radius was the feature responsible for limiting peptide utilization (caused by a molecular sieve-like structure within the cell wall), and this value was roughly equivalent to 650 Da [Payne, 1980]. The molecular weight (MW) of trialanine is 231 Da, well within the sieve size-limit. Our results support the conclusion that peptides in soils are a readily accessible N-source for microbes, and due to the absence of a lag-phase in the mineralization (Figure 2), the implication is that they are taken up intact and this is consistent with these compounds being used as C substrates [Jones and Hodge, 1999; Jan et al., 2009; Jones et al., 2009b]. By inference, we can also assume that the N is most likely utilized directly. Indeed, Payne and Smith [1994] conclude in their extensive review that peptide transport is important across a wide range of bacteria, using similar transport and assimilation systems to those studied in detail in E. coli and Salmonella typhimurium. This implies that soil microbes may short-circuit the traditional N-cycle whereby proteins are extracellularly cleaved into polypeptides, which are in turn extracellularly cleaved to free amino acids, before microbial assimilation. To our knowledge this is the first study demonstrating such assimilation in soils.

[24] Not only are the di- and tri-peptides taken up rapidly, but the amount of N so-accumulated is at least as great, and often greater, when the di- and tri-peptides are compared with alanine (Table 3 and 4). This reinforces the idea of peptides being a significant source of N, and possibly C, for soil microbes. A rough calculation further supports this view. We compare the rates of N uptake from peptides expressed per unit area of land with likely rates of N turnover in these soils which are derived below. For the Cambic Podzol, the experimental uptake rate of trialanine was 75 mg N m−2 d−1; if all the peptidic-N at this site were trialanine the uptake rate calculated using the k1 for trialanine in this soil would be 121 mg N m−2 d−1 (Table 4). We calculate the demand from higher plants (from NPP and N content) as 11 mg N m−2 d−1, and from microbes (from microbial N and assuming a conservative t1/2 of 15 d) as 513 mg N m−2 d−1. Clearly uptake of trialanine could satisfy plant uptake, but account for only about 15% of the microbial requirement. While bulk N mineralization studies were not carried out in this study, a comparison with some published results reporting gross N mineralization rates of 200–300 mg N m−2 d−1 reveals that our estimate of N uptake rate is of the same order [Jarvis et al., 1996]. At a time when the debate about niche specificity and resource division between plants species [McKane et al., 2002] and microbes [van der Heijden et al., 2008; Jones et al., 2009a, 2009b] over N is of great interest, it is clear that peptidic-N must be considered as a potentially important, readily assimilable N-source for soil microbes. This is especially true given the positive correlation observed between NPP and microbial uptake of amino acids and peptides demonstrated in this study, and possibly indicates that microbial uptake of LMW DON may have a regulatory role in ecosystem productivity.

4.2. Utilization of Peptides by Soil Microbes in the Presence of Increasing Amino Acid Concentration

[25] We have demonstrated that di- and tri-alanine are readily assimilated individually by soil microbes, but will they be taken up in an excess of free amino acids, or does their presence reduce amino acid uptake? In our experiments, there was no significant interference of peptide uptake by FAAs, or vice versa, at any reasonable combination of concentrations. Only in the Haplic Podzol is trialanine mineralization decreased by very high alanine concentrations. Payne and Gilvarg [1978] state that although there is competition between amino acids for shared transporters, no such competition has been observed between free amino acids and peptide-bound amino acids which would theoretically use the same transporter. Several species (E. coli, Leuconostoc mesenteroides, Streptococci, Pseudomonas, Salmonella, Neurospora and Saccharomyces) show no inhibition of amino acid uptake by peptides, nor of peptides by free amino acids [Payne and Gilvarg, 1978]. These previous studies were performed in vitro on pure cultures. To our knowledge, no work has sought to address the importance of amino acids relative to peptides (or vice versa) for mixed microbial communities in soil. Very little is known about the process by which microbes compete for molecules directly within the soil medium, and which carbon and nitrogen substrates are favored. Jones et al. [2004] identified two divisible pools of DON: fast-cycling low MW (<1000 Da) such as amino acids, and slow-cycling recalcitrant high MW DON bound to polyphenols, etc. While the low MW DON pool represents a small proportion of total DON, it sustains most microbial nitrogen needs [Ford and Lock, 1985], and almost certainly contains more N compound types than FAAs. All this is consistent with peptides being potentially important in soil N cycling and microbial nutrition.

[26] The proportion of peptidic-N can vary significantly between soil types, with cultivation having some effect [Warman and Isnor, 1991]. Here, we quantified free peptidic-N at very low concentrations and at the same order of magnitude as FAA-N across five different soil types. It could have been expected that peptide and amino acid usage by microbes would be negatively correlated to total DON, NO3, or NH4+ [Farrell et al., 2011], reflecting increased uptake in N-limited soils. However, there were no significant correlations between peptide and amino acid uptake rates (Table 4) and these indices of N-availability. This suggests that the availability of forms of N, which may be less attractive to soil microorganisms, is a poor indicator of the flux through the pools of other more attractive forms of soil N, such as peptides and amino acids.

5. Conclusions

[27] We have demonstrated the rapid turnover of peptides in five contrasting soil types across an elevational gradient in which productivity decreased with increasing altitude. Across the gradient we observed higher rates of peptides turnover in comparison to the corresponding amino acid monomers. This higher rate of utilization occurred irrespective of soil N-status, and these rates were positively correlated with primary productivity. Alongside clear evidence from pure cultures, our in vivo results strongly suggest that the peptides are taken up intact by the soil microbial community. This finding could have a profound effect on our understanding of terrestrial N cycling, especially as the rates of peptide uptake are comparable with calculated acquisition rates of N by both plants and microbes. Previously, it was assumed that peptides must be extracellularly cleaved before microbial assimilation as amino acids. Our results are consistent with trialanine and alanine being taken up by independent transporters. Collectively, these findings pose questions as to where the bottleneck in the terrestrial N-cycle lies, and suggests that microbes and plants can compete directly for peptides at an early stage of the DON-degradation pathway.


[28] We gratefully acknowledge funding provided by the Natural Environment Research Council (grant NE/E017304/1) awarded to R.D.B., D.L.J., and J.F. Thanks also to Llinos Hughes, Mark Hughes, Jonathan Roberts, and Francis Guyver for their assistance in the laboratory and field.