Metabolic Pathways of Amino Acids, Monosaccharides and Organic Acids in Soils assessed by Position-Specific Labeling

I Metabolic Pathways of Amino Acids, Monosaccharides and Organic Acids in Soils assessed by Position-Specific Labeling Dissertation zur Erlangung des Grades Doktor der Naturwissenschaften (Dr. rer. nat.) an der Fakultät Biologie / Chemie / Geowissenschaften der Universität Bayreuth vorgelegt von Michaela A. Dippold (Dipl. Geoökologin & Dipl. Biochemikerin) geb. am 10.07.1982 in Bamberg Betreuer: Prof. Dr. Yakov Kuzyakov Bayreuth, den 11. September 2013 II Die vorliegende Arbeit wurde in der Zeit von 01.03.2010 bis 11.10.2013 in Bayreuth am Lehrstuhl für Agrarökosystemforschung unter der Betreuung von Herrn Prof. Dr. Yakov Kuzyakov angefertigt. Vollständiger Abdruck der von der Fakultät für Biolgie, Chemie und Geowissenschaften der Universität Bayreuth genehmigten Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.). Dissertation eingereicht am: 11.09.2013 Zugelassen durch die Prüfungskommission: 18.09.2013 Wissenschaftliches Kolloquium: 31.01.2014 Amtierender Dekan: Prof. Dr. Rhett Kempe Prüfungsauschuss: Prof. Dr. Y. Kuzyakov (Erstgutachter) Prof. Dr. C. Werner-Pinto (Zweitgutachter) Prof. Dr. E. Matzner (Vorsitz) PD Dr. M. Horn Drittgutachter: PD Dr. G. Wiesenberg Contents I I. Contents I. Contents ............................................................................................... I II. List of Figures....................................................................................IX III. List of Tables ................................................................................... XV IV. Abbreviations................................................................................. XVII V. Summary ....................................................................................... XVIII VI. Zusammenfassung.......................................................................... XX 1 Extended Summary.............................................................................................. 1 1.1 Introduction ...................................................................................................... 1 1.1.1 Low molecular weight organic substances in soils.......................................... 1 1.1.1.1 Role and relevance of low molecular weight organic substances in soil...1 1.1.1.2 Sources and sinks of LMWOS.................................................................2 1.1.2 Microbial utilization of low molecular weight organic substances.................... 4 1.1.2.1 Microbial uptake of LMWOS: the most competitive process determining the fate of LMWOS-C in soil ....................................................................4 1.1.2.2 Mineralization versus incorporation of LMWOS-C into microorganisms...5 1.1.3 Metabolic tracing by position-specific labeling ................................................ 6 1.1.4 Objectives ...................................................................................................... 8 1.2 Experiments and Methods............................................................................... 8 1.2.1 Field experiment............................................................................................. 8 1.2.2 Laboratory experiments.................................................................................10 1.2.2.1 Experiment 1: Transformations of free alanine.......................................10 1.2.2.2 Experiment 2: Transformations of sorbed alanine..................................11 1.2.2.3 Experiment 3: Plant uptake of intact alanine..........................................11 1.2.3 Methods to trace 13 C and 14 C in transformation products of LMWOS.............11 1.2.3.1 Bulk-isotope measurements by EA-IRMS..............................................12 1.2.3.2 Compound-specific isotope analysis of microbial biomarkers ................12 1.2.3.3 Radiochemical analyses........................................................................13 1.2.4 The Divergence Index ...................................................................................14 1.3 Results and Discussion..................................................................................15 1.3.1 Overview: main results of the studies ............................................................15 Contents II 1.3.2 Determination of metabolic pathways of amino acids, monosaccharides and organic acids.................................................................................................16 1.3.2.1 Similarities and differences of individual LMWOS..................................16 1.3.2.2 The main pathways for LMWOS metabolization by soil microorganisms19 1.3.2.3 Metabolic pathways for the formation of specific cellular compounds ....22 1.3.3 Identification of specific metabolic pathways..................................................24 1.3.3.1 Specific pathways of individual members of the microbial community in soils.......................................................................................................24 1.3.3.2 Pathways under various concentrations of LMWOS ..............................25 1.3.3.3 Pathways of sorbed LMWOS.................................................................27 1.3.3.4 Extraversus intracellular transformation pathways...............................29 1.3.4 Kinetics and ecological relevance of competing sinks for LMWOS................30 1.3.4.1 Sorption versus microbial utilization.......................................................30 1.3.4.2 Plant uptake versus microbial utilization ................................................30 1.4 Conclusions.....................................................................................................32 1.5 Reference List .................................................................................................35 1.6 Contribution to the included manuscripts and publications........................40 2 Publications and Manuscripts............................................................................44 2.1 Study 1: Fate of low molecular weight organic substances in an arable soil: from microbial uptake to utilisation and stabilisation ..................................44 Abstract ....................................................................................................................45 2.1.1 Introduction ...................................................................................................47 2.1.2 Material and Methods....................................................................................49 2.1.2.1 Experimental design..............................................................................49 2.1.2.2 Bulk soil δ 13 C analysis ...........................................................................50 2.1.2.3 Microbial biomass..................................................................................50 2.1.2.4 Phospholipid fatty acid analysis.............................................................51 2.1.2.5 Calculations and statistical analysis.......................................................52 2.1.3 Results..........................................................................................................53 2.1.3.1 Microbial community structure ...............................................................53 2.1.3.2 Microbial utilisation of LMWOS..............................................................54 2.1.3.3 Utilisation of LMWOS by functional microbial groups.............................55 2.1.4 Discussion.....................................................................................................59 2.1.4.1 Incorporation of LMWOS into SOM and microbial biomass....................59 2.1.4.2 Microbial community composition..........................................................62 2.1.4.3 Incorporation of LMWOS into PLFAs.....................................................63 Contents III 2.1.5 Conclusion ....................................................................................................66 Acknowledgements...................................................................................................67 References ...............................................................................................................68 Supplementary Data.................................................................................................72 2.2 Study 2: Improved δ 13 C analysis of amino sugars in soil by Ion Chromatography – Oxidation – Isotope Ratio Mass Spectrometry .............74 Abstract ....................................................................................................................75 2.2.1 Introduction ...................................................................................................76 2.2.2 Material and Methods....................................................................................78 2.2.2.1 Soil ........................................................................................................78 2.2.2.2 Chemicals, reagents and external and internal standards......................78 2.2.2.3 Soil hydrolysis and ion removal .............................................................79 2.2.2.4 Purification by cation exchange column.................................................79 2.2.2.5 Development of the measurement by IC-O-IRMS..................................79 2.2.2.6 Evaluation of amino sugar quantification via IC-O-IRMS........................81 2.2.2.7 Evaluation of δ 13 C determination via IC-O-IRMS....................................82 2.2.3 Results and Discussion .................................................................................83 2.2.3.1 Chromatography....................................................................................83 2.2.3.2 Recovery, linearity, precision and detection and quantification limits .....84 2.2.3.3 3.3 Amount dependence and correction factors of δ 13 C values..............85 2.2.3.4 3.4 Accuracy, precision and isotopic LoQ of δ 13 C determination............86 2.2.3.5 Advantages of IC-O-IRMS.....................................................................89 2.2.4 Conclusions...................................................................................................89 Acknowledgments.....................................................................................................90 Reference List...........................................................................................................91 Supplementary Data.................................................................................................93 2.3 Study 3: Biochemical pathways of amino acids in soil: Evaluation by position-specific labeling and 13 C-PLFA analysis.........................................96 Abstract ....................................................................................................................97 2.3.1 Introduction ...................................................................................................98 2.3.2 Material and Methods..................................................................................100 2.3.2.1 Field experiment..................................................................................100 2.3.2.2 Analytical methods...............................................................................101 2.3.2.3 Divergence Index.................................................................................104 2.3.2.4 Statistical analysis ...............................................................................105 Contents IV 2.3.3 Results........................................................................................................105 2.3.3.1 Incorporation of uniformly labeled amino acids....................................105 2.3.3.2 Incorporation of position-specifically labeled amino acids....................106 2.3.3.3 Divergence Index.................................................................................108 2.3.4 Discussion...................................................................................................110 2.3.4.1 Incorporation of carbon from amino acids in soil and microbial biomass 110 2.3.4.2 Incorporation of tracer into the microbial groups ..................................111 2.3.4.3 Discrimination of individual carbon positions by microbial utilization differs depending on oxidation state, amino acid and time..............................113 2.3.5 Conclusions.................................................................................................115 Acknowledgements.................................................................................................116 Reference List.........................................................................................................117 Supplementary Data...............................................................................................120 2.4 Study 4: Biogeochemical transformations of amino acids in soil assessed by position-specific labeling ........................................................................122 Abstract ..................................................................................................................123 2.4.1 Introduction .................................................................................................124 2.4.2 Material and Methods..................................................................................126 2.4.2.1 Soil ......................................................................................................126 2.4.2.2 Chemicals and radiochemicals ............................................................126 2.4.2.3 Experimental setup..............................................................................126 2.4.2.4 Radiochemical analyses......................................................................129 2.4.2.5 Calculation of the kinetics of alanine utilization ....................................129 2.4.2.6 Calculation of the distribution of alanine-C in transformation products.130 2.4.2.7 Statistics..............................................................................................131 2.4.3 Results........................................................................................................132 2.4.3.1 Evaluation of results quality .................................................................132 2.4.3.2 Sorption of alanine to the soil matrix....................................................133 2.4.3.3 Kinetics of biotic alanine utilization.......................................................134 2.4.3.4 Biotic transformation products of alanine.............................................136 2.4.3.5 Position-specific differences of the alanine transformation pathways...137 2.4.4 Discussion...................................................................................................138 2.4.4.1 Sorption of alanine occurs as a whole molecule ..................................138 2.4.4.2 Kinetics of extracellular transformation and microbial uptake...............139 2.4.4.3 Exoenzymatic transformation products................................................140 Contents V 2.4.4.4 Metabolic pathways and their intracellular transformation products .....141 2.4.5 Conclusions and Outlook.............................................................................143 Acknowledgement ..................................................................................................144 Reference List.........................................................................................................145 Supplementary Data...............................................................................................148 2.5 Study 5: Sorption affects amino acid pathways in soil: Implication from position-specific labeling of alanine............................................................151 Abstract ..................................................................................................................152 2.5.1 Introduction .................................................................................................154 2.5.2 Material and Methods..................................................................................156 2.5.2.1 Soil ......................................................................................................156 2.5.2.2 Sorbents..............................................................................................156 2.5.2.3 Chemicals and radiochemicals ............................................................157 2.5.2.4 Pre-experiments ..................................................................................157 2.5.2.5 Experimental Setup .............................................................................158 2.5.2.6 Chemical and radiochemical analyses.................................................159 2.5.2.7 Calculations and modeling...................................................................159 2.5.2.8 Calculation of the C-1/C-2,3-ratio and the Divergence Index DI i ..........160 2.5.2.9 Statistics..............................................................................................161 2.5.3 Results........................................................................................................162 2.5.3.1 Sorption and microbial utilization of uniformly labeled alanine .............162 2.5.3.2 Kinetics of position-specific utilization of sorbed alanine C...................164 2.5.3.3 Incorporation of C from alanine positions in stabilized pools and decomposition to CO 2 ..........................................................................167 2.5.4 Discussion...................................................................................................169 2.5.4.1 Sorption mechanisms of amino acids ..................................................169 2.5.4.2 Bioavailability of sorbed alanine...........................................................172 2.5.4.3 Pathways of microbial metabolization of sorbed alanine......................173 2.5.4.4 Stabilization of amino acid C by sorption .............................................176 2.5.5 Conclusions and Outlook.............................................................................177 Acknowledgments...................................................................................................178 Reference List.........................................................................................................179 Supplementary Data...............................................................................................183 2.6 Study 6: Biochemistry of hexose and pentose transformations in soil analyzed by position-specific labeling and 13 C-PLFA.................................184 Abstract ..................................................................................................................185 Contents VI 2.6.1 Introduction .................................................................................................186 2.6.2 Material and Methods..................................................................................188 2.6.2.1 Sampling Site ......................................................................................188 2.6.2.2 Analytical methods...............................................................................189 2.6.2.3 Divergence Index.................................................................................192 2.6.2.4 Statistical analysis ...............................................................................192 2.6.3 Results........................................................................................................193 2.6.3.1 Incorporation of uniformly labeled monosaccharides ...........................193 2.6.3.2 Incorporation of position-specifically labeled monosaccharides ...........193 2.6.3.3 Tracer uptake of functional microbial groups........................................195 2.6.3.4 Divergence Index.................................................................................196 2.6.4 Discussion...................................................................................................196 2.6.4.1 Glucose and Ribose incorporation into soil and microbial biomass......196 2.6.4.2 Microbial utilization of individual positions of glucose and ribose molecules ............................................................................................198 2.6.4.3 Specific pathways of glucose and ribose utilization by individual microbial groups .................................................................................................199 2.6.4.4 Metabolic tracing by position-specific labeling of monosaccharides.....202 2.6.5 Conclusions and Outlook.............................................................................203 Acknowledgements.................................................................................................204 References .............................................................................................................205 Supplementary Data...............................................................................................208 2.7 Study 7: Metabolic pathways of fungal and bacterial amino sugar formation in soil assessed by position-specific 13 C-labeling ......................................210 Abstract ..................................................................................................................211 2.7.1 Introduction .................................................................................................213 2.7.2 Material and Methods..................................................................................215 2.7.2.1 Experimental Site ................................................................................215 2.7.2.2 Experiment Design ..............................................................................215 2.7.2.3 Sampling and Sample Preparation ......................................................216 2.7.2.4 Bulk Soil and Microbial Biomass Analysis............................................216 2.7.2.5 Amino sugar δ 13 C analysis...................................................................217 2.7.2.6 Divergence Index.................................................................................219 2.7.2.7 Statistics..............................................................................................220 2.7.3 Results........................................................................................................220 2.7.3.1 Glucose 13 C incorporation into soil and microbial C pools....................220 Contents VII 2.7.3.2 Incorporation of C from various positions of glucose molecule into individual amino sugars .......................................................................221 2.7.3.3 Replacement of cell wall pool by glucose 13 C.......................................223 2.7.4 Discussion...................................................................................................224 2.7.4.1 Fungal versus bacterial contribution to the amino sugar fingerprint and glucose utilization ................................................................................224 2.7.4.2 Pathways of amino sugar formation.....................................................225 2.7.4.3 Specific pathways of fungi and bacteria...............................................228 2.7.5 Conclusions and Outlook.............................................................................229 Acknowledgements.................................................................................................230 References .............................................................................................................231 Supplementary Data...............................................................................................235 2.8 Study 8: Formation and transformation of fatty acids in soil assessed by position-specific labeling of precursors......................................................236 Abstract ..................................................................................................................237 2.8.1 Introduction .................................................................................................239 2.8.2 Material and Methods..................................................................................241 2.8.2.1 Experimental Site ................................................................................241 2.8.2.2 Experiment Design ..............................................................................242 2.8.2.3 Sampling and Sample Preparation ......................................................242 2.8.2.4 Bulk Soil and Microbial Biomass Analysis............................................242 2.8.2.5 PLFA δ 13 C analysis..............................................................................243 2.8.2.6 Fatty acid grouping..............................................................................245 2.8.2.7 The Divergence Index DI i .....................................................................245 2.8.2.8 Statistics..............................................................................................246 2.8.3 Results........................................................................................................246 2.8.3.1 Incorporation of 13 C in soil and microbial biomass................................246 2.8.3.2 Incorporation of C from various positions of acetate and palmitate into individual PLFAs..................................................................................247 2.8.3.3 Incorporation of acetate and palmitate 13 C into PLFAs of individual microbial groups ..................................................................................249 2.8.4 Discussion...................................................................................................252 2.8.4.1 Utilization and turnover of acetate and palmitate by soil microbial community...........................................................................................252 2.8.4.2 Pathways of fatty acid formation from acetate in soil............................253 2.8.4.3 Pathways of fatty acid transformations in soils.....................................254 List of Figures XIV Fig. 2 Percentage of 14 C incorporation in roots and shoots after uniform 14 C labeling with acetate and alanine. Letters indicate the significant differences (p<0.001) of acetate and alanine C between plants ...............................................................278 Fig. 3 Percentage of 14 C incorporation in roots and shoots after position-specific labeling with alanine. The alanine positions were C-1 (carboxyl group), C-2 (amino-bound group) and C-3 (methyl group). Letters indicate significant differences (p<0.001) between alanine C positions. .............................................................................279 Fig. 4 Ratio of 14 C/ 15 N for individual alanine C positions incorporated in plant biomass. The alanine positions were C-1 (carboxyl group), C-2 (amino-bound group) and C3 (methyl group). Letters indicate significant differences (p<0.001) between alanine C positions.............................................................................................280 Fig. 5 Illustration of the fate of alanine tracer molecules, which are either taken up intact or degraded/mineralized to fragments and subsequently incorporated into plant biomass or microorganisms. Microbial metabolism of alanine by microorganisms is adapted from Dippold & Kuzyakov (2013)..........................................................282 List of Tables XV III. List of Tables Extended Summary Table S1 Treatments of position-specific 13 C and 15 N labeling. Applied amount of 13 C and 15 N, their isotopic enrichment, as well as the respective compound and labeled position are presented. Nat. abund. means application of non-enriched substances, x means“no application of 15 N in this treatment. ...............................10 Table S2 Applied 13 Cand 14 C-labeling approaches as well as analytical methods for the individual studies; First line shows whether samples were derived from field or laboratory experiments; PS indicates position-specific labeling............................12 Table S3 Title of the individual studies as well as their objectives and main conclusions. ...15 Table S4 Analogies in the behavior of individual C positions of LMWOS entering the main branch of the basic C metabolism (glycolysis, pyruvate dehydrogenase and citric acid cycle). Analogies were concluded from the basic metabolic pathways shown in Figure S6. Positions within one column are equivalent within these pathways. 21 Publications and Manuscripts: Study 1: Table 1 Absolute and relative abundance (absolute in µg per g and relative in % of total PLFAs) of the fatty acids of the microbial groups, classified by factor analysis (factor loadings see Supplementary Table 2).......................................................54 Study 2: Table 1 Recovery (%), relative standard deviation (RSD) and parameters of regression analysis as well as detection (LoD) and quantification limits (LoQ) for the quantification of amino sugars assessed from the standard addition experiment. 85 Table 2 Comparison of δ 13 C(Std) EA-IRMS (EA-IRMS PeeDeeBe calibrated δ 13 C value of standard substances spiked to the sample) and δ 13 C(Std) IC-O-IRMS (fitted δ 13 C value of the spiked standards from the mixing model of the standard addition method) reflecting the accuracy of IC-O-IRMS measurement. Fitted δ 13 C values for soil from mixing model (δ 13 C(soil) calculated ) and real measurement of un-spiked soil δ 13 C(soil) IC-O-IRMS are also presented. Precision is shown 1) by the standard deviation of the measurement repetitions and 2) by calculating the area dependent standard deviation according to equation 6 for the measured peak area. Isotopic LoQ reflects the minimum amount per vial needed to receive a standard error of the measurement repetitions lower than 0.5‰. Gal=galactosamin, Glc=glucosamine and MurA=muramic acid..........................................................87 Study 3: Table 1 Concentrations of amino acid solutions for soil labeling .....................................101 Table 2 Total C content and 13 C incorporation of uniformly labeled amino acids into soil, microbial biomass and sum of PLFA (Σ-PLFA)...................................................105 Study 4: Table 1 Parameters of the Michaelis-Menten kinetics for treatments with inhibition of respiration (eq. 2) and treatments without inhibition (eq. 6). R 2 is the coefficient of determination and stars show significance of the respective non-linear fitting result (respectives curves are plotted in Figure 4)........................................................136 List of Tables XVI Study 5: Table 1 Effective cation exchange capacity and specific surface area of the five sorbents and the soil used for this experiment..................................................................157 Table 2 Initially sorbed alanine C and fitted parameters to the four-pool model for microbial utilization of sorbed alanine (Fig. 1) fitted to the data of uniform alanine labeling ..........................................................................................................................162 Table 3 Fitted parameters of the four-pool model for microbial utilization of sorbed alanine (Fig. 1) for the individual alanine C positions......................................................165 Study 6: Table 1 Locations of 13 C in position specifically labeled glucose and ribose and their amounts added to soil in the field experiment.....................................................188 Table 2 Total C content and 13 C incorporation of uniformly labeled monosaccharides into soil, microbial biomass and sum of PLFA (Σ-PLFA). ..........................................193 Study 7: Table 1 Amount and glucose 13 C recovery in total organic C (TOC), microbial biomass C (C mic ) and the total amino sugars (Σ AminoSugars ) as well as the three individual amino sugars................................................................................................................221 Table 2 Theoretic C pattern of newly formed amino sugars after simple pathway combinations of basic C metabolism..................................................................228 Study 8: Table 1 Total organic C (TOC), microbial biomass C (C mic ) and the sum of all measured PLFAs (list of fatty acids see Supplementary, Table A1) in mg C per g soil (dry weight)...............................................................................................................246 Study 9: Table 1 The physicochemical properties of the Ap-horizon of the haplic Luvisol. ............273 Table 2 Shoot/root ratio of 15 N from individual N sources................................................278 Table 3 Intact uptake of alanine by chicory, lupine and maize and estimated contribution of intact alanine uptake to total N nutrition of these plants with respect to the other N sources. .............................................................................................................281 Abbreviations XVII IV. Abbreviations ANCOVA Analysis of Covariance ANOVA Analysis of Variance C Carbon CEC Cation Exchange Capacity DI Divergence Index DOC Dissolved Organic Carbon DON Dissolve Organic Nitrogen EA Elemental Analyzer FAME Fatty Acid Methyl Ester GC Gas Chromatography GC-C-IRMS Gas Chromatography-Combustion-Isotope Ratio Mass Spectrometer HPLC High Pressure Liquid Chromatography IC Ion Chromatography IC-O-IRMS Ion Chromatograph-Oxidation-Isotope Ratio Mass Spectrometer IRMS Isotope Ratio Mass Spectrometer IS Internal Standard LMWOS Low Molecular Weight Organic Substances MS Mass Spectrometer N Nitrogen PEEK Polyetheretherketon PLFA Phospholipid Fatty Acid SIM Selected Ion Mode SOC Soil Organic Carbon SOM Soil Organic Matter Summary XVIII V. Summary Transformation of low molecular weight organic substances (LMWOS) is one of the most important steps in biogeochemical cycles since all high molecular substances pass this stage during their decomposition. Microbial utilization is the most relevant sink for LMWOS in soils and thus knowledge about microbial transformations of LMWOS is crucial for understanding the soil organic carbon (SOC) cycle and predicting its reaction to changes in controlling environmental parameters. Previous studies focused on determining fluxes through the LMWOS pool, but they rarely identified transformation steps. This thesis aims to establish position-specific isotope labeling as a tool in soil science to trace the pathways of LMWOS transformations. In a medium-term field experiment six position-specific 13 C-labeled LMWOS from the three main LMWOS classes were applied: two amino acids (alanine and glutamate), two monosaccharides (glucose and ribose) and two organic acids (acetate and palmitate). 13 C remaining in soil and that incorporated into microbial biomass and specific microbial cellular compounds (phospholipid fatty acids (PLFA) and amino sugars) was determined by bulk and compound-specific 13 C analyses. Therefore, a new instrument coupling, an ion chromatograph with an isotope ratio mass spectrometer (IC-O-IRMS), and the respective methods for amino sugar analysis were established. The effect of altered environmental conditions and the relevance of further LMWOS sinks (sorption or plant uptake) were evaluated in several additional laboratory experiments based on position-specific 14 C-labeling. The divergence index (DI) was established to compare the position-specific fate of individual substances in various studies independent of the isotopic approach or experimental design used or the pool investigated. Microbial utilization was the fastest process in the removal of LMWOS from soil solution and neither plant uptake nor sorption could out-compete microorganisms. The incorporation of individual molecule positions in soils, microbial biomass and distinct compound classes was clearly defined by the microbial metabolism: Glycolysis, oxidation by pyruvate dehydrogenase and the citric acid cycle were identified as the main metabolic processes. However, in addition to these oxidizing catabolic pathways, the anabolic pathways, i.e. building-up new cellular compounds, occurred in soils simultaneously. This involved an intensive C recycling and turnover within the microorganisms that was observed not only for cytosolic compounds but also for cell wall polymers. Intensive modifications and transformation within metabolic side branches, like the fatty acid formation and transformation pathways, were identified. These results for fatty acid transformations are crucial for their application as plant biomarkers in studies on palaeoenvironmental reconstruction. The combination of position-specific 13 C-labeling with compound-specific isotope analysis of microbial biomarkers allowed the further identification of specific pathways of in- Summary XIX dividual functional microbial groups in soils. Fungal metabolism was shown to be slower than bacterial intracellular C recycling and turnover, which provides the metabolic reason for the slow-cycling fungal-based and fast-cycling bacteria-based branch of the soil food web. Shifts in C allocation through various metabolic pathways were dependent on environmental factors: a gradient of C metabolism from starvation pathways via maintenance metabolism to metabolic pathways characteristic for microbial growth was observed with increasing substrate concentration. Sorption, also limiting the bioavailability of a substrate, caused similar shifts in metabolic pathways: the lower the bioavailability (e.g. due to sorption), the more C was allocated towards anabolic biosynthesis, i.e. into microbial products. Thus, these studies revealed that position-specific labeling is not only a valuable tool in biochemistry for metabolic flux analysis, but also enables the reconstruction of metabolic pathways of LMWOS within diverse microbial communities in complex media such as soil. Processes occurring simultaneously in soil i.e. 1) within individual, reversible metabolic pathways, 2) in various microbial groups or 3) in specific microhabitats (like on mineral surfaces, at the soil-plant interface or at hot-spots versus bulk soil) could be traced by position-specific labeling in soils in situ. The main metabolic pathways of microbial LMWOS transformation by cataand anabolism were traced by position-specific labeling. These pathways and their regulating factors are crucial for assessing C flows towards mineralization versus the formation of microbial biomass, the prerequisite for the formation of microbially-derived SOC. This molecular knowledge of transformation steps and their regulating factors is crucial to predict (i.e. by new process-based modelling approaches) and manipulate C allocation and stabilization in soils. Zusammenfassung XX VI. Zusammenfassung Die Transformation niedermolekularer organischer Substanzen (LMWOS) ist der zentrale Schritt in biogeochemischen Kreisläufen, da alle hochmolekularen Substanzen während ihres Abbaus den LMWOS Pool passieren. Mikroorganismen stellen die bedeutendste Senke für LMWOS dar, weshalb mikrobielle Transformationen von LMWOS essentiell für den Kohlenstoffkreislauf im Boden sind. Bisherige Studien quantifizierten meist Flüsse durch den LMWOS Pool, arbeiteten aber kaum an der Aufklärung der Transformationsprozesse. Im Rahmen dieser Dissertation soll die positionsspezifische Isotopenmarkierung als neue bodenkundliche Methode zur Aufklärung von LMWOS-Transformationswegen etabliert werden. In einem Feldexperiment wurden sechs positionsspezifisch 13 C markierte LMWOS der drei wichtigsten Substanzklassen appliziert: zwei Aminosäuren (Alanin und Glutamat), zwei Monosaccharide (Glucose und Ribose) und zwei organische Säuren (Acetat und Palmitat). Die Analyse von verbleibendem 13 C im Boden, 13 C in der mikrobiellen Biomasse und in spezifischen Zellbausteinen (Phospholipidfettsäuren (PLFA) und Aminozucker) erfolgte durch gesamtund komponentenspezifische 13 C Methoden. Hierfür wurde eine neue Instrumentenkopplung – ein Ionenchromatograph mit einem Isotopenmassenspektrometer (IC-O-IRMS) – etabliert und die darauf abgestimmte Aminozucker-Aufreiningungsmethode eingearbeitet. Der Effekt sich ändernder Umweltfaktoren sowie die Relevanz weiterer LMWOS-Senken (Sorption und Pflanzenaufnahme) wurden anhand mehrerer zusätzlicher Laborexperimente mit positionsspezifischer 14 C Markierung evaluiert. Die Einführung des Divergenz Index (DI) ermöglichte es den positionsspezifischen Einbau in verschiedenen Studien unabhängig vom applizierten Isotop, dem experimentellen Design und dem untersuchten Pool zu vergleichen. Mikroorganismen waren die dominante Senke für LMWOS und weder Pflanzenaufnahme noch Sorption konnten in Rate und Kinetik mit mikrobiellen Aufnahmesystemen konkurrieren. Der Einbau einzelner Molekülpositionen in Boden, mikrobielle Biomasse und bestimmte Substanzklassen war durch den mikrobiellen Metabolismus bestimmt, v.a. durch Glykolyse, Oxidation durch Pyruvat-Dehydrogenase und Citratzyklus. Allerdings liefen parallel zu diesen oxidierenden, katabolen Stoffwechselwegen auch anabole Reaktionen, d. h. der Aufbau neuer Zellkomponenten, ab. Dies führte zu einem starken C-Umsatz und Recycling, nicht nur im Cytosol sondern z.B. auch von Zellwandpolymeren. Intensive Umsätze innerhalb metaboler Seitenäste, wie der Fettsäurebiosynthese, wurden identifiziert. Diese Ergebnisse zur Fettsäuretransformation sind wesentlich für die Anwendung von Fettsäuren als pflanzliche Biomarker in Paläoumweltstudien. Die Kombination positionsspezifischer 13 C Markierung mit komponentenspezifischer Isotopenanalytik mikrobieller Biomarker erlaubte des Weiteren die Identifikation spezifischer Stoffwechselwege einzelner mikrobieller Gruppen. Pilze zeigten einen langsameren intrazel- Zusammenfassung XXI lulären C-Umsatz als Bakterien, was die metabole Grundlage für den langsam-zyklierenden, pilzbasierten und den schnell-zyklierenden, bakterienbasierten Zweig des Bodennahrungsnetzes liefert. Die Verschiebungen der Kohlenstoffflüsse durch verschiedene Stoffwechselwege wurden in Abhängigkeit von Umweltfaktoren identifiziert: Mit Zunahme der Substratkonzentration konnte ein Gradient von C-Mangel-Stoffwechselwegen über den Erhaltungsmetabolismus hin zu charakteristischen Wachstums-Stoffwechselwegen beobachtet werden. Eine Verringerung der Substratverfügbarkeit durch Sorption verursachte eine ähnliche Verschiebung der metabolen C-Flüsse: Je niedriger die Verfügbarkeit, desto mehr C wird in Biosynthesewege also mikrobielle Produkte, verlagert. Diese Studien konnten zeigen, dass positionsspezifische Markierung nicht nur eine wertvolle Methode in der Biochemie darstellt, sondern auch die Aufklärung der Verstoffwechslung von LMWOS durch diverse mikrobielle Gemeinschaften in komplexen Medien wie dem Boden ermöglicht. Parallel ablaufende Prozesse in Böden wie z. B. 1) der Rückfluss durch reversible Stoffwechselwege, 2) Umsätze in verschiedenen mikrobiellen Gruppen oder 3) Umsätze in spezifischen Mikrohabitaten (an Mineraloberflächen, am Boden-PflanzeInterface oder an Hot-spots versus dem Gesamtboden) können mittels positionspezifischer Markierung im Boden in situ verfolgt werden. Der Umsatz von LMWOS in Kataund Anabolismus wurde im Rahmen dieser Dissertation rekonstruiert. Das Verständnis für diese Stoffwechselwege und ihre Regulationsfaktoren ist entscheidend für die Beurteilung von C-Flüssen zwischen Mineralisation und dem Aufbau mikrobieller Biomasse – der Voraussetzung zur Bildung mikrobieller, organischer Bodensubstanz. Das Wissen über Transformationsschritte und ihre regulierenden Faktoren ist essentiell für die Vorhersage (z. B. mittels prozessbasierter Modellierung), aber auch für die Manipulation der C-Sequestrierung und Stabilisierung in Böden. Zusammenfassung XXII Extended Summary 1 1 Extended Summary 1.1 Introduction 1.1.1 Low molecular weight organic substances in soils 1.1.1.1 Role and relevance of low molecular weight organic substances in soil Soil organic carbon (SOC) is the largest terrestrial carbon (C) pool, with C stocks of around 1500 Pg (Batjes, 1996). On average 30-120 kg C m -2 is stored up to 1 m soil depth and 0.5 to 1% of this stock is annually respired (van Hees et al., 2005a) but the largest portion is stable or inactive. Traditionally, SOC has been divided operationally by chemical fractionation into structurally diverse fulvic and humic acids (van Hees et al., 2005a). More recent results revealed that a limited range of defined substance classes and their polymers build up the soil organic matter (SOM) (Schmidt et al., 2011; von Luetzow et al., 2006). A minor portion of SOC constitutes dissolved organic matter (DOC), in most cases less than 2 mol C m -2 (van Hees et al., 2005a). Only up to 10% of DOC consists of identifiable compounds of low molecular weight. These low molecular weight organic substances (LMWOS) are defined as soluble substances with a molecular weight lower than 250 Da (Boddy et al., 2007) and mainly consist of aliphatic and aromatic carboxylic acids, amino acids and peptides, mono-, diand small oligosaccharides, amino sugars, phenolic substances and siderophores (McKeague et al., 1986). Although the portion of LMWOS in SOC is extremely low, they play a major role in ecosystem functions. Regarding the C cycle, the importance of LMWOS is not determined by their pool size (Fischer et al., 2007), but by their huge fluxes (>20 mol C m -2 y -1 ) that pass through this pool. During decomposition of plant-derived organic matter the high molecular weight organic substances are degraded by exoenzymes into low molecular monomers and pass the pool of LMWOS. They can then be oxidized to CO 2 by microbial respiration. Van Hees et al (2005a) summarized for forest soils that although LMWOS comprise less than 0.05% of the C pool, they contribute to more than 10-20% to the soil respiratory fluxes, thus demonstrating the high relevance of this active, fast cycling C pool for the SOC turnover. In addition to their function as energy and C source for microorganisms, they fulfill several important functions in soils: 1) Contribution to weathering and solubilization of nutrients for plants; 2) formation of soil structures like aggregates; 3) acceleration of re- Extended Summary 8 1.1.4 Objectives The main objective of this thesis was to establish position-specific 13 Cand 14 Clabeling and metabolic tracing as a tool in soil science, which enables a processorientated view on the LMWOS cycle and can be applied for field and laboratory experiments. More specifically, the following objectives were aimed towards: 1) Determination of metabolic pathways of two representatives of the three main substance classes of LMWOS (amino acids, monosaccharides and organic acids) by position-specific labeling 2) Coupling of position-specific labeling with compound-specific isotope analysis of microbial biomarkers a. to follow the C incorporation into various cellular compound classes b. to identify specific metabolic pathways of individual members of the soil microbial community 3) Identification of specific metabolic pathways depending on certain environmental conditions: a. pathways of LMWOS sorbed on various soil components b. pathways under various concentrations of LMWOS c. extraversus intracellular transformation pathways 4) Assessment of kinetics and ecological relevance of competing sinks for LMWOS: a. sorption versus microbial utilization b. plant uptake versus microbial utilization 1.2 Experiments and Methods 1.2.1 Field experiment The field experiment was carried out on an agricultural field site close to Hohenpoelz (49°54' Northern latitude; 11°08' Eastern longitude, 500 m a.s.l.) in northern Bavaria. Mean annual temperature is +7 °C, mean precipitation is 870 mm and soil type is a haplic Luvisol (IUSS Working group WRB, 2007). 784 columns (Figure S2) were installed according to a randomized block design, where the four blocks represent the four replications of each treatment. Extended Summary 9 Fig. S2 Schematic (left) and photo (middle) of a labeling column of the field experiment. Right photo shows one of the four blocks, each with 196 columns. In total, six individual LMWOS were applied: two amino acids (alanine and glutamate), two monosaccharides (glucose and ribose) and two organic acids (applied as acid anion: acetate and palmitate). Each of their purchasable isotopomers as well as their uniformly 13 C-labeled form was applied in separate columns (see Table S1). In addition, 15 N-labeled glutamate and alanine were added as well as labeled mineral nitrogen ( 15 NO 3 - and 15 NH 4+ ). Consequently, this is the first experiment which allows metabolic tracing by two parallel approaches: 1) metabolic tracing based on individual C positions (Dijkstra et al., 2011a) and 2) metabolic tracing based on multiple isotope labeling (Knowles et al., 2010) (results of the second approach were not included in this thesis). Each of the six substances had one background treatment where the same amount of non 13 C-enriched substance was applied. In each of the 22 labeled and 6 background treatments, the amount of C and N applied was as low as possible and identical for all columns to avoid potential disturbance of the microbial community. Each treatment was designed for seven sampling dates and in four replications (resulting in 784 soil columns). To prevent rainfall and thus leaching, a roof was installed above the field site during labeling and for the first 10 days. Excluding additional rainfall water due to the roof and assuming no preferential flow in the freshly tilled soil, leaching can be regarded as negligible and mineralization is the only process removing 13 C from the soil. Only data from the first two sampling dates, reflecting the short-term dynamics of LMWOS, are presented in this thesis. Sampling was performed by removing the entire column from the field. Length (i.e. volume) of soil column, fresh weight of soil and water content were determined. After homogenization, each sample was split and the subsamples were prepared and stored according to the different analyses. 10 cm 13 cm 10 cm 10 cm 13 cm 10 cm Extended Summary 10 Table S1 Treatments of position-specific 13 C and 15 N labeling. Applied amount of 13 C and 15 N, their isotopic enrichment, as well as the respective compound and labeled position are presented. Nat. abund. means application of non-enriched substances, x means“no application of 15 N in this treatment. Position-specific 13 C labeling 15 N labeling substance class substance position(s) applied amount of 13 C (µmol per column) substance applied amount of 15 N (µmol per column) C-1 94.2 x C-2 93.0 x C-3 94.7 alanine (nat. abund.) x alanine uniformly 96.3 15 N alanine (98 at%) 28.7 C-1 93.0 x C-2 56.3 glutamate (nat. abund.) x amino acids glutamate uniformly 91.6 15 N glutamate (98 at%) 28.0 C-1 91.5 x C-2 93.4 x C-4 93.5 x C-6 91.8 (NH 4 ) 2 SO 4 (nat. abund.) x glucose uniformly 93.4 15 N (NH 4 ) 2 SO 4 (98 at%) 27.8 C-1 93.3 x C-5 93.0 x monosacharides ribose uniformly 91.8 (NH 4 ) 2 SO 4 (nat. abund.) x C-1 94.6 x C-2 94.1 x acetate uniformly 95.8 KNO 3 (nat. abund.) x C-1 47.5 x C-2 47.1 x C-16 44.3 KNO 3 (nat. abund.) x organic acids palmitate uniformly 49.5 15 N KNO 3 (98 at%) 28.6 1.2.2 Laboratory experiments Besides the field experiment, several laboratory experiments with modifications of environmental conditions were performed. In general, in laboratory experiments positionspecific 14 C-labeled tracers were applied. 1.2.2.1 Experiment 1: Transformations of free alanine The first experiment aimed at identifying the transformation pathways of amino acids depending on two factors: 1) the concentration of alanine (0.5, 5, 50, 500 and 5000 µM), and 2) the extraand intracellular as well as abiotic processes (i.e. sorption) of alanine removal from soil solution, separated by selective inhibition. Alanine transformation products were operationally separated by sequential extraction into ionexchangeable and ligand-exchangeable transformation products, whereas irreversiblebound transformation products remained in the soil. Extended Summary 11 1.2.2.2 Experiment 2: Transformations of sorbed alanine In the sorption experiment, availability and microbial utilization pathways of absorbed alanine were investigated in two steps: 1) Labeled alanine was adsorbed to five sterilized sorbents commonly present in soils: two iron oxides with different crystalline structure: goethite and hematite; two clay minerals with 2:1 layers – smectite, and 1:1 layers – kaolinite; and active coal. 2) Thereafter, the sorbed alanine was mixed with the soil and incubated for 3 days. The effect of sorption on microbial utilization, especially their metabolic pathways, was elucidated. 1.2.2.3 Experiment 3: Plant uptake of intact alanine Plant uptake was carried out with dual-isotope, position-specific labeled alanine in rhizotubes (Biernath et al., 2008; Rasmussen et al., 2010). Uptake of C and N from individual positions by Zea mays, Lupinus albus and Cichorium intybus was traced. As a control, 14 C acetate and mineral 15 NH 4+ and 15 NO 3were applied. Thus, passive uptake of LMWOS as well as the relevance of N-LMWOS uptake compared to mineral nitrogen was assessed. Position-specific labeling enabled the differentiation of intact alanine uptake versus the uptake of its transformation fragments. 1.2.3 Methods to trace 13 C and 14 C in transformation products of LMWOS An overview over the methods, applied in this thesis, to trace 13 C, 15 N and 14 C in specific transformation products and more unspecific SOC fractions is presented in Table S2. Methods are briefly described in the following chapters and in detail in the Material and Methods section of the respective studies. Extended Summary 12 Table S2 Applied 13 Cand 14 C-labeling approaches as well as analytical methods for the individual studies; First line shows whether samples were derived from field or laboratory experiments; PS indicates position-specific labeling. Field experiment Laboratory experiment Study Study 1 Study 2 Study 3 Study 6 + 7 Study 8 Study 4 Study 5 Study 9 Topic LMWOS comparison method development amino acid monosaccharides organic acids free amino acids sorbed amino acids plant uptake of amino acids Labeling Uniformly 13 C PS 13 C PS 13 C PS 13 C PS 14 C PS 14 C PS 14 C & 15 N 13 C in bulk soil X X X X 15 N in bulk soil X 15 N in plant biomass X microbial biomass 13 C X X X X 13 C in PLFA X X X X 13 C in amino sugars X X 14 C in soil solution X X 14 C in bulk soil X X X 14 C in soil extracts X 14 C in CO 2 X 1.2.3.1 Bulk-isotope measurements by EA-IRMS 13 C and 15 N remaining in soil was quantified by determination of δ 13 C and δ 15 N values of bulk soil samples. Incorporation of 13 C in microbial biomass was calculated from the δ 13 C value of fumigated and unfumigated soil extracts, gained by the chloroformfumigation-extraction method (Brookes et al., 1985; Wu et al., 1990). For all δ 13 C measurements, the samples were freeze-dried and measured by Elemental Analyzer-Isotope Ratio Mass Spectrometer (EA-IRMS). 13 C incorporation was calculated according to the mixing model referenced on the respective δ 13 C values of the background treatment. 1.2.3.2 Compound-specific isotope analysis of microbial biomarkers 13 C incorporation into microbial membrane lipids, the phospholipids fatty acids (PLFA), was determined by compound-specific isotope analysis. For this, the Bligh-andDyer extract of intact polar lipids was performed. PLFA were purified by liquid-liquid extraction and column chromatography and derivatized to their fatty acid methyl esters (FAMEs). δ 13 C value of FAMEs was determined by gas chromatography-combustion- Extended Summary 13 isotope ratio mass spectrometry (GC-C-IRMS) and 13 C incorporation was calculated by mixing models with the respective background treatments as a reference. In addition, 13 C incorporation into microbial cell wall monomers, the amino sugars, was measured in this thesis. However, GC-C-IRMS methods are not sufficiently reliable and existing liquid chromatography-oxidation-isotope ratio mass spectrometry methods were not able to determine the δ 13 C value of bacterial muramic acid. Therefore, a new instrument coupling, using an ion chromatograph (IC) instead of a classical liquid chromatograph for IC-O-IRMS measurements and amino sugar δ 13 C determination was established in study 2 (see Figure S3). Thus, previous purification methods (Bode et al., 2009; Glaser and Gross, 2005; Indorf et al., in press) had to be optimized and an IC-OIRMS measurement had to be elaborated (see study 2). Briefly, after acid hydrolysis, iron and salts were removed by precipitation (Zhang and Amelung, 1996) and neutral compounds were separated from amino sugars by a cation exchange resin (Indorf et al., 2013). Liquid chromatography was optimized (Bode et al., 2009) and 13 C incorporation into microbial cell walls was calculated using the analogy of incorporation into PLFA. Fig. S3 Overview of the instrument coupling: Ion Chromatograph is shown on the left side with the pump, autosampler and detector-chromatography compartment. Connection to isolink occurs via a PEEK capillary with interposed colloid filter. Scheme of LC Isolink is adapted from Krummen et al. (2004). 1.2.3.3 Radiochemical analyses Incorporation of 14 C into the transformation products was performed after various sampling and extraction methods based on scintillation counting. For this purpose, solid samples were combusted and 14 CO 2 was trapped in NaOH. Bound 14 C was determined by scintillation counting. 14 C in soil extracts or suspensions could be directly measured Extended Summary 14 after mixing a subsample with a scintillation cocktail. Direct measurement of decomposed 14 CO 2 was performed on experiments in well plates. Therefore, a CO 2 trap based on 24segment filter paper was constructed above the 24-well plate. Construction of this trap was optimized and efficiency evaluated in study 5. 1.2.4 The Divergence Index It remains challenging to compare the position-specific fate of individual LMWOS, especially if C is transformed into strongly differing C pools. To enable the comparison of the individual studies independent of the isotopic approach ( 13 C or 14 C) or experimental design used or the pool investigated, the Divergence Index DI i was introduced in this thesis (equation 1). [ ] [ ] ∑ = = ⋅ = ni 1i i i i C Cn DI (1) This index shows the fate of individual C atoms from the position i within a transformation process relative to the mean transformation of the n total number of C atoms in the substance. Thus, a DI i of 1 means that the transformation of this C atom in the investigated pool corresponds to that of uniformly labeled substance (average of all C atoms of the substance). The DI i ranges from 0 to n, and values between 0 and 1 reflect reduced incorporation of the C into the investigated pool, whereas values between 1 and n show increased incorporation of the C atom into this pool as compared to the average. This index is not dependent on absolute amounts or proportions of the substance used in individual processes. Therefore, it enables comparison of the distribution of individual C atoms over the whole range of investigated concentrations, the size of C pools, the process rates, etc. Extended Summary 15 1.3 Results and Discussion 1.3.1 Overview: main results of the studies An overview over the objectives and the main results and conclusions of the individual studies is presented in Table S3. Table S3 Title of the individual studies as well as their objectives and main conclusions. Study Objectives Main Conclusions Study 1: Fate of low molecular weight organic substances in soil: from microbial uptake to utilization and stabilisation • Overview of microbial metabolism and C allocation within the microbial metabolism • Identification of specifics in the LMWOS utilization of various microbial groups in soil • The entry steps of an LMWOS to the basic C metabolism accounts for the fate of the 3 classes of LMOWS in soil Study 2: Improved δ 13 C analysis of amino sugars in soil by Ion Chromatography - Oxidation - Isotope Ratio Mass Spectrometry • Establishment of a new instrument coupling of ion chromatograph with isotope ratio mass spectrometers for routine measurements of waterdissolvable metabolites by IRMS • Development of an IC-O-IRMS method for parallel amino sugar quantification and δ 13 C determination • IC-O-IRMS enables reliable and routine measurements of amino sugars • Parallel quantification and δ 13 C determination of basic and acidic amino sugars is possible in soils • IC-O-IRMS has great advantages compared to classical LC-O-IRMS Study 3: Biochemical pathways of amino acids in soil: Assessment by position-specific labeling and 13 CPLFA analysis • Identification of transformation pathways of two representative amino acids (alanine and glutamate) • Assessment of specifics in amino acid metabolism of individual microbial groups in soils • Basic C metabolism accounts for the majority of the observed alanine transformations • Incorporation of glutamate C-2 reflects specific microbial pathways Study 4: Biogeochemical transformations of amino acids in soil assessed by position-specific labeling • Effect of concentration (i.e. C availability) on alanine transformations • Kinetics of sorption, extracellular transformation and microbial uptake and utilization of alanine • Biotic processes outcompete sorption in soils • Extracellular LMWOS transformations are only relevant at low concentrations or in special microhabitats • Concentration strongly affects the metabolic pathway and fate of alanine C in soils Study 5: Sorption affects amino acid pathways in soils: Implication from positionspecific labeling of alanine • Microbial availability of LMWOS sorbed to various sorbents • Effect of sorption on microbial utilization and transformation pathway of LMWOS • Transformation of sorbed LMWOS follows classical biochemical pathways: no abiotic transformations occur • C allocation through specific pathways in microbial metabolism is strongly affect by the sorption mechanism and bioavailability of LMWOS Extended Summary 16 Study 6: Biochemistry of hexose and pentose transformation in soil analyzed by position-specific labeling and 13 C-PLFA • Tracing metabolism pathways for hexoses and pentoses in soils • Identification of specifics in bacterial and fungal metabolism in the use of monosaccharide C for PLFA formation • Glycolysis and pentose phosphate pathway as well as backflux could be traced in soils in situ • Glucose – as a ubiquitous substrate – is spread over entire metabolism, intensively recycled and a large portion is used for microbial biomass formation Study 7: Metabolic pathways of fungal and bacterial amino sugar formation in soil assessed by position-specific 13 Clabeling • Reconstruction of the main pathways for amino sugar biosynthesis in soils • Identification of specifics in bacterial and fungal metabolism during cell wall formation • Formation of amino sugars follows several pathways: 1) direct intact glucose utiliation, 2) glycolysis and backflux and 3) pentose phosphate pathway and backflux • Fungi showed a lower metabolic activity than bacteria in maintenance metabolism Study 8: Formation and transformation of fatty acids in soils assessed by positionspecific labeling of precursors • Reconstruction of the main pathways for fatty acid formation and transformation in soils • Identification of microbial group specific fatty acid metabolism • Assessment of microbial transformation of the free fatty acid pool in soils • Acetate 13 C was rarely used for building of fatty acid backbones but mainly for elongations or introductions of functional groups • Fatty acids were intensively modified in soils according to the demand of the microbial community which has to be considered for paleoenvironmental applications Study 9: Organic N uptake by plants - Reevaluation by positionspecific labeling of amino acids • Determination of the overestimation of uniformly labeling approaches • Evaluation of the relevance of intact amino acid uptake for the N nutrition of agricultural plants • Quantification of intact amino acid uptake based on uniform labeling causes an 1.2-3fold overestimation • Microbial utilization strongly dominated the fate of organic substances in soils • The majority of amino acid uptake by plants was explained by passive uptake of microbial transformation products 1.3.2 Determination of metabolic pathways of amino acids, monosaccharides and organic acids 1.3.2.1 Similarities and differences of individual LMWOS Study 1 aimed at investigating the similarities and differences of the fate of the applied LMWOS, to gain insights in their microbial utilization and decomposition. The percentage of 13 C-LMWOS remaining in the SOM and incorporated into the microbial biomass pool after 3 days was similar for the substances investigated (Figure S4). However, after 10 days, significant differences between the incorporation of individual substances Extended Summary 17 arose. The initial rapid uptake was quite similar for all investigated LMWOS as they are ubiquitous substrates for many members of the microbial community. However, intracellular metabolism may have accounted for the observed differences in the fate of LMWOS-C in soil and this effect became evident after 10 days of continued transformation of the LMWOS (Figure S4). Fig. S4 13 C recovery (in % of applied 13 C) from six LMWOS in soil, microbial biomass and PLFA, 3 and 10 days after addition. Letters indicate significant differences in 13 C incorporation of the individual substances if occuring. LMWOS specifics of incorporation into the microbial biomass depend on their entry point into the basic C metabolism of microorganisms (Figure S5): Microbial uptake of both amino acids was similar on day 3, but much less glutamate C than alanine C remained in microbial biomass on day 10 (Figure S4). This reflects the fact that substrates with direct incorporation into the oxidizing citric acid cycle (Figure S5), such as glutamate, are preferentially oxidized for energy production compared to LMWOS, like alanine, which enter glycolysis. The high and rapid glucose uptake is connected to the fact that glucose is the most abundant sugar in soils (Derrien et al., 2006; Derrien et al., 2004; Fischer et al., 2010a). In comparison to amino acids and carboxylic acids, glucose was preferentially incorporated into microbial biomass. This reflects the preference of glycolysis substrates for anabolic utilization compared to catabolism. For carboxylic acids, the 13 C incorporated in microbial biomass declined by a factor of two from day 3 to day 10, which also reflects the preferred catabolic oxidation of substances entering the citric acid cycle (Figure S5). Incorporation of LMWOS 13 C into microbial membrane lipids, the PLFA, enables the identification of preferences of individual functional microbial groups for certain substrates Extended Summary 24 of plant-derived fatty acid fingerprints and their isotope signature for paleoenvironmental studies is hampered. Fig. S7 Metabolic pathways of fatty acid formation from acetate and fatty acid transformations of palmitate in soil. 1.3.3 Identification of specific metabolic pathways 1.3.3.1 Specific pathways of individual members of the microbial community in soils The investigated compound classes of microbial products are on the one hand specific for a certain biosynthetic pathway within microbial cells. On the other hand, PLFA and the amino sugars are also biomarkers for individual microbial groups in soils. This enables the identification of specific pathways of individual members of the soil microbial community. Amino sugars enable fungi and bacteria to be distinguished between (Engelking et al., 2007, Glaser et al., 2004): In study 7, bacterial muramic acid showed the higher dynamics of 13 C replacement of the cell walls (study 7, Figure 3). In addition the DI of individual C positions changed strongly from day 3 to day 10 for bacterial muramic acid but remained constant for fungal galactosamine (study 7, Figure 2). This reflects a more active maintenance metabolism in bacteria, which caused an increasing incorporation of glucose metabolization fragments within a period of 10 days. Therefore, this study proved for the first time that differences in the C turnover in the slowand fast-cycling branch of soil food webs can be attributed to metabolic processes (Moore et al., 2005). A more detailed fingerprint of the microbial community, especially for the prokaryotic groups, can be gained by individual PLFA (MooreKucera and Dick, 2008). Combining statistical grouping of the fatty acids with known fatty acid fingerprints enables functional Extended Summary 25 microbial groups to be distinguished in soil (studies 3, 6 and 8). The DI revealed for glutamate incorporation into PLFA that two functional microbial groups – gram-positive II and fungi – showed a specific C allocation: Glutamate C-2 was transferred from the citric acid cycle towards fatty acid synthesis. Special reactions (e.g. the glyoxylate bypass) are necessary to avoid the oxidation of glutamate C-2 (Caspi et al., 2012). None of the other microbial groups showed C-2 incorporation into PLFA (see study 3, Figure 4). Consequently, this was the first time that specific pathways of individual members of the soil microbial community could be traced in soils in parallel. 1.3.3.2 Pathways under various concentrations of LMWOS C allocation within metabolic pathways is strongly affected by environmental conditions. The bioavailability of C sources, especially LMWOS, is one of the main factors controlling the state of the microbial community - from maintenance to growth (Blagodatskaya et al., 2007, Fischer et al., 2010b; Schneckenberger et al., 2008). In study 4, alanine transformations were investigated within an alanine concentration gradient representing the full range of concentrations in soil: from bare soil (0.5 µM) to hot spots (5 mM). At low alanine concentration, uptake was described by Michaelis-Menten kinetics, whereas with increasing concentration, linear kinetics dominated microbial uptake (Jones and Hodge, 1999). This shift in uptake kinetics revealed that two mechanisms were responsible for the microbial uptake: An unsaturable, unspecific uptake of intact alanine at hot-spot concentrations and specific, active uptake mechanisms at low alanine concentrations. The DI of the non-extractable pool of microbial transformation products (containing macromolecules as well as lipids) reflected additional shifts in the intensity of the alanine metabolism pathways (Figure S8). A significantly increased incorporation of C-1 under lowest C availability may indicate C starvation pathways e.g. anapleurotic pathways. A convergence of the DI of C-2 and C-3 for highest alanine concentration is characteristic of many anabolic pathways for biomass formation under growth conditions (e.g. lipid biosynthesis from alanine for the formation of new cell membranes: see figures S5 and S7). Extended Summary 26 Fig. S8 Concentration-dependent position-specific transformation index DI i (N=6, ± SEM) of alanine C position incorporated into the not-extractable pool of microbial biomass compounds. These and further results from study 4 indicate an altered C allocation into individual pathways that is dependent on substrate availability: from anabolic pathways characteristic for C deficiency via maintenance metabolism towards pathways common for growing cells (Figure S9). Fig. S9 General biochemical pathways of amino acids metabolization in soil as depending on alanine availability. Line width represents the qualitatively estimated relative shifts of alanine C between certain pathways dependent on the alanine concentration. 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0.5 µM 5 µM 50 µM 500 µM 5000 µM DI i C-1 Alanine C-2 Alanine C-3 Alanine Extended Summary 27 1.3.3.3 Pathways of sorbed LMWOS Sorption is one of the most likely processes causing the long-term stabilization of C in soils (Gonod et al., 2006, Duemig et al., 2012). Nevertheless, a part of the sorbed LMWOS remains bioavailable: Study 5 showed for alanine that at least 20-50% of the mineral-sorbed alanine was microbially metabolized. However, sorption reduced the availability and consequently affected transformation pathways (Jones and Edwards, 1998). In study 5, transformations of position-specifically labeled alanine, sorbed to five sorbents (two iron oxides with different crystalline structure: goethite and hematite; two clay minerals with 2:1 layers – smectite, and 1:1 layers – kaolinite; and active coal) were investigated. Goethite and active coal showed the highest amount of sorbed alanine (~45% of added alanine), and the lowest portion of the sorbed alanine C was microbially utilized (26 and 22%, respectively), whereas clay minerals showed lower sorption (1026% of added alanine) and a higher portion that was microbially available (30-35%). The stronger the sorption by the individual sorbent, the lower the microbial utilization was (Jones and Hodge, 1999). The fate of individual molecule positions reflected that, at least for the four mineral phases, alanine was processed by the classical biochemical pathways: deamination, decarboxylation of C-1 by pyruvate dehydrogenase and further oxidation of C-2 and C-3 in the citric acid cycle (Djikstra et al., 2011). However, the intensity of microbial pathways depended on the bioavailability of the sorbed substrate: the less alanine was accessible, the less was oxidized by catabolism and the more alanine C was used for anabolism, i.e. the formation of microbial biomass. 0 2 4 6 8 0 10 20 30 40 50 60 70 80 time (h) ratio C-1/C-2,3 in respired CO2 heamatite goethite smectite kaolinite active coal Fig. S10 Ratio of C-1 to (C-2+C-3)/2 respiration of alanine C for the 5 applied sorbents calculated from the fitted, position-specific oxidation rate. Extended Summary 28 The ratio of C-1 oxidation by pyruvate dehydrogenase versus oxidation of C-2 and C-3 in the citric-acid cycle (Figure S10) depended on the microbial availability of alanine: High availability due to fast desorption of cation-exchangeable bound alanine caused an initial peak in C-1 oxidation by glycolysis for the two clay minerals and an abrupt shift to oxidation via the citric acid cycle. However, low microbial availability of alanine sorbed to iron minerals led to a parallel oxidation of all three positions by glycolysis and citric acid cycle (represented by the C-1/C-2,3 ratio in Figure S10). This slower oxidation rate was associated with an increase in C allocation towards anabolism (Djikstra et al., 2001) (Figure S11). Fig. S11 Metabolic pathways of alanine sorbed on clay minerals (smectite and kaolinite), iron oxides (hamatite and goethite) and active coal. Detailed explanations in text. Various colors show the pathways of C from individual positions of alanine. Line width represents the qualitatively estimated relative shifts in the fate of alanine C positions between certain pathways dependent on the sorbent class. Extended Summary 29 Alanine sorbed to active coal showed a deviating behavior with preferential stabilization of C-3 and oxidation of C-1 and C-2 (study 5, Figure 3 and 4). This indicates that in addition to basic microbial mechanism, further stabilization and modified transformations of sorbed alanine occured, e.g. by exoenzymatic degradation (Lehmann et al., 2011). Potential desorption processes and pathways are shown in Figure S11. In general, position-specific labeling revealed that the fate of amino acid C in soil is strongly affected by sorption: The stronger the sorption, the more metabolized LMWOS C is allocated into microbial biomass compounds (Figure S11). 1.3.3.4 Extraversus intracellular transformation pathways Besides intracellular microbial pathways, extracellular transformations may also play an important role in LMWOS transformations in soils, especially in microhabitats where intact cells have no access (von Luetzow et al., 2006). To distinguish extraand intracellular transformation pathways, selective inhibition of cellular, energy-dependent pathways was performed in study 4 (Figure S12). 0.02 0.03 0.04 0.05 0 5 10 15 20 25 30 35 40 time (h) biotic Ala-C removal (µmol) C-1 Alanine Microorganisms C-2 Alanine Microorganisms C-3 Alanine Microorganisms C-1 Alanine Exoenzymes C-2 Alanine Exoenzymes C-3 Alanine Exoenzymes Fig. S12 Removal of alanine from soil solution by extraand intracellular processes without inhibition (filled symbols, dashed line) and by extracellular transformation in respiration-inhibited treatments (open symbols, dotted line); Experimental points (means ± SEM, N=6) and fitted curves based on an exponential utilization model are presented. This approach could prove the existence of extracellular transformation of LMWOS in soils: Alanine was decomposed by a stepwise extracellular oxidation starting from the Extended Summary 30 carboxylic group, presumably by rather unspecific exoenzymes (Hofrichter et al., 1998). However, comparing the kinetics of extraversus intracellular processes (Figure S12) revealed that cellular uptake of LMWOS always out-competed extracellular transformations, which are quantitatively relevant only at very low alanine concentrations or in specific microhabitats. 1.3.4 Kinetics and ecological relevance of competing sinks for LMWOS 1.3.4.1 Sorption versus microbial utilization As position-specific labeling enables transformation pathways in soils to be distinguished, it can be used to assess the relevance of individual pathways and their fluxes. Sorption, as an LMWOS sink in soils, can interact and compete with microbial utilization (Fischer et al., 2010b; Kaiser and Kalbitz, 2012). Study 4 assessed the sorption of alanine in sterilized soil and found that sorption occured as intact molecules and no abiotic cleavage of alanine was detected (in contrast to results of Wang and Huang (2005)). In study 5, the microbial utilization of sorbed alanine was assessed: Desorption, at least from mineral phases, occurred mainly as intact molecules, too. The observed position-specific transformations in both studies were mainly characterized by the microbial metabolism. Therefore, sorption and desorption do not directly contribute to the transformation of LMWOS, but only prevent translocation (Kaiser and Kalbitz, 2012) or microbial transformation. If kinetics of sorption and microbial uptake were compared (study 4), microbial utilization outcompeted sorption in soils. 1.3.4.2 Plant uptake versus microbial utilization Besides microorganisms, plants are also known to have the ability for LMWOS uptake by their roots (Fischer et al., 1998). The relevance of this process is still in question for many ecosystems due to the effective competition of microorganisms for LMWOS (Hodge et al., 2000; Jones et al., 2005a). The main problem of evaluating the uptake of intact amino acids is methodological constraints: The commonly used dual-isotope labeling approaches coupled with bulk isotope measurements cause an overestimation of the calculated intact amino acid uptake (Sauheitl et al., 2009). Study 9 evaluated the intact uptake of the amino acid alanine by plants using position-specific labeling. Consequently, intact uptake could be distinguished from the uptake of transformation fragments. Position-specific 14 C-labeling revealed that a minor portion (less than 1.5% of the applied Extended Summary 31 amino acids) was taken up intact, whereas the majority of alanine (~98.5%) was used by soil microorganisms (Figure S13). Uptake calculated from uniform labeling reflected an overestimation for the factor 1.2-3 of the quantified intact uptake. Preferential uptake of C-3 by all three plant species indicated that the uptake of microbial transformation fragments occurred. Previous studies in this thesis revealed that microbial biomass compounds are also characterized by a dominance of alanine C-3 (see study 3 and 4). Consequently, microbial uptake and transformation can produce mineralized N as well as fragments of the C skeleton, which were partially available in the soil solution for root uptake (Jones et al., 2005). Labeling with the N-free LMWOS acetate showed a similar uptake of N-containing and N-free LMWOS (< 2% of the applied 14 C). This indicated that plant uptake of LMWOS mainly occurred via passive uptake mechanisms. These passive mechanisms can also account for the unspecific uptake of microbial transformation products. In summary, study 9 suggests that N uptake from organic sources is of minor importance for N nutrition of agricultural plants and even more for the fate of LMWOS in soils (Jones et al., 2005, Hodge et al., 2000). Fig. S13 Illustration of the fate of alanine (numbers represent % of applied tracer: this can either be taken up intact or degraded/mineralized to fragments and subsequently incorporated into microorganisms or plant biomass). Extended Summary 32 1.4 Conclusions The high flux of C through the pool of LMWOS clearly defines them as a crucial C pools in the SOC cycle. Many previous studies have analyzed the rates and turnover of LMWOS in soil, but the underlying mechanisms and pathways of C transformation of LMWOS remains unknown. Therefore, these studies were focused on biogeochemical pathways of three main groups of LMWOS in soil: amino acids, monosaccharides and organic acids. Tracing their transformations was achieved by combining for the first time position-specific 13 C-labeling with compound-specific isotope analysis (CSIA). The application of individual LMWOS revealed that entry steps into basic C metabolism account for specifics in the C partitioning between microbial cataand anabolism: substrates entering citric acid cycle were preferentially mineralized (>80% in 10 days) whereas e.g. monosaccharides entering glycolysis were preferentially allocated to anabolic pathways and incorporated into microbial biomass (less than 70% mineralized in 10 days). Position-specific 13 Cand 14 C-labeling provided a unique submolecular approach to reconstruct the main pathways of C transformation in soil and their specifics in individual microhabitats (e.g. at mineral surfaces or at the plant-root interface). The divergence index (DI) was developed and proven to be a valuable tool to compare the position-specific fate of individual substances independent of the used isotopic approach or experimental design used or the pool investigated. 13 C incorporation in various microbial compound classes was traced by compound-specific isotope analysis: fatty acids by GC-C-IRMS and amino sugars by IC-O-IRMS. Therefore, a new instrument coupling was applied and purification and measurement methods for soil amino sugar δ 13 C analysis were established and evaluated. Basic microbial C metabolism with glycolysis, pyruvate dehydrogenase oxidation and citric acid cycle could be traced in soil under field and laboratory conditions. Oxidizing, catabolic pathways are ongoing in soils in parallel to constructing, anabolic pathways (like gluconeogenesis): for example, up to 55% of the glucose allocated to amino sugar synthesis was not intact glucose but derived from glucose metabolites allocated by gluconeogenesis backflux towards amino sugar formation. Consequently, substrates entering glycolysis are intensively recycled within the cellular C pool, which was shown by a continued decrease of their divergence index. Specific tracers for individual biosynthetic pathways were identified, which allowed transformations to be followed within these side branches of the basic C metabolism: 1) the pentose phosphate pathway was detected by a combination of hexose and pentose 13 C labeling; and 2) turnover within the cellular lipid pool was proven by 13 C labeling of Extended Summary 33 short- (acetate) and long-chain (palmitate) precursors of PLFA. Within 10 days, 65% of the incorporated palmitate was transformed (e.g. by desaturation, elongation or branching) to other fatty acids and the fingerprint of the palmitate 13 C-derived fatty acids approached the PLFA pattern of the present microbial community. Knowledge of these fast fatty acid transformations is crucial for the application of fatty acid fingerprints and their isotopic values for palaeoenvironmental reconstructions. An intensive turnover was not only shown for lipids but also for cell wall polymers. Metabolic recycling activity and turnover was much higher for bacteria than for fungi, which was proven by both biomarker groups – PLFA and amino sugars. Therefore, this thesis experimentally revealed one underlying, mechanistic reason for the previously observed specifics in C turnover of the slow (fungi-based) and the fast (bacteria-based) cycling branch of the soil food web. For the first time, position-specific labeling was coupled with compound-specific isotope analysis of microbial biomarkers. This combination provides a novel opportunity to trace simultaneous, biosynthetic pathways of individual microbial groups in diverse microbial communities of soils. Furthermore, variations of environmental factors, like substrate concentration, were identified as the main regulatory factors for C allocation within microbial metabolism: Concentration gradients characteristic for soils from C-poor bulk soils to hotspots involved a shift of C allocation within metabolic pathways from C starvation pathways via maintenance pathways towards pathways that are characteristic for cells under growth conditions. Sorption, as a soil-specific process reducing the bioavailability of a substrate, affected microbial metabolism: the stronger a substrate is sorbed, the more of its C is allocated towards anabolism, e.g. is found in the microbial products. Understanding these shifts in metabolic pathways is crucial for the SOC cycle, as C allocation towards anabolism is the prerequisite for the formation and stabilization of microbially-derived SOM. Three soil-specific processes were traced in parallel with microbial utilization: sorption, exoenzymatic LMWOS utilization and plant uptake. Sorption, as well as desorption, occurred as intact molecules and did not account for LMWOS transformations. Exoenzymes caused a stepwise oxidation of the LMWOS C backbone. However, their kinetics could not compete with microbial uptake systems. Consequently, extracellular transformations can only be relevant in specific soil microhabitats, which are inaccessible for microorganisms. Intact uptake of amino acids by plants was assessed by dual-isotope position-specific 13 Cand 15 N-labeling. This new approach revealed the overestimation of intact amino acid uptake due to methodological constraints of previous studies and showed that less than 1.5% of the applied amino acids were taken up intact by plants. Consequently, none of the investigated abiotic or biotic processes could compete with Extended Summary 40 1.6 Contribution to the included manuscripts and publications The publications and manuscripts included in this PhD thesis were prepared in cooperation with various coauthors. The coauthors listed in these publications and manuscripts contributed as follows: Study 1: Fate of low molecular weight organic substances in an arable soil: from microbial uptake to utilisation and stabilisation Status at date of thesis submission: Submitted to Soil Biology and Biochemistry; Status at date of thesis printing: Submitted to Soil Biology and Biochemistry; Contributors: Anna Gunina 45% Accomplishment of experiments, laboratory analysis, data preparation, preparation of the manuscript Michaela Dippold 40% Experimental design, accomplishment of the experiments, data preparation, preparation of the manuscript Bruno Glaser 3% Discussions on experimental design, suggestions to improve manuscript Yakov Kuzyakov 12% Experimental design, discussions on the results, suggestions to improve manuscript Study 2: Improved δ 13 C analysis of amino sugars in soil by Ion Chromatography - Oxidation - Isotope Ratio Mass Spectrometry Status at date of thesis submission: In revision at Rapid Communications in Mass Spectrometry; Status at date of thesis printing: Published in Rapid Communications in Mass Spectrometry; Contributors: Michaela Dippold 55% Establishment of purification method , laboratory analysis, data preparation, preparation of the manuscript Stefanie Boesel 20% Establishment of measurement method, suggestions to improve manuscript Anna Gunina 12% Laboratory analysis, comments to improve manuscript Extended Summary 41 Yakov Kuzyakov 3% Suggestions to improve manuscript Bruno Glaser 10% Discussions on method development, suggestions to improve manuscript Study 3: Biochemical pathways of amino acids in soil: Assessment by position-specific labeling and 13 C-PLFA analysis Status at date of thesis submission: Accepted in Soil Biology and Biochemistry Status at date of thesis printing: Published in Soil Biology and Biochemistry Contributors: Carolin Apostel 42% Accomplishment of experiments, laboratory analysis, data preparation, preparation of the manuscript Michaela Dippold 40% Experimental design, accomplishment of the experiments, data preparation, suggestions to improve manuscript Bruno Glaser 3% Laboratory analysis, suggestions to improve manuscript Yakov Kuzyakov 15% Experimental design, discussions on the results, suggestions to improve manuscript Study 4: Biogeochemical transformations of amino acids in soil assessed by position-specific labelling Status at date of thesis submission: Accepted in Plant and Soil Status at date of thesis printing: Published in Plant and Soil Contributors: Michaela Dippold 85% Experimental design, accomplishment of experiment, data preparation, preparation of the manuscript Yakov Kuzyakov 15% Experimental design, discussions on the results, suggestions to improve manuscript Study 5: Sorption affects amino acid pathways in soil: Implications from position-specific labeling of alanine Status at date of thesis submission: Submitted to Soil Biology and Biochemistry Status at date of thesis printing: Accepted in Soil Biology and Biochemistry Extended Summary 42 Contributors: Michaela Dippold 60% Experimental design, data preparation, preparation of the manuscript Mikhail Biryukov 25% Accomplishment of experiment, data preparation, Yakov Kuzyakov 15% Experimental design, discussions on the results, suggestions to improve manuscript Study 6: Biochemistry of hexose and pentose transformations in soil analyzed by position-specific labeling and 13 C-PLFA Status at date of thesis submission: To be submitted to Soil Biology and Biochemistry Status at date of thesis printing: Resubmitted to Soil Biology and Biochemistry Contributors: Michaela Dippold 45% Experimental design, accomplishment of experiments, data preparation, preparation of the manuscript Carolin Apostel 45% Accomplishment of experiment, data preparation, preparation of the manuscript Yakov Kuzyakov 10% Discussions on the results, suggestions to improve manuscript Study 7: Metabolic pathways of fungal and bacterial amino sugar formation in soil assessed by position-specific 13 C-labeling Status at date of thesis submission: To be submitted to Biogeochemistry Status at date of thesis printing: Submitted to Soil Biology and Biochemistry Contributors: Michaela Dippold 60% Experimental design, accomplishment of experiments, paration, prepar data preparation, preparation of the manuscript Anna Gunina 20% Laboratory analysis, suggestions to improve manuscript Stefanie Boesel 5% Laboratory analysis, data preparation Bruno Glaser 3% Suggestions to improve manuscript Yakov Kuzyakov 12% Discussions on experimental design, suggestions to improve manuscript Extended Summary 43 Study 8: Formation and transformation of fatty acids in soils assessed by position-specific labeling of precursors Status at date of thesis submission: To be submitted to Organic Geochemistry Status at date of thesis printing: Submitted to Geochimica et Cosmochimica Acta Contributors: Michaela Dippold 85% Experimental design, accomplishment of experiment, data preparation, preparation of the manuscript Yakov Kuzyakov 15% Experimental design, discussions on the results, suggestions to improve the manuscript Study 9: Organic N uptake by plants - Reevaluation by position-specific labeling of amino acids Status at date of thesis submission: Submitted to Soil Biology and Biochemistry Status at date of thesis printing: Submitted to Journal of Experimental Botany Contributors: Daniel Moran 42% Accomplishment of experiment, data preparation, preparation of the manuscript Michaela Dippold 42% Experimental design, accomplishment of experiment, data preparation, preparation of the manuscript Bruno Glaser 3% Laboratory analysis, suggestions to improve the manuscript Yakov Kuzyakov 13% Experimental Design, Discussions on the results, suggestions to improve the manuscript Publications and Manuscripts 44 2 Publications and Manuscripts 2.1 Study 1: Fate of low molecular weight organic substances in an arable soil: from microbial uptake to utilisation and stabilisation Anna Gunina 1,2,3 , Michaela A. Dippold 1,2 , Bruno Glaser 4 , Yakov Kuzyakov 1,5 1 Department of Agricultural Soil Science, Georg-August-University of Göttingen 2 Department of Agroecosystem Research, University of Bayreuth 3 Faculty of soil science, Moscow Lomonosov State University 4 Department of Soil Biogeochemistry, Institute of Agricultural and Nutritional Science, Martin-Luther University Halle-Wittenberg 5 Department of Soil Science of Temperate Ecosystems, Georg-August-University of Göttingen Corresponding Author: Anna Gunina Max Planck Institute for Biogeochemistry Am Herrenberge 11 07745 Jena email: guninaa[email protected]m Tel.: 0157/85566093 Publications and Manuscripts 45 Abstract Microbial uptake and utilization are the main transformation pathways of low molecular weight organic substances (LMWOS) in soil, but detailed of transformations is strongly limited. As various LMWOS classes enter biochemical cycles at different steps, we hypothesize that the percentage of their carbon (C) used by microbial biomass and consequently stabilization in soil is different. Representatives of the three main groups of LMWOS: amino acids (alanine, glutamate), sugars (glucose, ribose) and carboxylic acids (acetate, palmitate) – were applied at naturally-occurring concentrations into a loamy arable Luvisol in a field experiment. Incorporation of 13 C from these LMWOS into microbial biomass (MB) and into phospholipid fatty acids (PLFAs) was investigated 3 d and 10 d after application. The microbial utilization of LMWOS for cell membrane construction was estimated by replacement of PLFA-C with 13 C. Mineralization of LMWOS to CO 2 comprised 20–65% of the initially applied 13 C, whereas 13 C incorporation into MB amounted to 10–24% at day 3 and was reduced to 1– 15% on day 10. Maximal incorporation of 13 C into MB was observed from sugars and minimal from amino acids. Strong differences in microbial utilization between LMWOS were observed mainly at day 10. Thus, despite similar initial rapid uptake by microorganisms, further metabolism within microbial cells accounts for the specific fate of C from various LMWOS in soils. 13 C from each LMWOS was incorporated into each PLFA. This reflects the ubiquitous utilization of all LMWOS by all functional microbial groups. The preferential incorporation of palmitate into PLFAs reflects its role as a direct precursor for fatty acids. Higher 13 C incorporation from alanine and glucose into PLFAs compared to glutamate, ribose and acetate reflects the preferential use of glycolysis-derived substances in the fatty acids synthesis. Gram-negative bacteria (16:1ω7c and 18:1ω7c) were the most abundant and active in LMWOS utilization. Their high activity corresponds to a high demand for anabolic products, e.g. to dominance of pentose-phosphate pathway, i.e. incorporation of ribose-C into PLFAs. The 13 C incorporation from sugars and amino acids in filamentous microorganisms was lower than in all procaryotic groups. However, for carboxylic acids, the incorporation was in the same range (0.1 – 0.2% of the applied carboxylic acid C) as that of gram-positive bacteria. This may reflect the dominance of fungi and other filamentous microorganisms for utilization of acidic and complex organics. Thus, we showed the divergence of C pathways from LMWOS over the 10 days, despite their similar initial uptake by microorganisms. Consequently, stabilization of C in Publications and Manuscripts 46 soil is mainly connected not with its initial microbial uptake, but with its incorporation into microbial compounds of various stability. Keywords: low molecular weight organic substances, 13 C-labelling, monosaccharides, amino acids, carboxylic acids, compound-specific isotope analysis, PLFAs turnover, soil microorganisms Publications and Manuscripts 47 2.1.1 Introduction Low molecular weight organic substances (LMWOS) comprise 5–10% of dissolved organic carbon (DOC) in soils (Ryan et al., 2001) and are products of rhizodeposition, above and belowground litter and microbial residue degradation. Microbial removal of LMWOS from solution in the upper soil horizons appears within minutes (Jones et al., 2004), whereas the half-life of C from LMWOS is much longer, from several hours to months or even decades (van Hees et al., 2005). This occurs due to rapid microbial uptake und further utilisation of LMWOS within the microbial biomass, which out-compete processes of physicochemical sorption of LMWOS at mineral surfaces and their leaching from the soil profile, probably by orders of magnitude (Fischer et al., 2010). Due to the strong link LMWOS dynamics with microbial utilization, the fate of LMWOS should be investigated at natural applied amounts, to avoid any changes in microbial response strategy. The main compound classes within the LMWOS are amino acids, sugars (mainly monosaccharides) and carboxylic acids (Fischer et al., 2010). Amino acids represent the largest pool of N in soils, mainly bound in proteins. About 30% of N obtained after acid hydrolysis from the protein pool (Stevenson, 1982) and a large portion of N released from soil organic matter (SOM) by enzymes are amino acids (Barraclough, 1997). Amino acidC half-lives are between 3–45 days and do not strongly differ between field and laboratory conditions (Glanville et al., 2012). The great variability in amino acid utilisation reported in the literature is a consequence of the diversity of metabolic pathways within microbial cells (Apostel et al., 2013) and also can be depend on activity of microorganisms (Jones et al., 2005). Numerous studies have reported that carbohydrates are the most abundant substance class, amounting for 5–25% of soil organic matter (SOM) (Benzingpurdie, 1980; Cheshire, 1979). Glucose is the most abundant carbohydrate derived either from the decomposition of plant residues (Derrien et al., 2006) or from root exudates (Derrien et al., 2004; Fischer and Kuzyakov, 2010). Half-life of glucose-derived C comprises around 15 days in field conditions (Glanville et al., 2012). On average, 60% of the added glucose is incorporated into cellular compounds (Fischer et al., 2010) Despite glucose supposed to be a ubiquitous substrate, which can be used by nearly all microorganisms (Macura and Kubatova, 1973), specifics of its utilisation in soil is still a topic of discussion (Reischke et al., 2014). The third most abundant class of LMWOS in soils is carboxylic acids. 80–90% of C from the applied carboxylic acids were decomposed during the first 7 days and only 10– 20% of C were incorporated into microbial biomass (MB) (Unteregelsbacher et al., 2012; Publications and Manuscripts 48 van Hees et al., 2002). The utilisation of carboxylic acids is substrate-dependent: acetate has a lower mineralization capacity than citrate and oxalate (van Hees et al., 2002) and citrate can be degraded faster than malate and oxalate (Strom et al., 2001). At the intramolecular level, –COOH groups can be oxidized to CO 2 very rapidly, whereas CH 3 - groups are preferentially used for biosynthesis (Dippold and Kuzyakov, 2013; Fischer and Kuzyakov, 2010). Thus, differences in carboxylic acid utilisation can be attributed to their various roles in cell metabolism as well as to their differences in chemical structure. Studies that simultaneously compare the fate of amino acids, sugars and carboxylic acids are not numerously reported in literature. In most cases studies either consider the microbial utilisation of LMWOS by the entire MB (Glanville et al., 2012; Rousk and Baath, 2011) or focus on contribution of various microbial groups to LMWOS utilization (Apostel et al., 2013; Rinnan and Bååth, 2009; Rinnan et al., 2013). Information concerning the evaluation of the contribution of functional microbial groups to LMWOS utilization can be obtained by coupling of 13 C or 14 C labeling with analysis of microbial biomarkers such as amino sugars (Amelung et al., 2001; Engelking et al., 2007; Glaser et al., 2004), phospholipids-derived fatty acids (Frostegard et al., 2011; Zelles, 1997) or DNA-based methods (Ibekwe et al., 2002, Radejewski et al. 2003). Coupling PLFA analysis with 13 C-labelling has shown that gram-negative (G-) bacteria are more active in the utilisation of plant C (low or high molecular weight) than grampositive (G+) bacteria, even if the latter group has a higher PLFA content in soil (GarciaPausas and Paterson, 2011; Waldrop and Firestone, 2004). Fungi contribute less to the utilisation of plant-derived C than bacteria (Waldrop and Firestone, 2004). In contrast, the use of 13 C pulse-labelling of plants to trace 13 C in PLFAs has shown that either fungi (Butler et al., 2003) or Gbacteria (Tian et al., 2013) are the most active in rhizodeposits consumer. Incorporation of 13 C from labelled straw into PLFAs has shown that fatty acids such as 16:0; 18:1w9, 18:2w6,9 were more 13 C-enriched, whereas other 16:1w5 or 10Me17:0 fatty acids contained negligible amounts of 13 C (Williams et al., 2006). Consequently, members of the microbial community are differentially involved in the assimilation of litteror root-derived C (Williams et al., 2006) and the activity of individual microbial groups appears to depend on the quality of substrate and on environmental conditions such as soil type, season and climatic conditions (Bray et al., 2012). Thus, general principles of LMWOS utilisation by individual groups of bacteria and fungi still remain open. The second factor controlling LMWOS fate is microbial metabolism: various classes of LMWOS enter different pathways and consequently are utilised differently. Sugars are mainly used directly by the basic glycolysis pathway (Caspi et al., 2008; Keseler et al., 2009), carboxylic acids enter from side branches of the citric acid cycle (Caspi et al., Publications and Manuscripts 49 2008; Keseler et al., 2009), and amino acids enter glycolysis or the citric acid cycle from individual side branches at different steps (Apostel et al., 2013; Knowles et al., 2010). Thus, we assume that universal substances such as sugars, entering glycolysis directly, will be metabolised very rapidly in comparison to carboxylic acids and amino acids entering the citric acid cycle. However, glycolysis also enables entry into many anabolic pathways, i.e., we hypothesise that sugars are used more for anabolism than carboxylic acids, which enter the oxidising citric acid cycle and can be directly metabolised for energy production. The highest diversity in pathways can be expected for amino acids, because they enter basic metabolism at various steps (Apostel et al., 2013). Since carboxylic acid utilisation is substrate-controlled, we expect divergence in the utilisation of shortand long-chain acids. Because three classes of LMWOS enter metabolic cycles at various points, we hypothesise that their role in the synthesis of cell components such as PLFAs should be different. Thus, the overall aim of this study was to estimate the short-term transformation of representatives of three main classes of LMWOS: monosaccharides (glucose and ribose), carboxylic acids (acetate and palmitate) and amino acids (alanine and glutamate) under field conditions, coupling 13 C substrate labelling with the analysis of specific cell PLFA biomarkers. 2.1.2 Material and Methods 2.1.2.1 Experimental design The field experiment was carried out at an agricultural field trial in Hohenpölz (49°54' N, 11°08'E, at 500 m a.s.l.). The mean annual temperature was +7 o C and mean annual precipitation was 870 mm. The site is cultivated by a rotation of triticale, wheat and barley. The soil is an arable loamy haplic Luvisol (IUSS Working group WRB, 2007) and had the following characteristics: pH 6.6, total C content 1.5%, C/N 10.7, CEC 13 cmol C kg -1 , clay content 22%. In August 2010, following harvest of the triticale and spudding of the soil, columns were inserted to a depth of 10 cm and six, 13 C uniformly-labelled substances: alanine, glutamate, glucose, ribose, sodium acetate and palmitate were injected into separate columns. The amounts of applied tracer were: alanine 96.3, glutamate 91.6, glucose 93.4, ribose 91.8, acetate 95.8 and palmitate 49.5 µmol 13 C column -1 . The amount of added C was kept as low as possible and constant for all columns, including the controls, where similar amounts of non-labelled C was applied (0.40–0.77 µg C g soil -1 ). Each column contained 1.5 kg soil. The field experiment had a randomised block design with four Publications and Manuscripts 56 Fig. 2 13 C incorporation from both amino acids (in % of applied 13 C) into PLFAs (top) and percent of 13 C replacement (in % of PLFA-C) (bottom) of microbial groups 3 and 10 days after alanine and glutamate application. Letters reflect significant differences between alanine and glutamate uptake into microbial groups. Utilisation of 13 C from sugars for PLFAs formation showed different trends in bacterial and fungal groups and much higher absolute 13 C incorporation compared to amino acids (Fig. 3, top). Between 0.01 and 0.70% of initially applied sugars 13 C was found in various taxonomic groups after three days, whereas only 0.001–0.25% was recovered 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 Alanine Glutamate Incorporation (% of applied 13C) G-1 G-2 G+1 G+2 G+3 Ac 16:1w5 F Pr a b a a a b b b Incorporation 0 0.02 0.04 0.06 a b 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 Days after LMWOS application Replacement by 13 C (% of PLFA-C) a b a b a a a b b b b Replacement a Publications and Manuscripts 57 from amino acids. The incorporation of 13 C from sugars into all microbial groups increased or remained constant between days three and 10. All bacterial species used glucose-C more efficiently than ribose-C except Ggroup 1, which preferred ribose. Among the filamentous microorganisms, fungi did not differ from 16:1w5 in glucose13 C incorporation into PLFAs, but fungi used more 13 C from ribose than 16:1w5. In general, the microbial specialisation for individual monosaccharides as building blocks for PLFAs was visible within bacterial but not within eukaryotic groups. Fig. 3 13 C incorporation from both monosaccharides (in % of applied 13 C) into PLFAs (top) and percent of 13 C replacement (in % of PLFA-C) (bottom) of microbial groups 3 and 10 days after glucose and ribose application. Letters reflect significant differences between glucose and ribose uptake into microbial groups 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Glucose Ribose Incorporation (% of applied 13 C) G-1 G-2 G+1 G+2 G+3 Ac 16:1w5 F Pr Incorporation a b a b a b 0 0.02 0.04 0.06 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 Replacement by 13 C (% of PLFA-C) Days after LMWOS application Replacement a b a a b b a b Publications and Manuscripts 58 Incorporation of 13 C from both carboxylic acids into PLFAs of Gbacteria group 1 was higher than from other LMWOS (Fig. 4, top). Other bacterial groups used 13 C from carboxylic acids less efficiently than 13 C from sugars and amino acids for PLFAs synthesis. Filamentous microorganisms (actinomycetes and fungi, but also 16:1w5) exceeded the prokaryotic groups of G+ bacteria in incorporation of 13 C from the most complex substrate, palmitate, into PLFAs. Fig. 4 13 C incorporation from both carboxylic acids (in % of applied 13 C) into PLFAs (top) and percent of 13 C replacement (in % of PLFA-C) (bottom) of microbial groups 3 and 10 days after acetate and palmitate application. Letters reflect significant differences between acetate and palmitate uptake into microbial groups. 0 0.05 0.1 0.15 0.2 Acetate Palmitate Incorporation (% of applied 13 C) G-1 G-2 G+1 G+2 G+3 Ac 16:1w5 F Pr Incorporation a a bb b a a b 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 3 10 Days after LMWOS application Replacement by 13 C (% of PLFA-C) Replacement a b a a a b b b a a b b 0 0.5 1 a b Publications and Manuscripts 59 In general, incorporation of LMWOS-C in bacterial species was higher than that in eukaryotes. 2.1.4 Discussion 2.1.4.1 Incorporation of LMWOS into SOM and microbial biomass Amino acids The mineralization of alanine and glutamate in our experiment was similar to literature data (Jones et al., 2005) and less than 50% of applied 13 С remained in the soil after 10 days. The half-life of alanine and glutamate-derived C reported for field conditions was 18 and 3 days, respectively (Glanville et al., 2012). This is much longer than in our experiment and we observed a similar mineralization of glutamate and alanine C within 10 days. These contrasting results might be attributable to differences in total microbial activity or community structure (Jones et al., 2005) in the studied soils as well as due to methodological differences. Similar 13 C amounts from glutamate and alanine remaining in the soil on day 10 reflected a similar mineralization of these amino acids and corresponds to similar microbial decomposition of differently charged amino acids (Jones and Hodge, 1999). The high amount of 13 C incorporated into the EMB pool at day three (Fig. 1) corresponded with the rapid and efficient uptake of free amino acids as intact molecules (Dippold and Kuzyakov, 2013; Geisseler et al., 2010; Jones and Hodge, 1999). After uptake, amino acids can either be oxidised for energy production, directly incorporated into proteins (Geisseler et al., 2010) or used in other metabolic pathways (Fig. 5) (Dippold and Kuzyakov, 2013; Knowles et al., 2010). The incorporation of alanine into EMB was higher than for glutamate on day 10, showing the more rapid mineralisation of glutamate C. Similarly, glutamate was utilised more rapidly than glycine and lysine over a broad concentration range (Jones and Hodge, 1999). This corresponds to the different entry point of these amino acids into metabolism (Knowles et al., 2010). Alanine enters the basic cellular metabolism at the connecting step between glycolysis and the citric acid cycle (Apostel et al., 2013; Caspi et al., 2008). Thus, it is easily distributed throughout all anabolic pathways for the synthesis of cell components, e.g., glyconeogenesis, protein synthesis, fatty acid synthesis and ribonucleotide synthesis (Fig. 5). In contrast, glutamate directly enters the citric acid cycle as oxoglutarate (Knowles et al., 2010). This demands energy for glyconeogenesis and therefore, fatty acid synthesis pathways will not be used if other more appropriate substrates are available. Thus, alanine C is preferentially incorporated into the more stable components of microbial cells – the cell walls and Publications and Manuscripts 60 the membranes – compared to glutamate. In contrast, glutamate plays a central role in the amino acid cycle, and oxoglutarate produced from glutamate by transamination will be rapidly decompose to CO 2 (Vinolas et al., 2001). Fig. 5 Primary metabolic pathways of the six representatives of three LMWOS classes (amino acids (blue), sugars (green) and carboxylic acids (red)). Thick arrows reflect the entering points of LMWOS in the metabolic pathways; black fine arrows show the basic C metabolism and shaded arrows reflect anabolic pathways for formation of cellular compounds. Sugars The half-life of glucose-C in our experiment (25% mineralized within 10 days) is within the range of previous studies: Glanville et al. (2012) reported a decomposition of 50% of glucose-C after 20 days, Saggar et al. (1999) measured a glucose-C decomposition of 51–66% within 35 days and Schneckenberger et al. (2008) observed a mineralisation of 26–44% of 14 C from glucose within 22 days. The incorporation of a significant proportion of applied 13C from sugars into EMB in our experiment is in agreement with the model of short-term glucose utilisation (Nguyen and Guckert, 2001). In this model, glucose taken up from solution is initially allocated to an intermediate pool and thereafter can be respired or used as a structural component. Thus, due to the demand for cellular products, glucose C will preferentially be transferred to anabolic pathways rather than be oxidised for energy production. Publications and Manuscripts 61 The lower mineralisation and incorporation of pentose compared to hexose on day three corresponds with its slower uptake rate, as it is a less common soil monosaccharide than glucose. The metabolisation of pentoses occurs mainly via the pentosephosphate-pathway, leading to incorporation into various cell components such as DNA or other ribonucleotides (Fig. 5). Both phenomena explain the lower utilisation of ribose compared to glucose. Preferential incorporation of ribose into biosynthetic ribonucleotide products has yet to be proven by substance-specific analysis, e.g., by stable isotope probing of DNA (Radajewski et al., 2000). Indeed, ribose mineralisation and incorporation into the EMB on day 10 were nearly the same as for glucose. This confirms that the hexose and pentose pathways are closely linked and that C from both monosaccarides is transferred within these pathways towards biosynthetic pathways according to the C demand of cells. Carboxylic acids Similar to amino acids and monosaccharides, the rapid uptake of acetate outcompetes its physicochemical sorption in soils (Fischer et al., 2010). Within the microbial biomass, acetate C can be subjected to “arrest metabolism” and stored in cells before use (Fischer and Kuzyakov, 2010). Acetate can also be transformed into carbohydrates, amino acids (Sorensen and Paul, 1971) and other cell components and thus is fixed in diverse microbial products. Incorporation of acetate into EMB was less than that of sugars, which confirms that acetate was used for respiration (ca. 80–90%) rather than for new cell biomass production (Jones and Edwards, 1998; van Hees et al., 2002). This occurs due to the oxidation of a high proportion of acetate C into the citric acid cycle. Furthermore, acetate must be activated prior to incorporation into the key metabolic pathways (van Hees et al., 2002). The transformation of acetate to anabolic products is thus unfavorable, as long as microorganisms have access to more freely available substrates. An exception to this pathway, however, is fatty acid synthesis, where acetate is a direct precursor. Palmitate is an anion of a short-chain fatty acid, the most dominant fatty acid in bacteria and fungi (Lawlor et al., 2000) and is a precursor for the synthesis of more complex fatty acids. Due to its high molecular weight and long aliphatic chain, we hypothesised that its decomposition is much slower than the decomposition of short chain carboxylic acids such as acetate. However, this hypothesis was not confirmed in our experiments. Previous studies showed that the degradation of palmitate was more rapid than that of similar or longer fatty acids (Moucawi et al., 1981). It was estimated that 41% of oleic and 31% of stearic acids were decomposed within four weeks (Moucawi et al., 1981), whereas 50% of palmitate decomposed within 10 days (Fig. 1). Publications and Manuscripts 62 In contrast to net mineralization, the incorporation of 13 C from palmitate into EMB was the lowest (Fig. 1). This might be due to hydrophobic interactions of palmitate with SOM that led to the lower uptake into microbial biomass compared to other LMWOS. However, if taken up, it was preferentially incorporated in PLFAs, but not used for the synthesis of other microbial compounds. Incorporation into PLFAs was higher than for any other LMWOS, which is in accordance with their direct precursor role for PLFA formation. In general, our results reflect that the uptake and utilisation of six LMWOS within 3 days was quite similar and comparable with the literature (Glanville et al., 2012). As far as substance-specific differences in incorporation into EMB were the most visible only at day 10 (Supplementary Table 3.), when the total amount of incorporated C decreased, the long-term fate of LMWOS-C in soils are caused by the metabolic pathways of LMWOS classes within microbial cells and not by rapid LMWOS uptake during the first few days. Therefore, we can consider the medium-term divergence of C depending on its initial form that entered the soil. 2.1.4.2 Microbial community composition The constant composition of PLFAs (Table 1) after the addition of very low amount of LMWOS-C shows that microbial community structure was under steady-state conditions (Blagodatskaya et al., 2007, , 2009). This corresponds to other studies with similar amounts of applied C (Brant et al., 2006) and leads to 13 C incorporation into microbial biomass and individual microbial groups reflected the typical utilization of these substances under natural soil conditions – i.e. a microorganisms under maintenance metabolism. The main classes of decomposers for the six substances were Gand G+ bacteria. Gbacteria are very common in the rhizosphere, which reflects their preference for LMWOS common in rhizosphere hotspots. In contrast, G+ bacteria are abundant in bulk soil (Soderberg et al., 2004). The soil environment of this study, with aerobic conditions, a neutral soil pH as well as the aboveand belowground litter remaining after the harvest, provide optimum conditions for the development of actinomycetes, which is assumed to be important for the primary degradation of recalcitrant SOM (McCarthy and Williams, 1992). In contrast, present environmental conditions and loadings of small amounts of complex substrates did not supported fungal growth (Reischke et al., 2014), which explains their low abundance as well as their low activity in LMWOS utilization in our experiment. Publications and Manuscripts 63 We detected relevant amounts of 16:1w5 fatty acid, which can be used to characterized the VAM fungi or gram-negative bacteria (Olsson, 1999; Zelles, 1997). Results of the factor analysis do not attribute the 16:1w5 to the group 1 or group 2 of Gbacteria, moreover they were loaded up similarly with fungi. Second the VAM are usually abundant in soils, because they form a symbiotic relationship with up to 80% of land plants (Madan et al., 2002). Third 16:1w5 behaved similarly with fungi in utilization of investigated LMWOS (especially for carboxylic acids). All these factors support the interpretation that 16:1w5 reflect VAM in this soil. For ensuring the interpretation of 16:1w5 as VAM fatty acid, the simultaneous analysis of 16:1w5 in PLFAs and neutral lipids should be done. As far as we didn’t measure this in our experiment, the interpretation of 16:1w5 as VAM is not assured. 2.1.4.3 Incorporation of LMWOS into PLFAs Amino acids The observed dominant role of bacteria in amino acid utilisation is in agreement with previous studies, which found that the relative incorporation of 13 C from glutamate (the added amount was 50 µg C g -1 soil) into bacteria was high, whereas incorporation into fungi was significantly lower (Brant et al., 2006; Rinnan and Bååth, 2009). Actinomycetes utilised amino acids in a forest soil, similar to G+ bacteria (Brant et al., 2006). Our results with agricultural soil support those of Brant et al. (2006) and confirm the theory that amino acid turnover is mainly controlled by microbial activity and not by microbial community structure (Jones et al., 2005). We demonstrated a preferred incorporation of alanine than glutamate into PLFAs and a higher replacement of PLFA-C by alanine 13 C than glutamate 13 C (Fig. 2, bottom). Despite a similar uptake of alanine and glutamate into EMB, their contrasting incorporation into PLFAs shows differences in intracellular metabolisation: alanine C is directly converted to acetyl-CoA, the direct precursor of fatty acids, whereas complex energyconsuming pathways are needed to transform glutamate C into acetyl-CoA (Fig. 5). In addition, the formation of acetyl-CoA from alanine causes a loss of one-third of its C backbone compared to a three-fifths loss of C from glutamate if converted to acetyl-CoA (Apostel et al., 2013). This also contributes to the lower incorporation of glutamate-C into PLFAs. Thus, our results confirm those of previous studies on metabolic tracing, showing that intracellular metabolisation is the master process that determines the fate of amino acid C in soils (Apostel et al., 2013; Dippold and Kuzyakov, 2013; Knowles et al., 2010). Publications and Manuscripts 64 Sugars The preference of bacteria for glucose utilization compared to fungi, corresponds with the dominance of bacteria within the soil microbial community, but can also be attributed to the more efficient uptake of LMWOS by bacteria (Moore et al., 2005) especially if low concentrations of LMWOS are applied (Reischke et al., 2014). This was revealed previously by a higher relative glucose incorporation into bacteria (Brant et al., 2006). The preferential incorporation of sucrose into bacterial fatty acids (16:1 ω 7 and 18:1 ω 7) was also reported by Nottingham et al. (2009), who noted the importance of the 16:1w7 biomarker in the control of priming effects. Thus, Gbacteria (corresponding to our G-1 group) represent a group whose growth is based on easily available substrates and which are the most competitive for LMWOS in many ecosystems (Treonis et al., 2004). Hence, the majority of studies show that bacteria are the most relevant group for the uptake and degradation of easily available substrates e.g., following the initial stage of litter decomposition, whereas fungi decompose more complex substrates that remain at later stages (Moore-Kucera and Dick, 2008). However, studies based on nuclear magnetic resonance showed the significant utilization of 13 C from glucose for the formation of unsaturated triacylglycerols, typical storage metabolites of eukaryotes (Lundberg et al., 2001). Based on these results, it has been suggested that fungi are the most active organisms in glucose degradation. Interpretations in our study are based on membrane lipids – a substance class whose structure, function and biosynthetic pathways are similar between many prokaryotes and eukaryotes. Thus, a comparison of the utilization pattern is probably more reliable if functional and biosynthetically comparable compounds are included (Rinnan and Bååth, 2009). The uptake pattern of ribose was relatively similar to that of glucose (Fig. 3, top), with predominant utilization by Gbacteria. This primary incorporation of pentose by Gbacteria was also characterised by 13C-xylose utilization (Waldrop and Firestone, 2004) a similar structure and thus presumably a similar uptake and metabolism. The high percentage incorporation of ribose 13 C into EMB compared to the relatively small amount of 13 C detected in PLFAs can be explained by the use of ribose for the synthesis of other cell polymers. After modification in the pentose-phosphate pathway and phosphorylation, ribose is likely to become a subunit for ribonucleotides and less used for fatty acid biosynthesis (Fig. 5). Ribonucleotides are extracted after chloroform fumigation and this can explain high 13 C incorporation in the microbial pool. Only G-1, the most active group in LMWOS utilization, incorporated high amounts of ribose into PLFAs, i.e. pentosephosphate intermediates. This accounts for high intracellular turnover of this most active microbial group. Publications and Manuscripts 65 Carboxylic acids Acetate is a ubiquitous substrate in soil: it is the main product of lipid degradation, the main substance of plant litter anaerobic decomposition (Reith et al., 2002), present at high concentrations in slurry (Laughlin et al., 2009) and is known as direct precursor role for the formation of fatty acids . The amount of acetate incorporated into membranes of Gbacteria 1 (16:1w7c and 18:1w7c) was 5-fold higher than for most of the other PLFAs (Fig. 4, top). Similar to the other LMWOS, this might be a result of their higher abundance within the microbial community and their rapid uptake of LMWOS. A similar high recovery of 13C from acetate in 16:1w7c and 18:1w7c PLFAs was reported in experiments with anoxic brackish sediment (Boschker et al., 2001). Our experiment with well-aerated agricultural soil showed that the high competitiveness of these Gbacteria for acetate does not depend on the oxygen supply. In addition, experiments with sediment and groundwater samples showed that only few genera were involved in acetate degradation (Pombo et al., 2005). The 16:1w7c PLFA has been suggested as a biomarker for acetate-oxidising sulphate-reducing bacteria. This anaerobic degradation can only occur in O 2 -deficient microhabitats such as aggregate cores and is very unlikely to play a relevant role in freshly tilled soil. Fatty acids that characterise G+ bacteria (such as i15:0, i16:0, i17:0, a17:0) were also enriched, but to a much lower degree than Gbacterial fatty acids (Fig. 4, top). Similar results were obtained with anoxic brackish sediments, where 10Me16:0, cy17:0, i15:0 and a15:0 PLFAs were enriched in 13 C from acetate and were related to sulphatereducing bacteria (Boschker et al., 2001). Carboxylic acids were the only substrate class where fungal uptake and incorporation could compete with those of prokaryotic, G+ groups. Thus, although fungi are less competitive for the most LMWOS, they prefer acidic substrates within the LMWOS (Rinnan and Bååth, 2009). This correlates with their preference for acid soil conditions, where acidic (non-neutralised), more complex substrates, dominate (Haider, 1996). A comparison of palmitate and acetate utilization in soils is important because acetate is a direct microbial precursor for palmitate synthesis. There are three pathways for the incorporation of palmitate into phospholipids: 1) partial step-by-step degradation of C2-units without total breakdown of palmitate can occur and subsequently, only parts of the molecule are used for further biosynthesis (Rhead et al., 1971); 2) the resynthesis pathway includes the complete degradation of the molecule to acetyl-CoA and the following synthesis of new fatty acids by a series of enzymatic reactions (Rhead et al., 1971); 3) the alternative is the utilization of palmitate directly without further transformation, because it is the most abundant fatty acids in microorganisms and might be slightly modified by elongation or desaturation. According to the first pathway, palmitate should have a Publications and Manuscripts 72 Supplementary Data Supplementary Table A1: Fatty acids in the external standard Supplementary Table A1: Results of factor analysis: Factor loadings and grouping of fatty acids derived from factor loadings and PLFAs literature. Publications and Manuscripts 73 Supplementary Table A3: Nested ANOVA between classes of LMWOS and single substances nested in class of LMWOS for soil, microbial biomass and PLFAs. Degrees of freedom (df), values (F) and significance level (p) are shown for the two time points. Publications and Manuscripts 74 2.2 Study 2: Improved δ 13 C analysis of amino sugars in soil by Ion Chromatography – Oxidation – Isotope Ratio Mass Spectrometry SHORT TITLE: Amino sugar δ 13 C analysis by Ion Chromatography - Isotope Ratio Mass Spectrometry Michaela A. Dippold 1,2,3 , Stefanie Boesel 2 , Anna Gunina 1,3 , Yakov Kuzyakov 3,4 , Bruno Glaser 2 1 Department of Agroecosystem Research, BayCEER, University of Bayreuth 2 Department of Soil Biochemistry, Institute of Agricultural and Nutritional Science, Martin-Luther University Halle-Wittenberg 3 Department of Agricultural Soil Science, University of Göttingen, Germany 4 Department of Soil Science of Temperate Ecosystems, Georg-August University of Göttingen Corresponding Author: Michaela Dippold Department of Agricultural Soil Science Georg-August University of Goettingen Buesgenweg 2 97077 Goettingen email: [email protected] Tel.: 0921/552187 Fax.: 0921/552246 Publications and Manuscripts 75 Abstract Rationale: Amino sugars build up microbial cell walls and are important compounds of soil organic matter. To evaluate their sources and turnover, δ 13 C analysis of soil-derived amino sugars by liquid chromatography was recently suggested. However, amino sugar δ 13 C determination remains challenging due to 1) a strong matrix effect, 2) CO 2 -binding by alkaline eluents, and 3) strongly different chromatographic behavior and concentrations of basic and acidic amino sugars. To overcome these difficulties we established an ion chromatography-oxidation-isotope ratio mass spectrometry method to improve and facilitate soil amino sugar analysis. Method: After acid hydrolysis of soil samples, the extract was purified from salts and additional components impeding chromatographic resolution. Amino sugar concentrations and δ 13 C values were analyzed by coupling an ion chromatograph to an isotope ratio mass spectrometer. The accuracy and precision of quantification and δ 13 C determination were assessed. Results: Internal standards enabled correction for losses during analysis, with a relative standard deviation < 6%. The higher magnitude peaks of basic compared to acidic amino sugars required an amount-dependent correction of δ 13 C values. This correction allowed to decrease the accuracy of δ 13 C determination of < 1.5‰ and their precision of < 0.5‰ for basic and acidic amino sugars in a single run. Conclusion: This method enables parallel quantification and δ 13 C determination of basic and acidic amino sugars in a single chromatogram due to the advantages of coupling an ion chromatograph to the isotope ratio mass spectrometer. Small adjustments of sample amount and injection volume are necessary to optimize precision and accuracy for individual soils. Keywords : compound-specific isotope ratio mass spectrometry, amino sugars, soil organic matter, microbial biomarker analysis, ion chromatography, IC-O-IRMS coupling. Publications and Manuscripts 76 2.2.1 Introduction The great relevance of microbial compounds within soil organic matter (SOM) became evident within the last decade. Microbial cell wall compounds seem to be the most relevant microbial-derived compound class within slow cycling SOM, as they 1) are highly polymeric substances (Amelung, 2003) and 2) stabilized by interaction with soil surfaces (Amelung et al., 2001; Miltner et al., 2011). Thus, an increasing interest arose to investigate their turnover and accumulation in soils (Miltner et al., 2011). Beside its contribution to the soil organic C (SOC) pool, amino sugars are - together with proteins – the compound classes linking the C and N cycles in soil and contribute significantly to the soil organic N (Amelung, 2003). In addition, amino sugars provide information about the microbial community structure. Bacterial cell walls consist of peptidoglycan – a polymer of N-acetylmuramic acid and N-acetylglucosamine whereas fungal cell walls consist of chitin, a N-acetylglucosamine polymer (Engelking et al., 2007; Glaser et al., 2004). The origin of mannosamine and galactosamine, additional amino sugars found in hydrolysis extracts of soils, are still debated. In contrast to cell membrane compounds like phospholipids, which turn over rapidly in soils (Rethemeyer et al., 2004), amino sugars are more stable. Contribution of living biomass versus necromass in soils (Glaser et al., 2004) or fungal and bacterial biomass (Joergensen and Wichern, 2008) as well as reliable and generally accepted results on their turnover time in soils are still rare (Amelung et al., 2008; Glaser, 2005) as no methods for 14 C measurements of amino sugars, neither in their natural abundance nor 14 Clabeled, have been reported in the literature to our knowledge. Recent approaches have focused on determinations of δ 13 C or δ 15 N values of amino sugars. These studies started from the quantification of amino sugars by gas chromatography (Guerrant and Moss, 1984; He et al., 2006; Zhang and Amelung, 1996) and continued with gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS) (Glaser and Gross, 2005). However, δ 13 C-determination by GC-C-IRMS has aggravating shortcomings (Decock et al., 2009): 13 C fractionation occurs during measurement; the resulting offset and amount dependence of the isotope signal can in part be corrected by the use of external standard (Glaser and Amelung, 2002; Schmitt et al., 2003). However, the greater the amount of introduced derivative C compared to C atoms of interest, the larger the error in the  13 C determination that still remains after applying correction functions (Decock et al., 2009; Gross and Glaser, 2004). As amino sugars are water-soluble low molecular weight organic substances, they can also be quantified by high performance liquid chromatography (HPLC) (Appuhn et al., 2004; Indorf et al., 2011). Therefore, current methodological developments have fo- Publications and Manuscripts 77 cused on the establishment of liquid chromatography-oxidation-isotope ratio mass spectrometry (LC-O-IRMS) methods (Krummen et al., 2004) for δ 13 C measurement of amino sugars (Bode et al., 2009), which have already revealed its high potential for application in soil science (Bai et al., 2013; Bode et al., 2013; Indorf et al., 2012). Many LC-O-IRMS methods and in particular the amino sugar method are not routinely used. Conventional liquid chromatographs are constructed for organic eluents and problems occur if continuously used with strong acids or bases. However, performing LCO-IRMS analysis for δ 13 C determination does not allow any organic eluents i.e. organic C. Hence, liquid chromatography is restricted to ion exchange columns which implies the use of salt solutions or acids and bases as eluents (Basler and Dyckmans, in press; Bode et al., 2009). Thus, metallic ions can be dissolved from stainless steel pumps or capillaries and salt crystallization occurs within the system (Bode et al., 2009; Rinne et al., 2012). This causes a loss in the performance of the columns as well as blockages of the system. To prevent such problems, time and money consuming purging steps have to be implemented between sample measurements (Bode et al., 2009; Rinne et al., 2012). In addition, any contamination by HCO 3has to be avoided for δ 13 C determination as HCO 3increases C background (i.e. baseline) and it will influence the δ 13 C value of the analytes. However, liquid chromatographs are per se not constructed to avoid gas diffusion into the system. Thus, pre-degassing of eluents have to be performed to enable carbonate-free chromatography – especially if bases are used as eluents. In addition, basic amino sugars (glucosamine, galactosamine and mannosamine) show greatly different chromatographic behavior than the acidic muramic acid. Thus, a high gradient with the eluents has to be driven, leading to strong elution of the matrix, especially for soils (Bode et al., 2009). In addition, concentrations of muramic acid are ten to hundred times lower than those of basic amino sugars. This hampers quantification due to the limited linear range of the detectors as well as δ 13 C determination due to a limited range of peak area with reproducible results. Therefore, current methods use a double measurement with different chromatographies to first measure muramic acid and afterwards basic amino sugars (Bode et al., 2009). This double measurement as well as the additional effort required for solvent-free HPLC methods renders routine measurement of δ 13 C values of amino sugars nearly impossible. The aim of this study was to establish an ion chromatography–oxidation–isotope ratio mass spectrometry (IC-O-IRMS) method for quantification and δ 13 C determination of soil-derived amino sugars. We hypothesized that using an ion chromatograph would strongly facilitate IRMS measurement of many biomarkers, as some basic requirements like carbonate-free measurement or metal-free systems are already fulfilled by the instrument. In addition, we intended to optimize amino sugar purification to reduce cationic Publications and Manuscripts 78 contamination and matrix peaks originating from soil. The aim was to provide a method enabling a routine application of δ 13 C amino sugar measurements, which are crucial regarding the increasing interest in microbial contributions to stable SOM. 2.2.2 Material and Methods 2.2.2.1 Soil Topsoil (0-10 cm) from the Ap horizon of a silt loamy haplic Luvisol (WRB, 2006) was collected from a long-term cultivated field in Bavaria (49.907 N, 11.152 E, 501 m. a. s. l, mean annual temperature 6-7 °C, mean annual precipitation 874 mm). The soil had a pH KCl of 4.88 and pH H2O of 6.49, TOC and TN content were 1.77% and 0.19%, respectively, and potential cation exchange capacity was 13.6 cmol c kg -1 . Field fresh soil was sieved to 2 mm and all roots were removed with tweezers. Soil was then freeze dried, ball milled and 500 mg of the resulting powder were used for each hydrolysis. 2.2.2.2 Chemicals, reagents and external and internal standards All chemicals for hydrolysis and purification were obtained from Sigma-Aldrich (St. Louis, MO, USA) with a minimum grade of “pro analysis” (>99.0% purity). For ion chromatography, a 50-52%, ultra-pure NaOH solution was purchased from Sigma Aldrich (St. Louis, MO, USA). NaNO 3 -solution (0.01 M) was produced from metal-free sodium nitrate, puratronic (99.999% purity, Alfa Aesar, Karlsruhe, Germany). For oxidation, a 0.26 M sodium persulfate solution and 10% phosphoric acid solutions were used (Sigma Aldrich, St. Louis, MO, USA). Methylglucamine p. a. (5 mg mL -1 ) and fructose p.a. (1 mg mL -1 ) (Sigma Aldrich, Louis, MO, USA) were used as the first and second internal standards (IS1 and IS2), respectively. Stock solutions for external standards contained methylglucamine, glucosamine, mannosamine and galactosamine at concentrations of 5, 14, 1.5 and 20 mg l -1 (Sigma Aldrich, Louis, MO, USA) and muramic acid (Toronto Research Chemicals Inc., Toronto, Canada) at 7.5 mg l -1 . The IAEA-calibrated δ 13 C value of each external standard was determined by repeated Elemental Analyzer-Isotope Ratio Mass Spectrometry (Flash 2000 HT Plus Elemental Analyzer and Delta V Advantage Isotope Ratio Mass Spectrometer, both from Thermo-Fisher, Bremen, Germany) measurement of these substances and calibrated against certified standards of the International Atomic Energy Agency IAEA (IAEA-CH6: -10.4‰, IAEA-CH7 -31.8‰ and USGS41 37.8‰) versus Pee Dee Belemnite (PDB). Publications and Manuscripts 79 2.2.2.3 Soil hydrolysis and ion removal Soil hydrolysis and ion removal were performed according to Zhang and Amelung (1996), which was optimized for δ 13 C determination by Glaser and Gross (2005). Briefly, hydrolysis was performed with 10 mL of 6 M HCl at 105 °C for 8 h. The filtrate extract was dried completely and redissolved in 20 mL H 2 O. One hundred microliters of the IS1 methylglucamine (i.e. 50 µg) were then added. The pH was adjusted to 6.6-6.8 with 0.6 M KOH and precipitated iron was removed by centrifugation (4000 rpm for 15 min). After freeze-drying the residue was redissolved in 5 mL of dry methanol and salt precipitates were removed by centrifugation (4000 rpm for 10 min). The supernatant was dried under a gentle stream of N 2 and stored frozen until column purification. 2.2.2.4 Purification by cation exchange column Liquid chromatography requires a column purification to remove hydrolysable noncationic compounds like monosaccharides and carboxylic acids from the extract. A cation exchange column (AG 50W-X8 Resin, H + form, mesh size 100-200, Biorad, Munich, Germany) was used as suggested by Indorf et al. (2013): a thin layer of clean glass wool was installed under 4 cm of cation exchange resin in the glass column (inner diameter: 0.8 cm). Resin was filled in by rinsing with ~10 mL of 0.1 M HCl solution to ensure the H + form of the sorbent, covered with a thin layer of glass wool and preconditioned with 5 mL of water. Dried extracts were redissolved in ~1 mL of water with one drop of 0.1 M HCl to ensure the cationic form of muramic acid. After transferring the sample onto the column, neutral and anionic compounds were eluted with 8 mL water. The cationic fraction containing the amino sugars was eluted by 15 mL 0.5 M HCl, freeze-dried and transferred with 5 mL of dry methanol. After evaporation of the methanol by a gentle stream of dried N 2 , the sample can be stored frozen (-20 °C) for at least one month. For subsequent measurement, the samples were re-dissolved in 200 µL water with the addition of 50 µL of IS2 solution and measured within 24 hours after re-dissolving. 2.2.2.5 Development of the measurement by IC-O-IRMS All measurements were performed by a Dionex ICS-5000 SP ion chromatography system coupled by an LC IsoLink to a Delta V Advantage Isotope Ratio Mass Spectrometer (Supplementary Figure 1) (all components from Thermo-Fisher, Bremen, Germany). Chromatographic conditions were modified and optimized with the aim of reaching baseline separation and a resolution factor Rs greater than 1. Publications and Manuscripts 80 ( ) 12 12 ww5.0 tt Rs +⋅ − = equation 1 t 1 and t 2 are the retention times of two neighboring peaks and w represents their respective peak width at the tangents baseline (Figure 1). Nine microliters of the water-dissolved sample or external standard were injected via a 25 µL injection loop and the injection time was defined as 0 sec. Chromatography was performed by a CarboPac TM PA 20 analytical anion exchange column (3 x 150 mm, 6.5 µm) which was preceded by a PA 20 guard column (Bode et al., 2009) (both from Dionex, Amsterdam, The Netherlands). The elution sequence contained a preconditioning before injection (15 min with 200 mM NaOH and 10 min with 8 mM NaOH). Elution sequence lasted for 35 min in total and was performed at constant temperature of 30 °C and a flow rate of 0.4 ml min -1 . 8 mM NaOH was increased after 11 min to 8 mM NaOH with a pulse of 2.5 mM NaNO 3 until 15 th minute. Then NaNO 3 was decreased and NaOH concentration increased for final 20 min of chromatogram (details in Supplementary Table 1). We measured external standards at four concentrations (e.g. 50, 100, 175 and 250 µL of the stock solution) at least once before and once after a sample batch. A sample batch consisted of 4–6 samples, each measured 4 times. A sample batch was always measured once in its entirety and then repeated three times. Integration was performed by Isodat 3.0 (Thermo Fisher Scientific, Bremen, Germany) with the following parameters: start slope 1 mV/s, end slope 2 mV/s, peak min 50 mV, peak resolution 50% and an individual background. Publications and Manuscripts 81 Fig. 1 Chromatogram of external standard (top) and un-spiked sample (down). First and second internal standards as well as basic amino sugars (galactosamine, mannosamine and glucosamine) and acidic muramic acid are marked. Peak resolution Rs is included for the triplett of basic amino sugars in the upper chromatogram of the external standard and Rs for muramic acid and its preceeding matrix peak is shown in the chromatogram of the sample. 2.2.2.6 Evaluation of amino sugar quantification via IC-O-IRMS To validate the method by standard addition, the standard mixture serving as external standards was added to the hydrolysis extracts. The amounts of substance added were in the range of 0, 1.3, 1.7, 2.1 and 3 times of the expected concentrations. The data from the standard addition experiment were statistically evaluated according to Birk et al. (2012): For each substance (including IS1) a linear regression was fitted by the method of least squares to the measured amounts as a function of the added amounts per sample (Supplementary Figure 2). The y-intercept represented the fitted amount of substance in soil and the slope gave the mean recovery of a substance. The significance of regression was tested and Steven’s Runs Test was performed to identify deviations from linearity. Significant differences between recoveries were detected by covariance analysis (ANCOVA) of the slopes. All regression parameters were calculated Publications and Manuscripts 88 LoQ without having an overload in the glucosamine peak. Under such special conditions a double measurement with high concentrated sample for determination of muramic acid δ 13 C values and diluted concentration for determination of glucosamine δ 13 C values might be necessary. Average amino sugar δ 13 C values differ for ~0.1 to 1.1‰ within the basic (Bode et al., 2009; Bode et al., 2013) and for more than 3-5‰ between the basic and acidic amino sugars (Bode et al., 2009; Bode et al., 2013; Glaser, 1999) and differ for around 7‰ from bulk SOC (Glaser and Gross, 2005). The achieved accuracies of individual amino sugars enable to distinguish amino sugar from their C sources even under natural abundance conditions. Resulting precisions (0.5‰) are lower than differences between basic and acidic amino sugars and consequently enables to identify microbial group specifics in amino sugar formation (e.g. specifics in the used substrates or the fractionations in biochemical formation pathways). Especially in experiments leading to a higher δ 13 C differences in amino sugars like C3 to C4 C source changes (Indorf et al., 2012), FACE experiments (Glaser and Gross, 2005) or application of labeled substrates (Bode et al., 2013) this method can fully distinguish C sources and individualities in cell wall formation of fungi and bacteria. In summary, this method enables a combined determination of δ 13 C values of amino sugars for the majority of soils. However, adjustments to new sample types are necessary to identify the optimum amount of sample to hydrolyze or the final volume to inject so that the optimum range for accuracy and precision of the δ 13 C values are met. Fig. 3 Amount-dependent function for estimation of standard error of δ 13 C (δ final (Ai) determination calculated according to equation 6. Publications and Manuscripts 89 2.2.3.5 Advantages of IC-O-IRMS Many previous studies reported severe problems with LC-O-IRMS, e.g. the impossibility of measuring muramic acid in non-spiked samples due to very low peak areas or the requirement for time-consuming purging steps to maintain performance of the PA 20 column (Bode et al., 2009). The absence of these issues in the currently proposed method can mainly be attributed to the advantages of IC over HPLC. Ion chromatographs are free of metals: all elements that are in contact with sample or eluents are made from polyether ether keton (peek). Thus, metal contamination can originate only from the sample. However, our method contains iron and salt precipitation steps, removing all (potentially column destroying) cations. This not only reduces measurement time but also reduces costs as, e.g., in-line high pressure filters protecting the column from colloids and metal ions are not needed. Even after 600 injections, no decrease in performance of the PA 20 column was detected and the pre-column did not need to be exchanged. In addition, the CO 2 -tight construction of Ion Chromatographs is a great advantage of δ 13 C determination as no shifts in the δ 13 C value due to increasing carbonate background occurred. Therefore, even CO 2 -binding eluents, like NaOH, do not cause problems for chromatography and isotope ratio mass spectrometry. In addition, Ion Chromatographs are routinely equipped with a degasser, which keeps the eluents and oxidizing reagents of the Isolink CO 2 -free. Thus, although acquisition costs may be higher, the improved performance, higher sample throughput and lower follow-up costs reflect the clear advantages of ion chromatographs for improving LC-O-IRMS. 2.2.4 Conclusions Amino sugars are important biomarkers for research on bacterial and fungal contribution to SOM. This new method enables parallel quantification and δ 13 C determination of the most frequent amino sugars in soils and thus sets the preconditions for wider adoption of δ 13 C amino sugar determination in soil science. The combination of iron and salt removal from gas chromatography protocols with purification via cation exchange resins adapted from liquid chromatography methods proved to be an optimal sample preparation for ion chromatography including chromatographic separation, system stability and longevity of system components. In addition, using ion chromatograph sets clear advantages over HPLCs as metal and carbonate exclusion from the system avoids column contamination as well as disturbance of δ 13 C determination by a carbonate background. Publications and Manuscripts 90 These improvements over previous methods enabled parallel quantification and δ 13 C determination of high-concentrated basic amino sugars and low-concentrated muramic acid. Recoveries ranged from 57 to 66% and could be corrected by using methylglucamine as the first internal standard. The quantification limit of muramic acid, the compound with the lowest concentration, was around 0.05 mg per vial for quantification and for isotope measurement. When muramic acid exceeded these values, glucosamine, the most concentrated compound, was still in a linear range for quantification and δ 13 C measurement. The accuracy of IC-O-IRMS was better than 1‰ for basic amino sugars and better than 1.5‰ for muramic acid compared to calibrated EA-IRMS values. Precision was amount-dependent and less than 0.5‰ over a comparatively broad range of areas. However, the dependence on the matrix and the ratio of muramic acid to glucosamine in individual samples necessitates adjustment in soil amount or injection volume to achieve the optimal accuracy and precision of δ 13 C. The quality of the quantification and δ 13 C determination as well as sample throughput of this method should enable this method to be used routinly in soil science. The advantages of IC-O-IRMS compared to HPLC-O-IRMS are evident and might also bring advantages for analysis of other biomarkers. Acknowledgments We thank the DFG for financing the IC-O-IRMS instrument and the project DFG KU 1184 19/1. Publications and Manuscripts 91 Reference List Amelung, W., 2003. Nitrogen biomarkers and their fate in soil. Journal of Plant Nutrition and Soil Science 166, 677-686. Amelung, W., Brodowski, S., Sandhage-Hofmann, A., Bol, R., 2008. Combining biomarker with stable isotope analyses for assessing the transformation and turnover of soil organic matter, Advances in Agronomy, Vol 100, 155-250. Amelung, W., Miltner, A., Zhang, X., Zech, W., 2001. Fate of microbial residues during litter decomposition as affected by minerals. Soil Science 166, 598-606. Apostel, C., Dippold, M., Glaser, B., Kuzyakov, Y., 2013. Biochemical pathways of amino acids in soil: Assessment by position-specific labeling and C-13-PLFA analysis. Soil Biology & Biochemistry 67, 31-40. Appuhn, A., Joergensen, R.G., Raubuch, M., Scheller, E., Wilke, B., 2004. The automated determination of glucosamine, galactosamine, muramic acid, and mannosamine in soil and root hydrolysates by HPLC. Journal of Plant Nutrition and Soil Science 167, 17-21. Bai, Z., Bode, S., Huygens, D., Zhang, X., Boeckx, P., 2013. Kinetics of amino sugar formation from organic residues of different quality. Soil Biology & Biochemistry 57, 814821. Basler, A., Dyckmans, J., in press. Compound-specific delta C-13 analysis of monosaccharides from soil extracts by high-performance liquid chromatography/isotope ratio mass spectrometry. Rapid Communications in Mass Spectrometry DOI: 10.1002/rcm.6717. Birk, J.J., Dippold, M., Wiesenberg, G.L.B., Glaser, B., 2012. Combined quantification of faecal sterols, stanols, stanones and bile acids in soils and terrestrial sediments by gas chromatography-mass spectrometry. Journal of Chromatography A 1242, 1-10. Bode, S., Denef, K., Boeckx, P., 2009. Development and evaluation of a highperformance liquid chromatography/isotope ratio mass spectrometry methodology for delta(13)C analyses of amino sugars in soil. Rapid Communications in Mass Spectrometry 23, 2519-2526. Bode, S., Fancy, R., Boeckx, P., 2013. Stable isotope probing of amino sugars - a promising tool to assess microbial interactions in soils. Rapid Communications in Mass Spectrometry 27, 1367-1379. Decock, C., Denef, K., Bode, S., Six, J., Boeckx, P., 2009. Critical assessment of the applicability of gas chromatography-combustion-isotope ratio mass spectrometry to determine amino sugar dynamics in soil. Rapid Communications in Mass Spectrometry 23, 1201-1211. Engelking, B., Flessa, H., Joergensen, R.G., 2007. Shifts in amino sugar and ergosterol contents after addition of sucrose and cellulose to soil. Soil Biology & Biochemistry 39, 2111-2118. Glaser, B., 1999. Eigenschaften und Stabilität des Humuskörpers der "Indianerschwarzerden" Amazoniens. Lehrstuhl für Bodenkunde und Bodengeographie. Glaser, B., 2005. Compound-specific stable-isotope (delta C-13) analysis in soil science. Journal of Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 168, 633-648. Glaser, B., Amelung, W., 2002. Determination of C-13 natural abundance of amino acid enantiomers in soil: methodological considerations and first results. Rapid Communications in Mass Spectrometry 16, 891-898. Glaser, B., Gross, S., 2005. Compound-specific delta C-13 analysis of individual amino sugars - a tool to quantify timing and amount of soil microbial residue stabilization. Rapid Communications in Mass Spectrometry 19, 1409-1416. Glaser, B., Turrion, M.B., Alef, K., 2004. Amino sugars and muramic acid - biomarkers for soil microbial community structure analysis. Soil Biology & Biochemistry 36, 399-407. Publications and Manuscripts 92 Gross, S., Glaser, B., 2004. Minimization of carbon addition during derivatization of monosaccharides for compound-specific delta C-13 analysis in environmental research. Rapid Communications in Mass Spectrometry 18, 2753-2764. Guerrant, G.O., Moss, C.W., 1984. Determination of Monosaccharides as Aldononitrile, O-Methyloxime, Alditol, and Cyclitol Acetate Derivatives by Gas-Chromatography. Analytical Chemistry 56, 633-638. He, H.B., Xie, H.T., Zhang, X.D., 2006. A novel GUMS technique to assess N-15 and C13 incorporation into soil amino sugars. Soil Biology & Biochemistry 38, 1083-1091. Indorf, C., Bode, S., Boeckx, P., Dyckmans, J., Meyer, A., Fischer, K., Jörgensen, R.G., in press. Comparison of HPLC Methods for the Determination of Amino Sugars in Soil Hydrolysates. Analytical Letters 46, 2145-2164. Indorf, C., Dyckmans, J., Khan, K.S., Joergensen, R.G., 2011. Optimisation of amino sugar quantification by HPLC in soil and plant hydrolysates. Biology and Fertility of Soils 47, 387-396. Indorf, C., Stamm, F., Dyckmans, J., Joergensen, R.G., 2012. Determination of saprotrophic fungi turnover in different substrates by glucosamine-specific delta C-13 liquid chromatography/isotope ratio mass spectrometry. Fungal Ecology 5, 694-701. Joergensen, R.G., Wichern, F., 2008. Quantitative assessment of the fungal contribution to microbial tissue in soil. Soil Biology & Biochemistry 40, 2977-2991. Krummen, M., Hilkert, A.W., Juchelka, D., Duhr, A., Schluter, H.J., Pesch, R., 2004. A new concept for isotope ratio monitoring liquid chromatography/mass spectrometry. Rapid Communications in Mass Spectrometry 18, 2260-2266. Miltner, A., Bombach, B., Schmidt-Brücken, B.K., M., 2012. SOM genesis: microbial biomass as a significant source. Biogeochemistry 111, 41-55. Rethemeyer, J., Kramer, C., Gleixner, G., Wiesenberg, G.L.B., Schwark, L., Andersen, N., Nadeau, M.J., Grootes, P.M., 2004. Complexity of soil organic matter: AMS C-14 analysis of soil lipid fractions and individual compounds. Radiocarbon 46, 465-473. Rinne, K.T., Saurer, M., Streit, K., Siegwolf, R.T.W., 2012. Evaluation of a liquid chromatography method for compound-specific delta C-13 analysis of plant carbohydrates in alkaline media. Rapid Communications in Mass Spectrometry 26, 2173-2185. Schmitt, J., Glaser, B., Zech, W., 2003. Amount-dependent isotopic fractionation during compound-specific isotope analysis. Rapid Communications in Mass Spectrometry 17, 970-977. WRB, I.W.G., 2006. World Reference Base for Soil Resources, 2nd ed. FAO, Rome. Zech, M., Glaser, B., 2009. Compound-specific delta O-18 analyses of neutral sugars in soils using gas chromatography-pyrolysis-isotope ratio mass spectrometry: problems, possible solutions and a first application. Rapid Communications in Mass Spectrometry 23, 3522-3532. Zhang, X.D., Amelung, W., 1996. Gas chromatographic determination of muramic acid, glucosamine, mannosamine and galactosamine in soils. Soil Biology & Biochemistry 28, 1201-1206. Publications and Manuscripts 93 Supplementary Data Figure Supplementary A1: Scheme of the instrument coupling: Ion Chromatograph is shown on the left side with pump, autosampler and detector-chromatography compartment. Connection to isolink occurs via a peek capillary with interposed colloid filter. Scheme of LC Isolink is adapted from Krummen et al. (2004). Figure Supplementary A2: Standard addition line of the quantified amino sugars: quantified amount per g soil is plotted against the amount of spiked amino sugar. Slope represents recovery of the individual analytes and y-axis gap represents soil content without recovery correction. Regression parameters are shown in Table 2. Publications and Manuscripts 94 Supplementary Figure A3: measured δ 13 C values of spiked samples are plotted against the percent of peak area, which is derived from the added standard: y-intercept of the fitted linear regression reflects the fitted value of soil whereas δ 13 C-value at 100% standard reflects the δ 13 C value of the added standard substance Figure Supplementary A4: area-dependant error terms of equation 6: left side shows the standard error of the measurement repetition of soil samples and right side shows the area-dependant error of the calibration/correction function from the external standard line Publications and Manuscripts 95 Supplementary Table A1: Solvent gradient and flow conditions of the IC-O-IRMS system time 20 mM NaOH 200 mM NaOH H 2 O 0.01 M NaNO 3 flow (ml min -1 ) -25 min 0% 100% 0% 0% 0.400 -10 min 8% 0% 92% 0% 0.325 11 min 40% 0% 35% 25% 0.400 15 min 45% 0% 45% 10% 0.400 18 min 25% 25% 50% 0% 0.380 35 min standby Publications and Manuscripts 96 2.3 Study 3: Biochemical pathways of amino acids in soil: Evaluation by position-specific labeling and 13 C-PLFA analysis Carolin Apostel 1 , Michaela Dippold 2,3 , Bruno Glaser 4 , Yakov Kuzyakov 1,2 1 Department of Soil Science of Temperate and Boreal Ecosystems, Georg-AugustUniversity Göttingen 2 Department of Agricultural Soil Science, Georg-August-University Göttingen 3 Department of Agroecosystem Research, BayCEER, University of Bayreuth 4 Department of Soil Biogeochemistry, Institute of Agricultural and Nutritional Science, Martin-Luther-University Halle-Wittenberg Corresponding Author: Carolin Apostel Department of Agroecosystem Research University of Bayreuth Universitaetsstrasse 30 95447 Bayreuth email: [email protected] Tel.: 0921/552187 Fax.: 0921/552246 Publications and Manuscripts 97 Abstract Microbial utilization is a key transformation process of soil organic matter (SOM). For the first time, position-specific 13 C labeling was combined with compound-specific 13 C-PLFA analysis to trace metabolites of two amino acids in microbial groups and to reconstruct detailed biochemical pathways. Short-term transformation was assessed by applying position-specifically 13 C labeled alanine and glutamate to soil in a field experiment. Microbial utilization of the amino acids’ functional groups was quantified by 13 C incorporation in total microbial biomass and in distinct microbial groups classified by 13 CPLFA. Loss from PLFAs was fastest for the highly oxidized carboxyl group of both amino acids, whereas the reduced C positions, e.g. C 3-5 , were preferentially incorporated into microorganisms and their PLFAs. The incorporation of C from alanines’ C 2 position into the cell membrane of gram negative bacteria was higher by more than one order of magnitude than into all other microbial groups. Whereas C 2 of alanine was still bound to C 3 at day 3, the C 2 and C 3 positions were partially split at day 10. In contrast, the C 2 of glutamate was lost faster from PLFAs of all microbial groups. The divergence index, which reflects relative incorporation of one position to the incorporation of C from all positions in a molecule, revealed that discrimination between positions is highest in the initial reactions and decreases with time. Reconstruction of microbial transformation pathways showed that the C 2 position of alanine is lost faster than its C 3 position regardless of whether the molecule is used anaor catabolically. Glutamate C 2 is incorporated into PLFAs only by two out of eight microbial groups (fungi and part of gram positive prokaryotes). Its incorporation in PLFA can only be explained by either the utilization of the glyoxolate bypass or the transformation of glutamate into aspartate prior to being fed into the citric acid cycle. During these pathways, no C is lost as CO 2 but neither is energy produced, making them typical C deficiency pathways. Glutamate is therefore a promising metabolic tracer in regard to ecophysiology of cells and therefore changing environmental conditions. Analyzing the fate of individual C atoms by position-specific labeling allows insight into the mechanisms and kinetics of microbial utilization by various microbial groups. This approach will strongly improve our understanding of soil C fluxes. Key words: metabolite tracing, transformation pathways, stable isotope applications, microbial community structure and functions, compound-specific isotope analysis Publications and Manuscripts 104 To correct for amount-dependent 13 C isotopic fractionation during measurements (Schmitt et al., 2003) and for the addition of C during derivatization, linear and logarithmic regressions of the external standards δ 13 C-values to their area were calculated. If both regressions were significant, that with the higher significance was applied. As the δ 13 Cvalue for the derivatizating agents was unknown, the correction was performed according to Glaser and Amelung (2002a) (Eq. 5). %)())(%)(( )( )( %)( ln/ln/ atCtAmatC CN CN atC FSEAlinFAMElinDKFAME FS FAME FS −− ++⋅−⋅= (5) with: C FS (at%) corrected 13 C amount of the fatty acid [at%] C FAME (at%) drift-corrected 13 C amount of the FAME [at%] m lin/ln slope of linear/logarithmic regression [at% · Vs -1 ] t lin/ln y-intercept of linear/logarithmic regression [at%] A FAME area of FAME [Vs] N(C) FAME number of C atoms in FAME N(C) FS number of C atoms in fatty acid C EA-FS (at%) measured 13 C-value of fatty acid [at%] 2.3.2.3 Divergence Index Discrimination of C from individual positions in one molecule during uptake and/or utilization was assessed. The extent of discrimination between pools, microbial groups and at two sampling times was compared as well. For both of these tasks, the differences in absolute uptake into C pools or microbial groups had to be relativized. Therefore, the divergence index (DI) was defined: ∑ ⋅ = n i i i C Cn DI 1 (6) with: n number of C atoms in molecule C i relative incorporation of tracer C [mol · mol -1 ] As required, the DI can be calculated with relative incorporation of tracer per bulk soil, microbial biomass, single PLFA or Σ PLFA of microbial groups. The DI compares the calculated actual incorporation of C from each position with the mean C incorporation from all positions. This can be understood as the result the experiments would have had if uniformly labeled tracers had been used. A DI of 1 would indicate no discrimination Publications and Manuscripts 105 between the positions; values above 1 indicate preferential incorporation, values below 1 show preferential degradation. 2.3.2.4 Statistical analysis For the repetitive measurements of d13C-values, a Nalomov outlier test with significance levels of 95% (when four repetitions were available) or 99% (when three repetitions were available) was performed. PLFAs were classified into corresponding microbial groups by a factor analysis of C contents of the entire dataset. Fatty acids with a loading of more than 0.5 (absolute value) on the same factor were categorized with regard to previous studies (Zelles, 1999; Zelles et al., 1995). All the data presented in this study were tested with a one-way analysis of variance (ANOVA); significances were determined with the Tukey Honest Significance Difference (Tukey HSD) post-hoc test with a significance level of 99.5%. All positions were tested for significant differences between recoveries in soil, microbial biomass and PLFAs. For every microbial group and soil pool, the difference in DI for the six position-specifically labeled positions was also tested for significance. All statistical tests were accomplished with R version 2.9.0 (17.04.2009). 2.3.3 Results 2.3.3.1 Incorporation of uniformly labeled amino acids The C content in the soil was 1230 µmol · g -1 (Table 2), which corresponds to 15.0 mg C · g -1 soil. Of this C, 3.5% is contained in microbial biomass, and 0.01% in the sum of PLFA ( Σ -PLFA). The incorporation of uniformly 13 C-labeled alanine and glutamic acid into soil and microbial biomass decreased between days 3 and 10. The recovery in Σ -PLFA remained stable or even increased. Recovery of applied alanine and glutamic acid C in soil decreased by about half between days 3 and 10. Recovery from applied glutamic acid in microbial biomass decreased by nearly 90% between days 3 and 10 (Table 2). Table 2 Total C content and 13 C incorporation of uniformly labeled amino acids into soil, microbial biomass and sum of PLFA ( Σ -PLFA). Publications and Manuscripts 106 2.3.3.2 Incorporation of position-specifically labeled amino acids With the tool of position-specific labeling, we were able to trace C from individual positions of alanine and glutamic acid into different soil C pools. On day 3 (Fig. 1, top), a clear discrimination against the carboxyl C of both amino acids and glutamic acids aminobound position in soil, microbial biomass and Σ -PLFA is evident. On day 10 (Fig. 1, bottom), the recovery of the carboxyl C in soil remained stable, while the recovery of all other positions in soil decreased by up to 60% of applied 13 C. This results in an equal recovery of all positions in soil on day 10. In microbial biomass and Σ -PLFA the recovery of 13 C from both carboxyl groups and glutamic acids amino bound position on day 10 was still lower than the 13 C recovery from other positions of both amino acids. In microbial biomass, the 13 C recovery of glutamic acids amino-bound position decreased by 35%. The same amount of 13 C from glutamic acid’s positions was recovered in Σ -PLFA on both days. Fig. 1 shows that on day 10, all C from the C 2 and C 3 positions of alanine in soil was located in the microbial biomass. The carboxyl C from alanine and glutamic acid, however, was stabilized in soil by other mechanisms. Fig. 1 Recovery of position-specifically 13 C labeled Ala and Glu in soil, microbial biomass and Σ -PLFA, 3 (top) and 10 days (bottom) after application. Letters indicate significant differences (p < 0.05) between recovery bulk soil (a), microbial biomass (a’) and Σ -PLFA (a’’) Publications and Manuscripts 107 To identify microbial groups, a PCA was performed on the PLFAs C-content from both sampling times. By comparing classification in the literature (Z ELLES 1999; Z ELLES et al. 1995), the fatty acid groups were matched to microbial groups and through factor loadings they were further subdivided (Supplementary Table 3). Recovery of applied positionspecifically labeled C from the two amino acids in most microbial groups (Fig. 2) shows the same pattern as recovery of applied C in Σ -PLFA: The 13 C recovery from the carboxyl groups is less than 0.1% of 13 C input of both amino acids on both days. On day 3, the recoveries of the amino-bound and the methyl C from alanine were similar. This pattern was different on day 10, when the recovery of alanines amino-bound C was lower than that of its methyl group. In the first days after being taken up by microorganisms and utilized in the cell membrane, only the C 1 position was split from the alanine molecule, while the C 2 and C 3 positions were utilized together. Fig. 2 Recovery of applied 13 C from positions of alanine (top) and glutamic acid (bottom) in microbial groups after 3 and 10 days. Letters indicate significant differences (p < 0.05) between carboxyl C (a), amino-bound C (a’) and methyl C of alanine or the residual molecule of glutamic acid (a’’). Publications and Manuscripts 108 The 13 C recovery of glutamic acid’s positions reveals that it is transformed differently than alanine. From both the amino-bound and the carboxyl C, less than 0.4% were recovered in Σ -PLFA on both days. In contrast, nearly 4% of the residual amino acid C were recovered in Σ -PLFA. In the microbial pathways both C 1 and the C 2 from glutamic acid were split from the residual molecule, which was then incorporated into PLFAs. The maximum incorporation of C into PLFAs from all positions of both amino acids was achieved by the group of gram negative I (18:1 ω 7c, 18:1 ω 9c) (Fig. 2). This group of gram negative prokaryotes took up 4.5 - 5.5% of the methyl C from alanine and also of the residual molecules C from glutamic acid. No other microbial group took up more than 2% from any position. Most prokaryotic groups incorporated more C from C 2 and C 3 positions of alanine than the anaerobic bacteria (cy17:0) and the two eukaryotic groups (Fungi (20:1 ω 9c, 18:2 ω 6,9) and VA-Mycorrhiza (16:1 ω 5c)). 2.3.3.3 Divergence Index The divergence index (DI) was used to compare the extent of incorporational discrimination of C from different positions between the pools (Fig. 3) and microbial groups (Fig. 4).The DI relativizes differences in absolute uptake. Regarding the DI in soil, microbial biomass and Σ -PLFA (Fig. 3), differences between days 3 and 10 after tracer application can be observed (Fig. 3). The relative incorporation in soil on day 3 shows a clear discrimination against the carboxyl positions; on day 10, there is no significant difference in DI between any position of alanine or glutamic acid. Although only the declined discrimination against alanines carboxyl C between day 3 and 10 is significant (p < 0.05), the reduced discrimination between all positions in soil between day 3 and day 10 shows that during the initial reactions, the pathways of C from different positions of the two amino acids differ greatly. However, ten days after application, the source position in the molecules is not determining for fixation in soil. In microbial biomass, a reduction of discrimination between positions may be taking place – the discrimination between positions is not significant anymore – but high standard errors prevent any certain conclusions. In Σ -PLFA, alanine’s amino-bound position had a DI equal to that of its methyl position on day 3 but a lower DI on day 10. The incorporation pattern of glutamic acids positions into Σ -PLFA did not change between day 3 and day 10. Publications and Manuscripts 109 Fig. 3 Divergence index (DI) reflecting incorporational discrimination between C positions into soil, microbial biomass and Σ -PLFA, 3 (left) and 10 (right) days after applying 13 C-labeled alanine (Ala) and glutamic acid (Glu). Letters indicate significant differences (p < 0.05) in the relative incorporation of the C positions into soil (a), microbial biomass (a’), Σ -PLFA (a’’) on day 3, and into Σ -PLFA (a**) on day 10 after tracer application. Despite the differences in absolute 13 C recovery in PLFAs between the microbial groups (Fig. 2), the DI (relative 13 C recovery) of most microbial groups PLFAs (Fig. 4) shows a similar pattern, which also generally reflects the pattern described by Σ -PLFA (Fig. 3). In most PLFAs, there was an average to above-average relative incorporation of the amino-bound group of alanine on day 3, which is prominent on day 10 (Fig. 4). The DI for the methyl position of alanine and the residual molecule of glutamic acid was aboveaverage in all microbial groups and on both days. The amino-bound C of glutamic acid was incorporated less than average in all microbial groups and on both days. Exceptions to this pattern were the groups of gram positive II (i15:0, i17:0) and fungi (20:1 ω 9c, 18:2 ω 6,9), which both showed no significant discrimination against any position on either days. Publications and Manuscripts 110 Fig. 4 Divergence Index (DI), reflecting discrimination between C positions of alanine (Ala) and glutamic acid (Glu), 3 (top) and 10 (bottom) days after application. Letters indicate significant differences (p < 0.05) between the relative incorporation of the C positions into the microbial group a: gram negative I, a’: gram negative II, a’’: gram positive I, a*: actinomycetes, a°: VA-mycorrhiza. 2.3.4 Discussion 2.3.4.1 Incorporation of carbon from amino acids in soil and microbial biomass On day 3, the 13 C recovery from alanine and glutamic acid in soil, microbial biomass and Σ -PLFA shows the same pattern (Fig. 3). C from the carboxyl group is recovered less than that from the amino-bound and methyl positions and the residual molecule of glutamic acid in all pools. This was expected as the carboxyl C has the highest oxidation state and is therefore most prone to being removed from the molecules. This is achieved by decarboxylation. Enzymes necessary for decarboxylation of amino acids have been found in soil (Braun et al., 1992; Tena et al., 1986) as well as in prokaryotic Publications and Manuscripts 111 and eukaryotic microorganisms (Caspi et al., 2008). On day 10, the amount of carboxyl C from both amino acids in soil remained stable, while the recovery of all other positions decreased. Although it seems contradictory this can also be explained by the high reactivity of the carboxyl C: not only can it be oxidized to CO 2 easily, but it can also react with other soil components and be thus stabilized. This possible stabilization mechanism is supported by results of Kuzyakov (1997), who found position-specifically labeled 14 C from the carboxyl position of alanine in humic and fulvic acids. As complex macromolecules, humic and fulvic acids contain a variety of functional groups such as hydroxyl groups, methylenes, ethers and esters in aliphatic chains (Simpson et al., 2002). It is possible for carboxyl C from microbial sources to react with humic macromolecules, e.g. by forming ester-linkages with hydroxy groups. Esters are highly inert, therefore the former carboxyl C will be stabilized from further degradation. In contrast to carboxyl C, the 13 C recovery from the amino-bound and methyl group of alanine in soil decreased by up to 60% between days 3 and 10. Compared to the decrease in recovery of these positions in soil, the amount incorporated into microbial biomass is still high on day 10. In microorganisms, alanine can be used catabolically, in the citric acid cycle and anabolically, e.g. to produce sugars or fatty acids (Fig. 5) (Caspi et al., 2008). The first reactions for both pathways are the same. Alanine is first deaminated and then decarboxylated, thereafter the resulting acetyl reacts with coenzyme A to form acetyl-CoA. The acetyl-CoA, which consists of the former amino-bound and methyl C from alanine, is then fed into the citric acid cycle or used for biosynthesis. This explains why C from those two positions is recovered in PLFAs, but C from the carboxyl group is not. After incorporation into PLFAs, C from the former amino-bound position is on the terminal position and thus most prone to being oxidized and decarboxylated (Caspi et al., 2008).This process is hinted at by the slight decrease in relative incorporation of alanines amino-bound position between day 3 and 10. The incorporation of the C from the methyl position of alanine and from the residual molecule of glutamic acid in Σ -PLFAs is still high on day 10. 2.3.4.2 Incorporation of tracer into the microbial groups The 13 C incorporation into PLFAs of microbial groups differed by more than one order of magnitude (Fig. 2). As hypothesized, the highest incorporation, with more than 5% 13 C uptake, was recorded for a group of gram negative prokaryotes (gram negative I). This fits well with the observations by Griffiths et al. (1999) that gram negatives react fastest to addition of LMWOS, which gives them a competitive advantage. Publications and Manuscripts 112 Fig. 5 Microbial transformation pathways of alanine (a) and glutamic acid (b, c, d). As there are different transformation pathways for glutamic acid, it is presented in 3 subfigures. The entrance of alanine (a) occurs from the bottom (in contrast to glutamic acid, b, c, d,) of the citric acid cycle because of its initial transformation to acetyl-CoA. Three other prokaryotic groups (gram negatives II, actinomycetes and gram positive II), also achieved moderate 13 C incorporation. The two eukaryotic groups – fungi and VA-mycorrhiza – were unable to take up as much of the applied amino acid C as the prokaryotic group. This is unsurprising because the turnover of the larger, more complex eukaryotes’ biomass is slower than that of prokaryotes’ (Bååth 1998, Rousk & Bååth 2007). Accordingly, enrichment of eukaryotic cell components takes longer (Moore et al., Publications and Manuscripts 113 2005). Apart from a slower turnover, the larger size of eukaryotic cells results in a smaller ratio of surface to volume. As PLFAs are utilized as cell membranes on the surface of the organism and the difference in the ratios of 13 C in PLFA to 13 C in microbial biomass for various microbial groups is unknown, there is no full comparability between cells of different size. It is also well known that fungi are specialized on more complex substrate than LMWOS. As in the eukaryotes, the anaerobic bacteria also incorporate only a maximum of 0.7% of the applied C. As the roof we installed prevented excess wetting, the soil was well aerated, so the anaerobic microorganisms could only persist inside anaerobic microhabitats such as microaggregates. Thus, only 13 C that permeated into those anaerobic microhabitats could be taken up by anaerobic microorganisms. 2.3.4.3 Discrimination of individual carbon positions by microbial utilization differs depending on oxidation state, amino acid and time As in soil and microbial biomass, discrimination of the individual C positions of both amino acids also took place in the microbial PLFA. As the percent of 13 C recovery (Fig. 2) between the microbial groups’ PLFA differs greatly, the discrimination between the positions of alanine and glutamic acid is best evaluated with the DI (Fig. 4). On day 3, there was no difference in relative incorporation of 13 C from the methyl and amino-bound C of alanine for most microbial groups. Nearly no 13 C from alanine’s carboxyl group was recovered in the PLFAs and the incorporation of alanine’s 13 C into microbial biomass is much lower than that of its amino-bound and methyl position. Accordingly, we can conclude that during the three days after applying the amino acid, the C 1 atom in alanine is split from the molecule quickly, whereas C 2 and C 3 remain bonded. Presumably, the alanine molecule is taken up and then metabolized in the main alanine utilization pathway: deamination to pyruvate and after decarboxylation by pyruvate dehydrogenase, transformation to acetyl-CoA (Fig. 5a). This molecule then either enters the citric acid cycle (de Kok et al., 1998) or fatty acid synthesis (Caspi et al., 2008). On day 10, the DI of the amino-bound C is slightly lower than that of alanine’s methyl C in most microbial groups, which can be explained by the further reactions in microorganisms: If the molecule is used catabolically in the citric acid cycle, then the acetyl-CoA condensates with oxalate to citric acid. After this reaction, the former amino-bound C of alanine is one of citric acid’s carboxyl groups. Thus, the chance for the amino-bound position to be degraded into CO 2 during the next step – the formation of 2-oxoglutarate (Camacho et al., 1995) – is about 1:3, whereas the methyl position is still incorporated in the nonreactive chain. After every circuit of the citric acid cycle, C from the alanine molecule can Publications and Manuscripts 120 Supplementary Data Supplementary Table A1: Fatty acids in the external standard Publications and Manuscripts 121 Supplementary Table A2: Results of factor analysis Publications and Manuscripts 122 2.4 Study 4: Biogeochemical transformations of amino acids in soil assessed by position-specific labeling Michaela A. Dippold 1,2 , Yakov Kuzyakov 2,3 1 Department of Agroecosystem Research, University of Bayreuth 2 Department of Agricultural Soil Science, Georg-August-University of Göttingen 3 Department of Soil Science of Temperate Ecosystems, Georg-August-University of Göttingen Corresponding Author: Michaela Dippold Department of Agroecosystem Research University of Bayreuth Universitätstraße 30 95447 Bayreuth email: [email protected] Tel.: 0921/552187 Fax.: 0921/552246 Publications and Manuscripts 123 Abstract BACKGROUND AND AIMS: Amino acid turnover in soil is an important element of terrestrial carbon and nitrogen cycles. This study accounts for their driver - the microbial metabolism - by tracing them via the unique isotopic approach of position-specific labeling. METHODS: Three 14 C isotopomers of alanine at five concentration levels combined with selective sterilization were used to distinguish sorption mechanisms, exoenzymatic and microbial utilization of amino acids in soil. RESULTS: Sorption and microbial uptake occurred immediately. Unspecific microbial uptake followed a linear kinetic, whereas energy-dependent uptake followed Michaelis-Menten. Less than 6% of the initially added alanine was sorbed to soil, but after microbial transformation products were bound to the soil matrix at higher proportions (525%). The carboxyl group (C-1) was rapidly oxidized by microorganisms, whereas C-2 and C-3 positions were preferentially incorporated into microbial biomass. Dependency of C metabolization on amino acid concentration reflected individual alanine transformation pathways for starvation, maintenance and growth conditions. CONCLUSIONS: This study demonstrates that position-specific labeling determines the mechanisms and rates of C cycling from individual functional groups. This approach reflected underlying metabolic pathways and revealed the formation of new organic matter. We therefore conclude that position-specific labeling is a unique tool for detailed insights into submolecular transformation pathways and their regulation factors. Keywords : Position-specific tracers, Amino acids stabilization, Sorption, Exoenzyme and uptake kinetics, Metabolic tracing, Soil organic matter formation, Sterilization and inhibition methods, Biochemical pathways Publications and Manuscripts 124 2.4.1 Introduction Studies on transformation of organic substances in soils are important for understanding of C and N cycles in terrestrial ecosystems. Plant residues and rhizodeposits are the main sources of organic matter in soils (Rasse et al., 2005). Therefore, many studies have focused on decomposition, microbial utilization and stabilization of C from these sources (von Luetzow et al., 2006). During decomposition of litter, macromolecular compounds are depolymerized by enzymes into low molecular weight organic substances (LMWOS) (Cadisch and Giller, 1996). LMWOS are the lightest (<250 Da) components of DOC (Boddy et al., 2007) from substance classes such as organic acids, amino acids, monoand disaccarides, amino sugars, phenols and many more (van Hees et al., 2005a). In addition to litter, rhizodeposition is a source of LMWOS in soil. Microorganisms determine the fate of LMWOS in soil because they either produce them, decompose them to CO 2 and NH 4+ (catabolism) or incorporate them in cellular compounds (anabolism). The importance of LMWOS is not connected with their pool size (Fischer et al., 2007), but with the huge fluxes that pass through this pool. Therefore, the transformation pathways of LMWOS represent a crucial step of soil C and N fluxes, and a molecular-level knowledge of these processes is needed (van Hees et al., 2005a). Within the LMWOS, amino acids play an important role because they are the quantitatively most important compounds coupling the C and N cycle. In topsoil, amino bound N constitutes 7-50% of the total organic N (Gardenas et al., 2011; Stevenson, 1982a). Thus, many recent studies focused on the fate of N-containing LMWOS (Hobbie and Hobbie, 2010; Jones et al., 2004b; Knowles et al., 2010; Kuzyakov, 1996; Lipson et al., 2001; Vinolas et al., 2001a) and investigated the three major pathways of amino acid utilization in soil: 1) sorption (Jones, 1999), 2) extracellular transformation, and 3) intracellular metabolization (Vinolas et al., 2001a; Vinolas et al., 2001b) which can be separated by selective inhibition of biotic processes. Sorption strongly depends on the functional group of the amino acid (Jones and Hodge, 1999): it can occur by ion exchange of positively charged amino groups, by ligand exchange of carboxyl groups and by hydrophobic interactions with alkyl groups. To date, nearly all studies assumed sorption of the entire molecule by soil sorbents. Only a few studies on glycine sorption indicated abiotic degradation of the sorbed amino acid (Wang and Huang, 2003, , 2005). Amino acids can be transformed extracellularly, mainly by exoenzymes attached to cell surfaces (Geisseler et al., 2010). Deamininating (Killham, 1986) and oxidizing (Bohmer et al., 1989; Braun et al., 1992) extracellular systems are described in the litera- Publications and Manuscripts 125 ture, but neither their relevance nor the differences between extraand intracellular pathways have been investigated (Burns, 1982). Intracellular amino acid metabolization follows the uptake by transport systems (Anraku, 1980; Hediger, 1994; Hosie and Poole, 2001). Uptake kinetics of some amino acids has already been elucidated (Vinolas et al., 2001a; Vinolas et al., 2001b). Barraclough (1997) showed that the majority of N mineralization of amino acids occurred inside the cells. Knowles et al. (2010) described for the first time the decoupling of N and C metabolization in soil, discovering a preferential retention of amino acid N with respect to C. Nonetheless, as they used uniformly labeled tracers, they could not determine the fate of the C skeleton. We hypothesize that the fate of amino acid C and N in soil is mainly determined by the dominating intracellular metabolization pathways. Therefore, identification of microbial metabolization is a crucial step for understanding and predicting C and N fluxes. In addition to abiotic factors such as temperature (Dijkstra et al., 2011c; Vinolas et al., 2001b) or soil properties (Gonod et al., 2006; Kemmitt et al., 2008), the concentration of a substrate is a key driver of the intracellular metabolization (Dijkstra et al., 2011a; Fischer and Kuzyakov, 2010b; Schneckenberger et al., 2008). Soil amino acid concentrations range from 0.5 µM in root-free bulk soil to 5 mM directly next to bursting cells (Fischer et al., 2007; Jones and Hodge, 1999). We expect cellular uptake and metabolism always dominate the amino acid removal from soil solution and that sorption only plays a relevant role at low substrate concentrations. For our study, we chose alanine as a representative amino acid for the neutral amino acids as it is one of the most dominant amino acids in soil solution (Fischer et al., 2007). In addition, alanine was chosen because it is very close to the basic C metabolism of the cell: by oxidative deamination alanine can be transferred to pyruvate, which is a suitable substrate for metabolic tracing experiments in plants and soils (Dijkstra et al., 2011a; Tcherkez et al., 2005). To elucidate intraand extracellular alanine transformation pathways, we used the approach of position-specific labeling. This tool is commonly used in biochemistry to investigate metabolization pathways, but has rarely been applied in soil science (Dijkstra et al., 2011a; Dijkstra et al., 2011b; Dijkstra et al., 2011c; Fischer and Kuzyakov, 2010b; Fokin et al., 1993, , 1994; Haider and Martin, 1975; Kuzyakov, 1997; Nasholm et al., 2001). It overcomes the limitations of uniform labeling because it allows differentiating between incorporation of fragments vs. incorporation of entire molecules. Coupling of position-specific labeling with soil sterilization enables us to separate abiotic splitting of alanine from extracellular and from cellular metabolism. We assume that extra– and intracellular transformation differ from each other as they are based on different enzymes. By comparison of the kinetics of alanine removal from soil solution in Publications and Manuscripts 126 the non-inhibited and respiration-inhibited treatments, the relevance of extraversus intracellular transformations of alanine was compared. We hypothesize that under soil conditions microbial uptake systems and intracellular metabolization dominate the fate of alanine in soil. Comparing our results with known microbial metabolization pathways enables the identification of metabolic changes depending on substrate concentration. 2.4.2 Material and Methods 2.4.2.1 Soil Topsoil (0-10 cm) from the Ap horizon of a silt loam haplic Luvisol (WRB, 2006) was collected from a field in Bavaria with a crop sequencing of barley, wheat and triticale (49.907 N, 11.152 E, 501 m asl, mean annual temperature 6-7 °C, mean annual precipitation 874 mm). The soil had a pH KCl of 4.88 and pH H2O of 6.49, total organic C and total N content were 1.77% and 0.19%, respectively, and potential CEC was 13.6 cmol c kg -1 . Soil was sieved to 2 mm, and all roots were removed with tweezers. Soil was stored at field moisture at 5 °C not longer than one week until the experiment started. 2.4.2.2 Chemicals and radiochemicals Stock solutions with 1, 10, 100, 1000, and 10000 µM alanine and an equal activity of 10 4 DPM ml -1 (Disintegrations Per Minute and ml) were prepared from U14 C-labeled alanine and the position-specifically labeled isotopomers 114 C-, 214 Cand 314 C-labeled alanine (American Radiolabeled Chemicals Inc, St. Louis, USA) as well as non-labeled alanine (Sigma-Aldrich, Taufkirchen, Germany). Sterilization solutions were produced with 1 mM NaN 3 to inactivate aerobic microbial respiration or with 1 mM NaN 3 and 1 mM HgCl 2 to denaturate all proteins and reach full inhibition of biotic processes. Effectiveness of the chosen azide inhibitor was evaluated by a qualitative 2,3,5-triphenyltetrazoliumchloride incubation (TCC, Sigma-Aldrich, Taufkirchen, Germany). Therefore 0.63 µg of the yellow dye TCC were added to the 1 ml of soil suspension in this preexperiment. 2.4.2.3 Experimental setup The effects of two factors on alanine transformations in soil were investigated: 1) the concentration of alanine, and 2) the extraand intracellular as well as abiotic processes of alanine removal from soil solution, separated by sterilization. Therefore, three Publications and Manuscripts 127 sterilization treatments were used (Fig. 1): 1) treatments without any inhibition, where three groups of processes occured: intracellular metabolism, extracellular transformation and physicochemical sorption, 2) treatments with inhibition of aerobic respiratory chains by azides (Burns, 1982), where only extracellular processes are active and sorption could occur, and 3) treatments with full inhibition, where microbial metabolism as well as exoenzymes were inhibited by HgCl 2 (Stevenson and Verburg, 2006; Wolf et al., 1989) and only sorption could remove alanine from the soil solution (Fig. 1). We define here as extracellular transformations all processes (decomposition, decarboxylation, condensation, etc.) localized in the soil solution or periplasm (Glenn, 1976) which don’t depend on intracellular energy metabolism (i.e. proton gradient or ATP) and can not be inhibited by NaN 3 . Biotic transformations sums up extraas well as intracellular processes. Fig. 1 Scheme of the experimental design for one of the five concentrations: in part 1 on the left side (incubation experiment) yellow-shaded plates shows fullyinhibited treatment to investigate sorption whereas green-shaded plates reflect biotic utilization (upper line with only extracellular activity and lower line with extraand intracellular activity). Yellow-shaded graphs demonstrate the calculation of the sorbed proportion of alanine by the sorption isotherm, which is derived from the fully inhibited treatment. Green-shaded graphs reveal the calculation of the biotic utilization by substracting the sorption from the percentage of alanine removal from supernatant. In part 2 on the right side (extraction experiment) purple-shaded plates reflect the fully-inhibited treatment and thus extraction of untransformed alanine by the sequential procedure. Blue-shaded plates show desorption of biotic alanine transformation products (upper plate with only extracellular activity and lower plate with extraand intracellular activity). Publications and Manuscripts 128 The experiment consisted of two parts (Fig. 1): In the first part – the incubation experimentthe processes removing alanine from the supernatant were investigated. The incubation was performed in 24-deep-well plates (6 ml volume per well) on a rotational shaker at 200 rpm with 200 mg field fresh soil per replication. Before adding the alanine, the soil was pre-incubated for 1 h with 0.5 ml of 1 mM sterilization solutions or distilled water, respectively. Pre-incubation was performed under intensive shaking to enable a homogenous sterilization of the entire soil volume under high oxygen supply. Thus, during pre-incubation anaerobic processes were prevented, the stored energy could be consumed and no new energy reserves were produced. In the treatment with extracellular processes, the intracellular respiration was inhibited with 0.5 ml 1 mM NaN 3 . Although chosen NaN 3 -concentrations are far above those described for respiratory chain inhibition (Kita et al., 1984) some activity may remain in the soil suspension. This was evaluated by a triphenyl-tetrazolium chloride assay. This dye is intracellularly reduced by various dehydrogenases (Kvasnikov et al., 1974; Mohammadzadeh et al., 2006). An active intracellular metabolism leads to the formation of insoluble red formazan crystals within living cells. In the treatment with full inhibition, denaturation of proteins was achieved by adding 0.5 ml of 1 mM HgCl 2 and 1 mM NaN 3 . After pre-incubation, 0.5 ml of the alanine-solution was added. All experiments were performed with uniformly labeled alanine and the three isotopomers. The soil suspension was shaken for 30 seconds, centrifuged at 2000 rpm and an aliquot of 50 µl was removed for 14 C measurement. After remixing, incubation was continued, and further 50 µl were sampled 5, 15, 30 and 60 min and 6 h, 12 h and 36 h after addition of 14 C labeled alanine. After incubation, the remaining supernatant was removed and soil was washed three times – first with distilled water, then with full inhibition solution and finally with distilled water. In the second part of the study - the desorption experiment - we evaluated the binding mechanisms of alanine C in soil (Fig. 1). In treatments with full inhibition, the extracted C reflects alanine C itself, as no biotic transformation occurred. In treatments with biotic activity, the microbial or extracellular transformation products were extracted. The washing step with HgCl 2 led to denaturation of membrane proteins and thus a loss of membrane integrity. This allowed the joint extraction of water soluble cytoplasm compounds and extracellular transformation products. Macromolecular compounds like proteins, polysaccharides or peptidoglycan as well as hydrophobic compounds like the membrane lipids could not be extracted by a salt solution. For the desorption experiment, 0.5 ml of 0.5 M CaCl 2 solution was added to the soil and shaken for 2 h. The solution was centrifuged, and supernatant was removed and stored for 14 C analysis. Desorption was repeated three times, and the supernatants were Publications and Manuscripts 129 combined to one solution, in which 14 C was analyzed. This desorption treatment with CaCl 2 enabled evaluating the amount of alanine being weakly bound, mainly by ion exchange. After extraction with CaCl 2 the same procedure was done three times with 0.5 M NaH 2 PO 4 solution to extract the alanine bound by ligand-exchange. To estimate irreversibly bound alanine C, the soil was freeze-dried and combusted at 600 °C for 10 min under a constant O 2 stream with a HT 1300 solid combustion module of the multi N/C 2100 analyzer (Analytik Jena, Jena Germany). 14 CO 2 released by combustion was trapped in 10 ml of 1 M NaOH. The irreversibly bound pool contains untransformed, irreversibly bound alanine C as well as macromolecular, hydrophobic or irreversibly bound microbial transformation products. 2.4.2.4 Radiochemical analyses 14 C activity of the supernatants was determined using a scintillation counter (Wallac 1450, MicroBeta ® TriLux, PerkinElmer, Walham MA; USA) by adding 50 µl of the supernatant directly to 0.6 ml scintillation cocktail (EcoPlus, Roth Company, Germany) in transparent 24-well plates. Remaining supernatant, washing solution and desorption solution were measured in glass scintillation vials with the LS 6500 scintillation counter (LS 6500, Beckman-Coulter, Krefeld, Germany) with a 1:2 ratio of solution to scintillation cocktail and a 1:8 ratio for the CaCl 2 and NaH 2 PO 4 solutions. 14 C activity in the NaOH solution was measured with a 1:2 ratio of sample to scintillation cocktail after 24 h of dark storage after disappearance of chemoluminescence. All measurements with the LS 6500 were also performed with blanks of the respective solutions (CaCl 2 , NaH 2 PO 4 or NaOH) and background corrected by subtracting this value from each measurement result. 2.4.2.5 Calculation of the kinetics of alanine utilization To calculate the biotic utilization the amount of sorbed alanine C has to be subtracted from the total removal from soil suspension. Therefore, the decrease in 14 C activity in the supernatant of the fully inhibited treatment (A % (t) in percent of added activity) was fitted to an exponential equation (Fig. 1) where B (% of added activity) and c (1/h) are the fitted parameters and D equ is the remaining percentage of activity in the supernatant at equilibrium. The remaining activity D equ was converted into the amounts of sorbed alanine per g soil (S equ in µmol g -1 ) and the dissolved alanine concentration c equ (µM) were calculated and all five concentration treatments were fitted by a linearized Freundlich sorption isotherm with the sorption affinity constants k and n (Fig. 1). Based on Fischer and Kuzyakov (2010b), the fitted sorption isotherm was used to calculate the Publications and Manuscripts 232 Fernandez, C.W., Koide, R.T., 2012. The role of chitin in the decomposition of ectomycorrhizal fungal litter. Ecology 93, 24-28. Fischer, H., Eckhardt, K.-U., Meyer, A., Neumann, G., Leinweber, P., Fischer, K., Kuzyakov, Y., 2010. Rhizodeposition of maize: Short-term carbon budget and composition. Journal of Plant Nutrition and Soil Science 173, 67-79. Fischer, H., Kuzyakov, Y., 2010. Sorption, microbial uptake and decomposition of acetate in soil: Transformations revealed by position-specific C-14 labeling. Soil Biology & Biochemistry 42, 186-192. Fokin, A.D., Knyazev, D.A., Kuzyakov, Y.V., 1993. Destruction of C-14-labeled and N-15labeled amino acids and nucleic bases in soil and the supply of their transformation products to plants. Eurasian Soil Science 25, 109-122. Fokin, A.D., Knyazev, D.A., Kuzyakov, Y.V., 1994. Incorporation of C-14 and N-15 amino acids and nucleic bases into humus and the turnover of atomic molecular composition. Eurasian Soil Science 26, 24-34. Gearing, P.J., Gearing, J.N., Maughan, J.T., Oviatt, C.A., 1991. Isotopic distribution of carbon from sewage-sludge and eutrophication in the sediments and food web of estuarine ecosystems. Environmental Science & Technology 25, 295-301. Glaser, B., Amelung, W., 2002. Determination of C-13 natural abundance of amino acid enantiomers in soil: methodological considerations and first results. Rapid Communications in Mass Spectrometry 16, 891-898. Glaser, B., Gross, S., 2005. Compound-specific delta C-13 analysis of individual amino sugars - a tool to quantify timing and amount of soil microbial residue stabilization. Rapid Communications in Mass Spectrometry 19, 1409-1416. Glaser, B., Turrion, M.B., Alef, K., 2004. Amino sugars and muramic acid - biomarkers for soil microbial community structure analysis. Soil Biology & Biochemistry 36, 399-407. Gottwald, W., 2000. Statistik für Anwender, 1. Auflage ed. WILEY-VCH Verlag GmbH, Weinheim, 227 pp. Haider, K., Martin, J.P., 1975. Decomposition of Specifically C-14-Labeled Benzoic and Cinnamic Acid-Derivatives in Soil. Soil Science Society of America Journal 39, 657662. Indorf, C., Bode, S., Boeckx, P., Dyckmans, J., Meyer, A., Fischer, K., Jörgensen, R.G., 2013. Comparison of HPLC Methods for the Determination of Amino Sugars in Soil Hydrolysates. Analytical Letters 46, 2145-2164. Indorf, C., Stamm, F., Dyckmans, J., Joergensen, R.G., 2012. Determination of saprotrophic fungi turnover in different substrates by glucosamine-specific delta C-13 liquid chromatography/isotope ratio mass spectrometry. Fungal Ecology 5, 694-701. Keiluweit, M., Bougoure, J.J., Zeglin, L.H., Myrold, D.D., Weber, P.K., Pett-Ridge, J., Kleber, M., Nico, P.S., 2013. Nano-scale investigation of the association of microbial nitrogen residues with iron (hydr)oxides in a forest soil O-horizon. Geochimica Et Cosmochimica Acta 95, 213-226. Keseler, I.M., Bonavides-Martinez, C., Collado-Vides, J., Gama-Castro, S., Gunsalus, R.P., Johnson, D.A., Krummenacker, M., Nolan, L.M., Paley, S., Paulsen, I.T., Peralta-Gil, M., Santos-Zavaleta, A., Glennon Shearer, A., Karp, P.D., 2009. EcoCyc: A comprehensive view of Escherichia coli biology. Nucleic Acids Research 37, D464D470. Kindler, R., Miltner, A., Thullner, M., Richnow, H.-H., Kaestner, M., 2009. Fate of bacterial biomass derived fatty acids in soil and their contribution to soil organic matter. Organic Geochemistry 40, 29-37. Kotnik, M., Anderluh, P.S., Prezelj, A., 2007. Development of novel inhibitors targeting intracellular steps of peptidoglycan biosynthesis. Current Pharmaceutical Design 13, 2283-2309. Kuzyakov, Y., Domanski, G., 2000. Carbon input by plants into the soil. Review. Journal of Plant Nutrition and Soil Science 163, 421-431. Kuzyakov, Y.V., 1997. The role of amino acids and nucleic bases in turnover of nitrogen and carbon in soil humic fractions. European Journal of Soil Science 48, 121-130. Publications and Manuscripts 233 Lauer, F., Koesters, R., du Preez, C.C., Amelung, W., 2011. Microbial residues as indicators of soil restoration in South African secondary pastures. Soil Biology & Biochemistry 43, 787-794. Lengeler, J.W., Drews, G., Schlegel, H.G., 1999. Biology of the Prokaryotes. Georg Thieme Verlag, Stuttgart, Germany. Macura, J., Kubatova, Z., 1973. Control of carbohydrate utilization by soil microflora. Soil Biology and Biochemistry 5, 193-204. Merzendorfer, H., 2011. The cellular basis of chitin synthesis in fungi and insects: Common principles and differences. European Journal of Cell Biology 90, 759-769. Meyer, A., Fischer, H., Kuzyakov, Y., Fischer, K., 2008. Improved RP-HPLC and anionexchange chromatography methods for the determination of amino acids and carbohydrates in soil solutions. Journal of Plant Nutrition and Soil Science 171, 917-926. Milewski, S., Gabriel, L., Olchowy, J., 2006. Enzymes of UDP-GlcNAc biosynthesis in yeast. Yeast 23, 1-14. Miltner, A., Bombach, B., Schmidt-Brücken, B. Kästner, M., 2012. SOM genesis: microbial biomass as a significant source. Biogeochemistry 111, 41-55. Moore, J.C., McCann, K., de Ruiter, P.C., 2005. Modeling trophic pathways, nutrient cycling, and dynamic stability in soils. Pedobiologia 49, 499-510. Nasholm, T., Huss-Danell, K., Hogberg, P., 2001. Uptake of glycine by field grown wheat. New Phytologist 150, 59-63. Nguyen, C., Guckert, A., 2001. Short-term utilisation of C-14-[U]glucose by soil microorganisms in relation to carbon availability. Soil Biology & Biochemistry 33, 53-60. Park, J.T., Uehara, T., 2008. How bacteria consume their own exoskeletons (Turnover and recycling of cell wall peptidoglycan). Microbiology and Molecular Biology Reviews 72, 211-227. Ranneklev, S.B., Baath, E., 2003. Use of phospholipid fatty acids to detect previous selfheating events in stored peat. Applied and Environmental Microbiology 69, 35323539. Rasse, D.P., Rumpel, C., Dignac, M.F., 2005. Is soil carbon mostly root carbon? Mechanisms for a specific stabilisation. Plant and Soil 269, 341-356. Rethemeyer, J., Kramer, C., Gleixner, G., Wiesenberg, G.L.B., Schwark, L., Andersen, N., Nadeau, M.J., Grootes, P.M., 2004. Complexity of soil organic matter: AMS C-14 analysis of soil lipid fractions and individual compounds. Radiocarbon 46, 465-473. Richmond, M.H., Perkins, H.R., 1962. Biosynthesis of muramic acid by intact Staphylococci. Biochemical Journal 85, 580-588. Rousk, J., Baath, E., 2007. Fungal biomass production and turnover in soil estimated using the acetate-in-ergosterol technique. Soil Biology & Biochemistry 39, 2173-2177. Said-Pullicino, D., Birk, J.J., Glaser, B., in press. Nitrogen immobilization in paddy soils as affected by redox conditions and rice straw incorporation. Geoderma http://dx.doi.org/10.1016/j.geoderma.2013.06.020. Scandellari, F., Hobbie, E.A., Ouimette, A.P., Stucker, V.K., 2009. Tracing metabolic pathways of lipid biosynthesis in ectomycorrhizal fungi from position-specific 13Clabelling in glucose. Environmental Microbiology 11, 3087-3095. Simpson, A.J., Simpson, M.J., Smith, E., Kelleher, B.P., 2007. Microbially derived inputs to soil organic matter: Are current estimates too low? Environmental Science & Technology 41, 8070-8076. Uehara, T., Park, J.T., 2008. Growth of Escherichia coli : Significance of peptidoglycan degradation during elongation and septation. Journal of Bacteriology 190, 3914-3922. Veuger, B., van Oevelen, D., Boschker, H.T.S., Middelburg, J.J., 2006. Fate of peptidoglycan in an intertidal sediment: An in situ C-13-labeling study. Limnology and Oceanography 51, 1572-1580. von Luetzow, M., Koegel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G., Marschner, B., Flessa, H., 2006. Stabilization of organic matter in temperate soils: mechanisms and their relevance under different soil conditions - a review. European Journal of Soil Science 57, 426-445. Publications and Manuscripts 234 Waring, B.G., Averill, C., Hawkes, C.V., 2013. Differences in fungal and bacterial physiology alter soil carbon and nitrogen cycling: insights from meta-analysis and theoretical models. Ecology Letters 16, 887-894. Wu, J., Joergensen, R.G., Pommerening, B., Chaussod, R., Brookes, P.C., 1990. Measurement of soil microbial biomass C by fumigation extraction - an automated procedure. Soil Biology & Biochemistry 22, 1167-1169. Zhang, X.D., Amelung, W., 1996. Gas chromatographic determination of muramic acid, glucosamine, mannosamine, and galactosamine in soils. Soil Biology & Biochemistry 28, 1201-1206. Publications and Manuscripts 235 Supplementary Data Supplementary Table A1: Solvent gradients and flow conditions of the IC-O-IRMS measurment time 20 mM NaOH 200 mM NaOH H 2 O 0.01 M NaNO 3 flow (ml min -1 ) -25 min 0% 100% 0% 0% 0.400 -10 min 8% 0% 92% 0% 0.325 11 min 40% 0% 35% 25% 0.400 15 min 45% 0% 45% 10% 0.400 18 min 25% 25% 50% 0% 0.380 35 min standby Publications and Manuscripts 236 2.8 Study 8: Formation and transformation of fatty acids in soil assessed by position-specific labeling of precursors Michaela Dippold 1,2 , Yakov Kuzyakov 1,3 1 Department of Agricultural Soil Science, Georg-August-University of Göttingen 2 Department of Agroecosystem Research, University of Bayreuth 3 Department of Soil Science of Temperate Ecosystems, Georg-August-University of Göttingen Corresponding Author: Michaela Dippold Department of Agricultural Soil Science Georg-August University of Goettingen Buesgenweg 2 37077 Goettingen email: [email protected] Tel.: 0551-3933546 Fax.: 0551-3933310 Publications and Manuscripts 237 Abstract Fatty acids are frequently used as plant and microbial biomarkers to trace the pathways of C stabilization and soil organic matter (SOM) formation. Whereas microbial and plant fatty acid fingerprints are well investigated, their transformations in soils remain unclear. However, knowledge of the transformation pathways in soils is crucial for the interpretation of fatty acid fingerprints, especially because the formation and decomposition processes are simultaneously ongoing. Therefore, we analyzed the formation of microbial fatty acids from their precursor acetate and the transformation of palmitate in soil by coupling position-specific 13 C labeling with compound-specific 13 C analysis. Position-specifically and uniformly 13 C-labeled acetate and palmitate were applied in an agricultural Luvisol. Pathways of fatty acids were traced by analyzing microbial utilization of C from individual molecule positions of acetate and palmitate and their incorporation into phospholipid fatty acids (PLFAs). Acetate 13 C incorporation into microorganisms and that remaining in the soil were characterized by basic microbial metabolism: C-1 is preferentially oxidized to CO 2 in the citric acid cycle, whereas C-2 is preferentially incorporated into microbial compounds. If palmitate was used in basic C metabolism, it was split into C2-units (acetyl-CoA), and odd and even positions of palmitate were transformed in a manner similar to acetate. However, as palmitate is the preferred precursor for PLFA formation, more than 6% of the added palmitate was incorporated into microbial cell membranes. Newly formed fatty acids were first, on day 3, dominated by basic, straight chain fatty acids. With increasing time, the pattern of newly formed PLFA approached the fingerprint of the microbial community. Therefore, the C backbone of palmitate was not split, but modified (e.g. desaturated, elongated or branched) according to the fatty acid demand of the soil microbial community. If acetate 13 C was used for PLFA formation, the construction of new C backbones of fatty acids rarely occurred. However, acetate 13 C was incorporated into microbial PLFAs by elongations or branchings of already existing fatty acids. Therefore, the previous assumption, that fatty acids are generally newly formed from the added substrates has to be discarded and future PLFA studies have to consider the reuse of existing plant and microbial-derived fatty acids. Discrimination of acetate positions by PLFA formation was lowest in the microbial groups with the highest competitiveness for acetate uptake. In contrast, palmitate uptake and transformations were highly specific for the individual microbial groups in soil. For both substrates, it could be concluded that more direct, less complex metabolic pathways are characteristic of fast-growing microbial groups with high turnover. Publications and Manuscripts 238 This study proves the fast microbial turnover of the free fatty acid pool in soils, as well as the high turnover and transformation of cellular PLFAs. Knowledge about these microbial transformations of fatty acids in soils is crucial for the interpretations of microbial as well as plant-derived fatty acid fingerprints. Furthermore, tracing the formation and transformation of lipids in soils improves our understanding of C fluxes and the stabilization of microbial as well as plant-derived lipids in soils. Keywords: position-specific tracers, metabolic tracing, biomarker approaches, fatty acid formation and transformation, phospholipids, lipid stabilization, paleoenvironmental reconstructions Publications and Manuscripts 239 2.8.1 Introduction Soil organic matter (SOM) is the largest active carbon (C) pool (1462-1548 Pg, (Batjes, 1996)) within the global carbon (C) cycle, but the genesis and transformation processes are poorly understood. The main input of C into soils occurs via plant litter or rhizodeposition (Rasse et al., 2005). Litter is composed of macromolecules such as cellulose, hemicellulose, lignin or proteins (Crawford et al., 1977; Sorensen, 1975); in addition, rhizodeposits contain low molecular weight organic substances (LMWOS) (Farrar et al., 2003). Along with water soluble low molecular weight organic compounds and their polymers, lipids are important constituents of plant biomass. They comprise around 3-10% of aboveground and 0.5-5% of belowground plant biomass and are thus an essential compound of plant C input into soils (Bliss, 1962; Ohlrogge and Browse, 1995; Wiesenberg et al., 2004). In addition, microbial biomass contains around 10% of lipids, mainly in their cell membranes and cell walls (Lengeler et al., 1999; Zelles et al., 1995) and significantly contribute to the lipidic SOM pool. Lipids comprise a higher percentage of SOM (Almendros et al., 1991; Rumpel et al., 2004) than their source material i.e. plant and microbial biomass. This accumulation of lipids already indicates their selective preservation in soils (Lichtfouse et al., 1998a). Moreover, lipids are assumed to play an even more important role in SOM formation and stabilization: lipids are stabilized by hydrophobic interactions with SOM and with each other, leading to a decreased wettability and subsequent hampering microbial decomposition (Lichtfouse et al., 1998a; von Luetzow et al., 2006). In addition, functional groups of lipids can covalently bind further molecules (Allard, 2006; Berthier et al., 2000) or encapsulate smaller molecules, leading to their preservation (Lichtfouse et al., 1998b; Piccolo, 2002; Sutton and Sposito, 2005). The long-term preservation of some lipid classes qualifies them as important biomarkers (Otto et al., 2005; White et al., 1997). N-alkanes are commonly assumed to be plant-derived (Kuhn et al., 2010; Lichtfouse, 1998) and are used as plant biomarkers to differentiate vegetation types (Bush and McInerney, accepted 2013; Schwark et al., 2002; Zech et al., 2012); long-chain fatty acids are used in a similar way (Wiesenberg and Schwark, 2006). More complex cutin-suberin-derived hydroxylated or poly-carboxylic acids enable aboveground litter input to be differentiated from belowground litter input (Mendez-Millan et al., 2011; Spielvogel et al., 2010). Sterols and other isoprenoid lipids, like terpenoids, are indicative for animal or plant-derived SOM (Otto and Wilde, 2001). Application of all these biomarkers tacitly assumes that they remain unmodified in soils over long periods. However, microorganisms are able to use lipids as substrates and decompose them to metabolites as well as build up their own lipids, therefore signifi- Publications and Manuscripts 240 cantly contributing to the lipid pool of SOM and the modification of initial lipidic compounds (Lichtfouse et al., 1995; Otto et al., 2005). This indicates already that the concept of generally untransformed lipid biomarkers is likely to be too simplified. Recent studies indicate that plant-derived biomarkers can be modified and overprinted by rhizomicrobial activity (Gocke et al., in press). For some biomarker classes like sterols microbial modifications of plant or animal-derived lipids are specifically used to trace the microbial community impact (Arima et al., 1969; Bull et al., 2002; Bull et al., 1999). For others, like alkanes, approaches used to correct for microbial overprint of plant-derived signals have been developed (Buggle et al., 2010; Zech et al., 2013). However, current knowledge on the microbial transformation of lipids in soils is rare and new experimental studies are needed. 14 C age of microbial lipid biomarkers revealed that they are not indicative for the age of this substance class, as microorganisms obviously use old lipidic substrates to build up their new lipids, e.g. phospholipid fatty acids (PLFA) (Rethemeyer et al., 2004). However, it is not yet clear whether microorganisms prefer the new synthesis of their lipids from low molecular weight precursors like acetate or whether they prefer to use available lipidic compounds, e.g. fatty acids, and simply modify them. According to the biochemical principle of the most economic pathways, cells tend to use preformed building blocks for biomass synthesis (Lengeler et al., 1999). Consequently, we hypothesize that lipids released by the decomposition of plant or microbial biomass should be the preferred substrates for further lipid synthesis by microorganisms. However, it is known that microbial polar lipids, like PLFA, are degraded to fatty acids after cell death (Lichtfouse et al., 1995) and, consequently, re-contribute to the free lipid pool in soils. The knowledge gap concerning the soil lipid cycle has caused an intensive discussion concerning the contribution of plant versus microbial lipids (Lichtfouse, 1998; Lichtfouse et al., 1995; Otto et al., 2005). Therefore, microbial modification of lipid biomarkers in soil, as well the transformation processes, are crucial for the application of biomarker fingerprints as well as isotope data from lipid biomarkers. Therefore, we traced the microbial formation of membrane lipids, PLFA, from their low molecular weight organic precursor – the acetate (Caspi et al., 2008; Keseler et al., 2009; Rock et al., 1981). In addition, we traced the utilization of the most abundant fatty acid – palmitic acid – as a microbial substrate for PLFA. Palmitic acid is a key compound for plant and microbial fatty acid metabolism. Investigating its microbial utilization and transformation pathways reveals a general view of the microbial modification of soil lipids. To elucidate the metabolic pathways of microbial fatty acids, we used the approach of position-specific labeling. This tool was originally derived from biochemistry to investigate metabolism pathways and has rarely been applied in soil science (Fokin et al., 1993, Publications and Manuscripts 241 , 1994; Haider and Martin, 1975; Kuzyakov, 1997; Nasholm et al., 2001). However, within the last years, increasing interest in the use of position-specific labeling to assess metabolic pathways in soil has arisen (Apostel et al., 2013; Dijkstra et al., 2011a; Dijkstra et al., 2011b; Dippold and Kuzyakov, 2013; Fischer and Kuzyakov, 2010b). This is because this approach is the only one which enables the fate of individual C positions to be traced through various pools or metabolites and consequently allows the reconstruction of individual transformation steps. Knowledge about fatty acid synthesis by microorganisms is mainly derived from experiments with pure cultures (Lennarz, 1970; Rock et al., 1981; Zelles et al., 1995). They can be newly synthesized from precursors like acetate or available lipid precursors can be modified by them (Lennarz, 1970; Rethemeyer et al., 2004; Rhead et al., 1971). If palmitate is given as a substrate, there are three possible mechanisms by which palmitate C can be used for PLFA synthesis: 1) the resynthesis pathway i.e. the complete degradation of the molecule to acetyl-CoA units and the following reconstruction of new fatty acids from C2-mojieties (Rhead et al., 1971); 2) Partial step-by-step degradation of the C2-units without total breakdown of palmitate: subsequently, only parts of the molecule are incorporated into newly formed fatty acids (Rhead et al., 1971); and 3) the untransformed utilization of palmitate as it is the most abundant fatty acid in microorganisms (Rhead et al., 1971; Zelles et al., 1995). Position-specific 13 C labeling enables these three pathways to be distinguished and to evaluate the transformation of the straight chain, unsaturated palmitate. This approach will deepen the understanding of the transformation of fatty acids and other lipidic biomarkers and improve the interpretation of fatty acid fingerprints in soils. 2.8.2 Material and Methods 2.8.2.1 Experimental Site The field experiment is located in Bavaria, close to Hohenpölz (49.907 N, 11.152 E) with 501 m.a.s.l, mean annual temperature 6.7 °C and mean annual precipitation of 874 mm. The agriculturally used field site is managed by a rotation of corn, barley, wheat and triticale. Soil type is a loamy Luvisol which has a pH KCl of 4.88, a pH H2O of 6.49, a TOC content of 1.77%, a TN content of 0.19% and a CEC of 13 cmol C kg -1 . Before the experiment started in August 2010 triticale, the last crop, was harvested, and the field site was grubbed for soil homogenization. Publications and Manuscripts 248 Fig. 2 Classes of phospholipid-derived fatty acids (top left) extracted from soil and recovery of position-specifically 13 C-labeled acetate (top right) and palmitate (bottom) in the different fatty acid classes 3 and 10 days after 13 C application. Experimental points (means ± SEM, N=4) are presented. Figure 2 shows that acetate was an appropriate precursor for the new formation of fatty acids: e.g. at day 3, the pattern of newly formed fatty acids from acetate C-2 was already quite similar to the PLFA profile in soils (Figure 2). Only branched fatty acids smaller than 16 C atoms were not formed from acetate, not even from acetate C-2. In general, acetate C-1 incorporation into each fatty acid was lower than C-2, reflecting that not only intact acetate was used as a precursor in fatty acid synthesis. Especially in branched fatty acids, a preferential incorporation of C-2 was observed (Figure 2). The C1 from acetate was not incorporated in any of the odd-numbered fatty acids (Figure 2). Palmitate 13 C incorporation into phospholipid-bound palmitate is higher than acetate 13 C incorporation (Figure 2). C-1 of palmitate is incorporated only in negligible amounts into fatty acids which were shorter than 16 carbons (Figure 2 and Figure 3). In addition, the terminal C-1 position of palmitate was not incorporated into odd-numbered PLFAs. This indicates the preference to use palmitate as a direct precursor for microbial phospholipid synthesis. Publications and Manuscripts 249 Fig. 3 Fingerprint of phospholipids-derived fatty acids in soil (top left) and relative recovery of individual positions from palmitate 13 C in fatty acid classes. Experimental points (means ± SEM, N=4) are presented. Figure 3 presents the transformations of incorporated palmitate by microbial PLFA formation. The portion of desaturated fatty acids was already high at day 3 and did not increase significantly from day 3 to day 10 (Figure 3). This shows a fast desaturation of the palmitate precursor. However, incorporation of palmitate 13 C into elongated and even more into branched fatty acids significantly increased from day 3 to day 10 for each of the C positions, reflecting slower kinetics of these processes. In general, a modification of palmitate according to the demand of the microbial community could be observed in this study. After 10 days the fingerprint of newly formed fatty acids already closely approached the PLFA distribution of the present microbial community. However, individual transformation steps occurred with different kinetics. 2.8.3.3 Incorporation of acetate and palmitate 13 C into PLFAs of individual microbial groups Preference for acetate and palmitate strongly differed for individual microbial groups. The PCA, based on the amounts of fatty acids, revealed two groups of gram negatives: whereas gram-negative 1 (18:1w9c, 18:1w7c, 14:1w5c) showed the highest incorporation of acetate, gram-negatives 2 (16:1w7c, cy19:0) reflected the highest uptake of palmitate (Figure 4). In general, the low molecular weight acetate was a better substrate for prokaryotic groups than for eukaryotic fungi or protozoa. A similar pattern was not observed for palmitate, which was preferentially used by prokaryotic gram-negatives and eukaryotic fungi and protozoa (Figure 4). The amount of incorporated 13 C decreased from day 3 to day 10 for each of the microbial groups and both carboxylic acids (Figure 4, the only exception is incorporation of palmitate into actinomycetes). Publications and Manuscripts 250 Fig. 4 Recovery of applied 13 C from positions of acetate (top) and palmitate (bottom) in microbial groups after 3 and 10 days. Experimental points (means ± SEM, N=4) are presented. Significant differences of incorporation of individual positions and incorporation between the days, calculated by nested ANOVA, are presented in Supplementary, Table A4 Specifics in the acetate and palmitate transformations are more visible if the divergence index (DI) is considered rather than the absolute 13 C incorporation (Figure 5). For acetate, each of the microbial groups showed the preferential incorporation of C-2. However, the discrimination between C-1 and C-2 was lowest for gram-negative groups (who had the highest absolute incorporation) and highest for eukaryotic groups (who had the lowest acetate 13 C incorporation) (Figure 5). The DI of palmitate did not show similar trends: each microbial group had individual preferences for incorporation of palmitate Publications and Manuscripts 251 positions into their PLFAs. Similar to acetate, the discrimination between positions was lowest for the two gram negative groups. For many microbial groups, the position-specific preferences and discrimination between positions strongly changed from day 3 to 10. This reflects an intensive turnover of palmitate 13 C, even if incorporated into PLFA. Fig. 5 Divergence index (DI) reflecting discrimination between C positions by incorporation into individual microbial groups 3 (left) and 10 (right) days after application of 13 C-labeled acetate (top) and palmitate (bottom). Experimental points (means ± SEM, N=4) are presented. Significant effects of C position and day on DI, calculated by nested ANOVA, are presented in Supplementary Table A5. Letters indicate significant differences (p < 0.05 derived from HSD post-hoc test) in the relative incorporation of the C positions into one group. Publications and Manuscripts 252 2.8.4 Discussion 2.8.4.1 Utilization and turnover of acetate and palmitate by soil microbial community The short-chain low molecular weight organic acids are a well-used microbial substrate (Jones et al., 2003). Our study shows that long chain carboxylic acids like palmitate are also good substrate in soils and are used in similar proportions by the microbial community (Figure 1). In specific pathways, such long chain carboxylic acids can function as direct precursors for lipid formation, e.g. that of PLFA. Then, their incorporation into microorganisms can even exceed those of low molecular weight substances (Figure 1). The preferential oxidation of C-1 of acetate is in accordance with previous studies and can clearly be linked to microbial metabolism, i.e. the oxidation of acetate in the citric acid cycle (Dippold and Kuzyakov, 2013; Fischer and Kuzyakov, 2010b) (Figure 6). A similar preferential oxidation of C-1 was observed for palmitate, especially after 10 days. This reflects that if palmitate is used in basic C metabolism, e.g. as an energy source, it is successively oxidized by fatty acid β -oxidation to acetyl-CoA (2 C atoms) units (Caspi et al., 2008; Keseler et al., 2009). Consequently, terminal C-1 and C-2 from palmitate form an acetate unit, and are transformed similarly to acetate in basic C metabolism. Fig. 6 Metabolic pathways of fatty acid formation from acetate and fatty acid transformations of palmitate in soil. At day 10, a higher portion of the 13 C remaining in soil is found in microbial biomass for acetate than for palmiate (p<0.05). A higher portion of the 13 C in microbial biomass for acetate than for palmiate (p<0.05) clearly shows the lower microbial availability of palmitate. Consequently, a higher relative proportion of the added 13 C remained extracellular, e.g. as SOM-associated palmitate. Nevertheless, a high discrimination between C-1 and C-2 was observed, which was even higher for palmitate than for acetate at day 10. This indicates that in addition to the microbial transformed palmitate also palmitate stabilized in soil gets transformed by terminal oxidation. This terminal oxidation of carboxylic acids Publications and Manuscripts 253 to odd and even alkanes has been previously described for plants and microorganisms (Dennis and Kolattukudy, 1992; Ladygina et al., 2006; Park, 2005), and specific as well as unspecific decarboxylases contribute to the decarboxylation of carboxylic acids in soils (Hofrichter et al., 1998). Extracellular transformations are known to be less relevant for well available, low molecular weight organic substances (Dippold and Kuzyakov, 2013). However, their quantitative relevance for hydrophobic substances, such as palmitate, which could be stabilized by hydrophobic interactions in soils and are consequently less available for microbial uptake, still remains open. To finally identify extracellular, terminal oxidation, selective inhibition of microbial, intracellular processes coupled with positionspecific lipid 13 C labeling has to be performed (Dippold and Kuzyakov, 2013). 2.8.4.2 Pathways of fatty acid formation from acetate in soil Whereas in microorganisms the C-2 from acetate is preferentially incorporated compared to C-1, this clear pattern is less expressed if acetate C is used for PLFA synthesis (Figure 1). This shows that many of the microbial compounds in cytosol (released by the chloroform-fumigation-extraction) are small water soluble products like carboxylic acids or nucleotides derived from citric acid cycle metabolites. During the citric acid cycle the C-1 of acetate gets oxidized in an early step, whereas several cycles are needed until acetate C-2 gets oxidized (Figure 6). However, acetate is a direct precursor for fatty acid synthesis, which is built up from the C2-unit acetyl CoA (Caspi et al., 2008; Keseler et al., 2009; Lengeler et al., 1999). Therefore, the direct formation of fatty acids from acetate would lead to an identical incorporation of both positions (Figure 5). This identical incorporation was not observed in this study. Instead, there are clear specifics for C-1 and C-2 incorporation into individual PLFAs (Figure 2b): The lower incorporation of C-1 into basic, straight chain fatty acids like palmitate can be explained by the utilization of already partially oxidized fragments of acetyl CoA for fatty acid synthesis. This is similar to the metabolism of glucose (Dippold et al., submitted). Although glucose is a direct precursor for amino sugar synthesis, glucose molecules were transferred into basic glucose C metabolism and only fragments of the partially oxidized and split molecule were subsequently used for amino sugar synthesis. Similar bidirectional pathways were observed for carbohydrate synthesis: formation of microbial sugars from glycine occurred parallel to direct formation from glucose and in parallel to glucose oxidation (Derrien et al., 2007). Such bidirectional pathways can also explain the C positions used for fatty acid formation in this study: acetate was partially oxidized by the citric acid cycle and fragments, only containing C-2, were transferred back from citric acid cycle metabolites towards acetyl-CoA for new fatty acid synthesis. Publications and Manuscripts 254 An alternative explanation would be that the majority of fatty acids are not newly formed from acetyl-CoA (equal incorporation of C-1 and C-2). Instead, acetate 13 C was only used to perform transformation and modifications at already existing fatty acids, e.g. elongations of partially degraded fatty acids or the introduction of branching points into molecules. Figure 2b shows that no acetate C-1 is incorporated in any odd fatty acid: this is rather unlikely, if fatty acids with 15, 17 or 19 C are newly formed from acetyl-CoA (using 7, 8 or 9 units of acetyl-CoA). However, if only a terminal acetyl-CoA is added and afterwards the terminal carboxylic group is split to reach an odd fatty acid, this would cause a low incorporation of C-2 and the absence of C-1 incorporation into odd fatty acids (Figure 2b). For non-growing microbial communities under maintenance conditions, like in this study (see Table 1), internal recycling of fatty acids is likely to occur: the utilization of direct precursors contributes to save energy and C (Lengeler et al., 1999). However, this can only finally be proven, if not only position-specific labeling but also position-specific detection of the isotopic label in PLFAs is performed. 2.8.4.3 Pathways of fatty acid transformations in soils The higher absolute incorporation of palmitate C compared to acetate C for the formation of PLFA showed that the more complex and direct precursor palmitate is preferred for synthesis and acetate is preferred for catabolism. This suggests already that palmitate was not fully degraded to acetyl-CoA and new fatty acids were built up from acetyl-CoA according to the resynthesis pathway (Rhead et al., 1971). Instead it is likely that modification of intact palmitate occurred. Figure 4 shows that the initially added 13 C palmitate is successively transformed to more diverse spectra of fatty acids. Comparing those transformed fatty acids with the PLFA fingerprint of the soil (Figure 3) shows that over a period of 10 days the newly transformed fatty acids are approaching the composition and consequently the demand of the microbial community. However, different kinetics of transformations are clearly shown by figure 3: Simple desaturation of palmitate occurred rapidly during the first 3 days after labeling. Thereafter, the proportion of desaturated fatty acids only marginally increased. More complex, biochemical processes, like elongations or even more branchings, occurred more slowly and therefore the proportion of these fatty acids strongly increased after day 3 . However, the use of intact palmitate and the following modifications are in accordance with high recycling of existing fatty acids in soil microorganisms observed by acetate 13 C (Figure 2). This is confirmed by the fate of individual palmitate positions: 1) there is almost no C-1 and C-2 incorporated in even numbered fatty acids smaller than palmitate, e.g. C14 fatty acid (tetradecanoic acid). This suggests that the terminal acetate (C-1 Publications and Manuscripts 255 and C-2) of palmitate is just split off to form the C14 fatty acid, whereas the basic C skeleton containing C-16 remained intact. 2) There is no palmitate C-1 in odd numbered fatty acids, suggesting that the terminal C-1 is only oxidized during the formation of odd numbered fatty acids from even numbered palmitate. 3) C-1, C-2 and C-16 are incorporated in similar amounts in desaturated fatty acids (Figure 2c). This suggests that the unsaturated, straight chain palmitic acid is just desaturated – or elongated and desaturated – for the formation of desaturated C16 and C18 fatty acids. Position-specific 13 C labeling cannot distinguish whether these modifications occur as free fatty acids or bound to the PLFA. Whereas elongations and shortenings need to occur with a free, non-esterified terminal carboxylic group (Caspi et al., 2008), modifications like 10-methyl branching, cyclization or desaturation are known to be possible if fatty acids are bound to a PLFA (Aguilar et al., 1998; Lennarz, 1970). To distinguish transformations of free fatty acids from those occurring bound in PLFAs within the membrane, the measurement of isotopic label in intact phospholipids and free fatty acids has to be performed with much shorter time intervals than those chosen for this study. However, irrespective of the detailed biochemical mechanism, this study proved: 1) an intensive modification and use of intact fatty acids taken up from soil, and 2) an intensive recycling of the microbial fatty acid pool (which can occur intracellular or intercellular after cell death). Therefore, the previous assumption that fatty acids are generally newly formed from the added substrates have to be discarded and future PLFA studies, because they have to consider the reuse of existing plant and microbial-derived fatty acids (see section 4.5). 2.8.4.4 Pathways of specific microbial groups in soils For the example of acetate, Figure 3 reflects the classical use of LMWOS by individual microbial groups in soils: gram-negatives are known to be the dominating group in the rhizosphere (Soderberg et al., 2004; Tian et al., 2013) and are most competitive for LMWOS (Apostel et al, 2013, Gunina et al, submitted). In contrast, gram-positives prefer old SOM (Kramer and Gleixner, 2006) and are less competitive for LMWOS. In general, the more complex organisms are structured, the lower their turnover and their competitiveness for fast uptake of LMWOS is: Bacteria have a shorter generation time (bacteria 20 min versus fungi 4-8 h to complete a life cycle under optimal conditions) and consequently higher cellular turnover (bacteria 2-3 times and fungi 0.75 times biomass turnover per year under soil conditions) (Moore et al., 2005; Rousk and Baath, 2007; Waring et al., 2013). The turnover is even slower for higher levels of the nutritional net, e.g. the protozoa. Consequently, the results for acetate utilization in this study confirm that fast growth Publications and Manuscripts 256 is based on the utilization of readily available substrates and is closely associated with a fast turnover of the respective microbial groups in the soil. Microbial groups with a faster turnover are commonly more competitive for LMWOS, like acetate, even if the microbial community in general is under maintenance conditions (Gunina et al., submitted). However, such a general rule is not valid for more complex, not ubiquitous substrates like palmitate: Figure 3 shows that, even within the gram-negatives, there is a clear preference of palmitate utilization by the gram-negatives 2. Separation of the two groups of gram-negatives was based on different groupings of the respective gramnegative fatty acids by explorative statistical tools (here: loadings of fatty acid contents on different factors in a principle component analysis). This tool, commonly used to characterize fatty acid fingerprints, supported here the separation and identification of two ecophysiologically different groups of gram-negatives: gram-negative 1 with a preference for LMWOS and gram-negative 2 with a preference for more complex, hydrophobic carboxylic acids (Figure 4). Not only the absolute uptake but also the metabolism was specific for the investigated microbial groups. The Divergence Index revealed the preference for acetate C-2 incorporation for each of the microbial groups. However, discrimination between C-1 and C-2 increased significantly for those groups with high LMWOS, i.e. acetate, uptake. (Figure 4 and Figure 5). This can have two possible reasons: 1) Fast growing microbial groups with rapid turnover are characterized by a more direct metabolismn using precursors without further transformation, and 2) The fast growing gram-negatives are mainly characterized by straight chain, monounsaturated C16 and C18 fatty acids (Zelles, 1999), which are formed by simple desaturation without complex metabolic processes like methylations or branchings (Lennarz, 1970) leading to discriminations between C-1 and C-2. In both cases, it can be concluded – at least for the PLFA formation pathway - that more direct, less complex metabolic pathways are characteristic for fast growing microbial groups with high turnover. This is also confirmed for palmitate incorporation into PLFA, where the gram-negatives showed a comparatively low discrimination between the palmitate positions (Figure 5). However, to prove these general or microbial group-specific transformation steps, a combination of position-specific labeling with position-specific analysis of the microbial transformation products is needed. Nevertheless, the strong difference in DI from day 3 to day 10 for palmitate confirms that a high, internal turnover e.g. by recycling and transformation of the fatty acids, took place after 13 C incorporation. Publications and Manuscripts 257 2.8.4.5 Consequences for the application of fatty acids as biomarkers The observed transformation of free fatty acids in soil by microorganisms causes consequences for the application of fatty acids as microbial and plant biomarkers. Alkanes can more easily be distinguished between plantand microbial-derived n-alkanes. This enables the alkane fingerprint to be corrected for microbial contribution (Buggle et al., 2010; Zech et al., 2013). In contrast, the differentiation between microbial and plantderived fatty acids is not as sharp: Vegetation type as well as microbial community affect the fatty acid sources in soils (Otto et al., 2005) and, in many cases, a reconstruction of the original source is not possible (Gocke et al., 2014). Furthermore, it is not clear, whether microbial enzymatic systems modifying n-alkanoic acids like palmitate are highly specific enzymes, which work only intracellularly or whether unspecific modification of plant-derived free fatty acids can occur. This would strongly limit the application of fatty acid fingerprints (Zhou et al., 2005) as well as their isotope signatures (Li et al., 2011) for paleo-environmental reconstructions. Therefore, further investigations, e.g. positionspecific labeling of long-chain plant-derived fatty acids and investigation of their microbial transformations, is needed. The transformation and internal recycling of fatty acids within microbial cells has important consequences for its application as microbial biomarkers. Changes within the fatty acid fingerprint in soils are commonly assumed to be related to changes in the microbial community structure (Zelles, 1999). However, acetate as well as palmitate labeling showed in this study that fatty acids are transformed and modified very fast in soils. Pure culture studies confirm that these modifications of fatty acids occur within living cells, if environmental conditions surrounding a living organism are changing, e.g. by temperature changes (Aguilar et al., 1998). These modifications of existing fatty acids can even occur in intact PLFA within the membranes (Aguilar et al., 1998; Lennarz, 1970). Therefore, further knowledge about the impact of internal fatty acid turnover for the interpretation of the PLFA fingerprint is needed (Frostegard et al., 2011). However, the high internal turnover of fatty acids within living microbial cells explains the discrepancy between the turnover of PLFA and of that of microbial biomass. PLFAs are assumed to have a half-life between one day and one week (Kindler et al., 2009; Ranneklev and Baath, 2003; Rethemeyer et al., 2004). In contrast, the turnover of the bacterial microbial community is assumed to occur 2-3 times per year (Moore et al., 2005; Rousk and Baath, 2007; Waring et al., 2013). An intensive intracellular turnover of PLFA explains the much faster turnover of PLFA. This is similar to observations for the turnover of microbial cell walls (Dippold et. al, submitted): for E. coli it was even shown that they recycle 60% of their peptidoglycan during cellular life (Park and Uehara, 2008; Uehara and Park, 2008). Malik et al. (2013) showed that the turnover of microbial biomass compounds de- Publications and Manuscripts 264 Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the characterisation of microbial communities in soil: a review. Biology and Fertility of Soils 29, 111 - 129. Zelles, L., Bai, Q.Y., Rackwitz, R., Chadwick, D., Beese, F., 1995. Determination of phospholipid-derived and lipopolysaccharide-derived fatty acids as an estimate of microbial biomass and community structures in soils. Biology and Fertility of Soils 19, 115-123. Zhou, W.J., Xie, S.C., Meyers, P.A., Zheng, Y.H., 2005. Reconstruction of late glacial and Holocene climate evolution in southern China from geolipids and pollen in the Dingnan peat sequence. Organic Geochemistry 36, 1272-1284. Publications and Manuscripts 265 Supplementary Data Supplementary Table A1: Fatty acids in the external standard Supplementary Table A2: Fatty acids in the external standard Publications and Manuscripts 266 Supplementary Table A3: Result of factor analysis Supplementary Table A4: Nested ANOVA for acetate and palmitate positions nested in the variable day, block as random variable and day. Degrees of freedom (df), F-values and significance level (p) are shown for the acetate and palmitate. If requirements for parametric tests (normal distribution + homogeneity of variances was not given, a Kruskal-Wallis ANOVA for the individual treatments was calculated (in this case H-Value is given instead of F value) Publications and Manuscripts 267 Supplementary Table A5: Nested ANOVA for acetate and palmitate DI, with the independent variables position (being nested in the variable day), block (as random variable) and day. Degrees of freedom (df), F-values and significance level (p) are shown for the acetate and palmitate Publications and Manuscripts 268 2.9 Study 9: Organic nitrogen uptake by plants: Reevaluation by position-specific labeling of amino acids Daniel Moran-Zuloaga #1,2 , Michaela Dippold #1,2 , Bruno Glaser 3 Yakov Kuzyakov 2,4 , # equal contribution 1 Department of Agroecosystem Research, BayCEER, University of Bayreuth 2 Department of Agricultural Soil Science, University of Göttingen, Germany 3 Department of Soil Biogeochemistry, Institute of Agricultural and Nutritional Science, Martin-Luther University Halle-Wittenberg 4 Department of Soil Science of Temperate Ecosystems, Georg-August University of Göttingen Corresponding Author: Daniel Moran-Zuloaga Department of Agricultural Soil Science Georg-August-University of Goettingen Buesgenweg 2 37077 Goettingen Tel: 0541-3933546 e-mail: [email protected] Publications and Manuscripts 269 Abstract Current studies suggested that besides inorganic nitrogen (N), many plants are able to take up organic N in form of amino acids. However, reliable methods to quantify the uptake of intact amino acids are still missing and the relevance of organic N uptake is doubted. We used position-specific 14 C labeling to investigate the uptake of intact amino acids and their role in the N nutrition of plants. Position specifically 14 C and 15 N labeled alanine, injected into the rhizosphere soil, enabled to trace the uptake of C from individual molecule positions by Zea maize, Lupinus albus and Cichorium intybus . As a control, uniformly 14 C labeled alanine, acetate and inorganic 15 NH 4+ and 15 NO 3were applied. The same uptake of uniformly 14 C labeled alanine and acetate showed that low molecular weight organic substances are taken up by roots may occur by passive mechanisms, without differences for N containing and N free organics. Differences in plant uptake of 14 C from individual positions in alanine molecule confirmed that soil microorganisms split alanine within 6 h into transformation fragments (including mineral NH 4+ ), which were then taken up by plants. Only 0.04 to 0.25% of the alanine added directly into the rhizosphere were taken up as intact molecule with the highest uptake observed for lupine – the plant adapted to organic N transport from Rhizobia . Microbial utilization strongly dominated the fate of low molecular weight organic substances in soils and the majority of amino acid 14 C uptake by plants was explained by passive uptake of microbial transformation products. Position-specific labeling is an innovative tool that enables to separate easily the intact uptake from uptake of molecule fragments. Thus, it improves the quantification of intact uptake by avoiding the up to 3fold overestimation of uniform labeling approaches. Keywords: Alanine; Position-specific, dual isotope labelling; Organic N uptake; Chicory; Lupine; Maize; Isotopic approaches; Nitrogen cycle Publications and Manuscripts 270 2.9.1 Introduction Over the past century, many studies have emphasized the role of dissolved inorganic nitrogen (DIN) in ecosystems (Matson et al., 1997; Vitousek et al., 1997; Vitousek et al., 1979). Ammonium (NH 4+ ) and nitrate (NO 3- ) are the main representatives of mineral nitrogen. Ammonium is a reduced form of DIN and can be directly utilized by plants after uptake whereas nitrate needs to be reduced first. Nitrate reduction demands energy from plants (Doubnerova and Ryslava, 2011; Liu et al., 2011; Tischner, 2000) leading to additional CO 2 fluxes through the plant-soil system (Gavrichkova and Kuzyakov, 2008, , 2010). Both DIN species can be lost from ecosystems: nitrate by leaching into the ground water, denitrification to N 2 O and N 2 , or reduction to ammonium and ammonium can be lost by volatilization or irreversible fixation by soil minerals. In ecosystems with low availability of DIN due to slow mineralization, like boreal or arctic ecosystems (Nasholm et al., 1998; Vitousek et al., 1979), plants may also rely on other N forms such as dissolved organic nitrogen (DON). This is not only a short-circuit in the traditionally assumed N nutrition pathways (the mineralization to NH 4+ and NO 3is omitted), but also reduces potential N losses from ecosystems, e.g. by leaching. In the past twenty years, there has been remarkable interest in DON as a plant N source (Chapin et al., 1993; Jones et al., 2005a; Nasholm et al., 1998; PaungfooLonhienne et al., 2012; Schimel and Chapin, 1996). Organic N can be found in many compounds in soil from macromolecules like proteins (Jones et al., 2005d) or humic substances (Szajdak et al., 2003) to low molecular weight organic substances (LMWOS) like amino acids (Doerr et al., 2012; Jones et al., 2005c; Lipson et al., 1999; Streeter et al., 2000), amino sugars (Roberts et al., 2007; Roberts and Jones, 2012) and nucleic acids (Kuzyakov, 1996). Many amino acids have very fast cycling rates and the half-life of amino acid C in soils is in the range of few hours (Jones et al., 2009; Kuzyakov, 1996). This fast cycling is connected with fast and almost complete uptake by microorganisms (Fischer et al., 2007). Another study demonstrated that LMWOS at average soil concentrations in soil solution (below 10 µmol l -1 ) were taken up by microorganisms at a rate of 82% after 3 min (Fischer et al., 2010b), and the half-life of amino acids in soil solution ranges between 48 min (Jones et al., 2004). Due to this fast utilization, soil microorganisms are stronger competitors for amino acids than plants (Biernath et al., 2008; Hodge et al., 2000; Jones et al., 2005a; Kuzyakov and Xu, 2013b), whereas in the long-term this N is released by the microorganisms and is available for plants. In contrast, Chapin et al. (1993) showed in the early 90ies the preferential use of organic N by an arctic sedge which started the discussion about the relevance of amino acids as plant N source. Publications and Manuscripts 271 Further studies showed that boreal forest vegetation actively take up amino acids, probably due to a lack of other N sources (Delgado-Baquerizo et al., 2011; Nasholm et al., 1998). The parallel uptake of DIN and DON is dependent on their availability (Kranabetter et al., 2007). Therefore DON is discussed to be less relevant for agricultural crops (Jones et al., 2005a). In order to evaluate the relevance of DON, a comparison between LMWOS with inorganic N uptake was recommended (Glass et al., 2002; Jones et al., 2005a; Streeter et al., 2000) especially for agroecosystems, where the role of DON is still controversial. Isotope labeling of LMWOS with 15 N coupled with 13 C or 14 C is a common tool to investigate uptake and allocation in plants as well as mineralization or microbial incorporation (Thede, 2010; van Hees et al., 2005). The uptake of amino acids by plants was mainly investigated by dual-labeling with 15 N and 13 C (Nasholm et al., 1998; Streeter et al., 2000). It is tacitly assumed in this approach that the uptake of 13 C corresponds to the uptake of the intact amino acid. However, dual isotope labeling has a methodological shortcoming leading to an overestimation of intact uptake: microorganisms produce labeled fragments from the added amino acids, and these fragments and mineralized N can be taken up in parallel (Rasmussen et al., 2010). This would contribute to the quantified intact uptake by the dual isotope labeling approach (Sauheitl et al., 2009a). The first evaluation of this overestimation was performed by the application of dual uniformly labeled amino acids and compound-specific 13 C and 15 N analysis during their root uptake. It was shown that due to uptake of labeled metabolites, bulk measurements caused an up to six-fold overestimation of the intact uptake (Sauheitl et al., 2009a). However, compound-specific 13 C and 15 N analysis has the disadvantage of being a time-consuming and expensive technique (Sauheitl et al., 2009a). To prove the uncertainties of the original 13 C/ 15 N approach, Rasmussen et al. (2010) proposed position-specific labeling as a potential tool to overcome the problem of molecule splitting. Thus, uptake as a whole molecule could be distinguished from uptake as partially degraded amino acid fragments i.e. decarboxylated fragments (Dippold and Kuzyakov, 2013). Some recent studies (Dijkstra et al., 2011a; Fischer and Kuzyakov, 2010) clearly showed that position-specific 13 C and 14 C labeling enables tracing the fate of individual functional groups in various soil pools. If the uptake of amino acid C occurs as a broad spectra of various transformation products (and not as intact amino acids), this would strongly reduce the importance of N nutrition by amino acid – from a quantitative as well as a regulative view concerning N deficiency. Here, we used the same technique of position-specific 14 C labeling to quantify the intact uptake of amino acid. We hypothesized that 1) the original 13 C/ 15 N approach over- Publications and Manuscripts 272 estimates the intact uptake of amino acids, and 2) organic N uptake is traceable in temperate ecosystems but is of minor relevance for the N nutrition of agricultural plants. In order to consider the physiological differences of plant functional types (Weigelt et al., 2005), we performed our experiment with three species: maize, chicory and lupine. These species differ in their N uptake and transformation, their physiology and morphology, especially in the root system: 1) the grass maize ( Zea mays L .) has a fibrous root system and reduces NO 3in roots and shoots (He et al., 2011), 2) the herb chicory ( Cichorium intybus L. ) reduces NO 3in roots (Goupil et al., 1998) and has a taproot system, where it can store N-containing compounds for the next year (Ameziane et al., 1997) and 3) the legume lupine ( Lupinus albus L. ) reduces NO 3in roots (Gavrichkova and Kuzyakov, 2008) and has the ability to reduce atmospheric N 2 in root nodules through symbiosis with Rhizobia . As organic N source, we used alanine as one of the most abundant amino acids (Fischer et al., 2010a) and ammonia and nitrate as inorganic N sources. To evaluate the preference of amino acid uptake compared to N-free LMWOS, we included additional treatments with acetate, which has a structural resemblance to alanine. If uptake of NLMWOS (alanine) occurs mainly by unselective mechanisms, it should be in a similar range to N-free LMWOS (acetate). The aims of this study were: 1) to determine the fate of amino acids in soil with a special focus on the plant uptake of an initial substance versus the uptake of its transformation products, 2) to assess the ecological and physiological role of intact uptake of amino acids by different plant species and 3) to evaluate the relevance of three N sources (alanine, ammonium and nitrate) for N nutrition of agricultural plants. 2.9.2 Material and Methods 2.9.2.1 Experiment preparation Soil sampling Soil samples were collected from an agricultural field site close to Hohenpölz (Bavaria, Germany at 49.907 N, 11.152 E, 501 m.a.s.l.) that had been long-term cultivated with cereals (barley, wheat, triticale). The soil is a loamy haplic Luvisol (FAO, 2006). Soil was collected from 0-10 cm, sieved to 2 mm and roots were removed. The physicochemical characteristics of the soil are described in Table 1. Publications and Manuscripts 273 Plant and material preparation After sieving, soil was immediately filled into transfer pipettes made of low density polyethylene 30 cm in length and 1 cm diameter, which were used as rhizotubes (Biernath et al., 2008; Kuzyakov and Jones, 2006). We used maize ( Zea mays L), lupine ( Lupinus albus L) and chicory ( Cichorium intybus L). Plant seeds were pre-germinated at constant temperature (30 ºC) and watered for 36 hours (Gavrichkova and Kuzyakov, 2008). Then, one sprout of each plant was inserted into the rhizotubes. The rhizotubes were submerged in a plastic container halffilled with cold water to maintain the soil temperature around 12ºC. Thus, microbial activity e.g. mineralization rates should resemble field conditions (Jones, 1999). The pipette was connected with an air inlet (tube) at the bottom and directly under the soil surface (Biernath et al., 2008) to avoid water saturation of the soil and provide the soil and roots with air. Table 1 The physicochemical properties of the Ap-horizon of the haplic Luvisol. Soil parameters Values pH KCl 4.88 ± 0.12 pH H 2 O 6.49 ± 0.11 Total Organic Carbon 1.77 ± 0.07% Total Nitrogen 0.19 ± 0.01% Cation-Exchange Capacity 13.6 cmol c kg -1 soil Microbial biomass C 42.5 ± 1.1 µmol C g -1 soil Microbial C/N ratio 9.9 ± 0.3 Chemicals and radiochemicals The radiochemical stock solution had concentrations of 50 µM for alanine and acetate, both with 10 6 DPM ml -1 14 C activity. Position-specific labeled alanine ([114 C], [214 C], [314 C]alanine, American Radiolabeled Chemical Inc., St Louis, USA), as well as uniformly labeled [U14 C]acetate (Biotrend Köln, Germany) and [U14 C]alanine (American Radiolabeled Chemical Inc., St Louis, USA) were used. Nitrogen labeling was performed with a 99 atom-% 15 N enriched tracer of either alanine CH 3 CH( 15 NH 2 )COOH as the organic N-source or ammonium sulfate ( 15 NH 4 ) 2 SO 4 or potassium nitrate K 15 NO 3 as inorganic N forms (Biotrend Köln, Germany). Amount of applied C and N was identical in each treatment and lower than average concentrations of alanine, acetate, NH 4+ or NO 3in agricultural soils. Publications and Manuscripts 280 Plant species had no significant effect on the amount of mineralized 14 C (Figure. supplementary). Alanine showed significantly higher mineralization of C-1 (76%) than C-2 (45%) and C-3 (52%). In general, we observed that after 6 h, the individual molecule positions of alanine had strongly differing fates concerning plant uptake as well as the proportions remaining in the soil. 2.9.3.4 Intact uptake of alanine assessed by position-specific labeling The 14 C/ 15 N ratio in the plant biomass (shoots and roots) reflects the proportion of 14 C of each individual position, which was taken up together with 15 N. Based on positionspecific 14 C labeling, this calculation can be performed for each C position of alanine (Fig. 4). This ratio showed the pattern C-3>C-2>C-1 for each plant. We considered that a molecule of alanine could only be taken up intact if all three positions were incorporated into the plant. Thus, the minimum of the 14 C/ 15 N ratio reflects the maximum intact uptake of alanine in plants, which was the case for the 14 C/ 15 N ratio of C-1 position. These values were in a similar range for the three investigated plant species: 7 to 14% of the alanine-N was taken up as intact alanine in the order maize<chicory<lupine (Table 3). Fig. 4 Ratio of 14 C/ 15 N for individual alanine C positions incorporated in plant biomass. The alanine positions were C-1 (carboxyl group), C-2 (amino-bound group) and C-3 (methyl group). Letters indicate significant differences (p<0.001) between alanine C positions. Publications and Manuscripts 281 In order to compare the contribution of the three applied N sources, we estimated the tracer N nutrition budget. Comparing the role of alanine within the three investigated N sources, intact alanine uptake reached a maximum level of 0.25% of N. Lupine showed the highest N uptake in the form of intact alanine followed by chicory and maize (Table 3). The range of plant-specific relevance of intact alanine uptake (0.04-0.25%) reflected the plant-specific ability for N nutrition by organic sources. Table 3 Intact uptake of alanine by chicory, lupine and maize and estimated contribution of intact alanine uptake to total N nutrition of these plants with respect to the other N sources. Chicory Lupine Maize % 15 N uptake as intact alanine of total alanine-derived 15 N uptake 10.21 ± 3.48 13.70 ± 5.19 7.20 ± 4.70 % intact alanine of the three investigates N sources (alanine+ ammonium+ nitrate) 0.07 ± 0.04 0.25 ± 0.12 0.04 ± 0.02 Factor of overestimation of intact uptake based on uniform labeling 1.47 ± 0.25 1.14 ± 0.08 2.81 ± 1.89 The uptake of intact alanine reached a maximum of 13.7% of the total 15 N uptake from alanine (Table 3). The majority of the alanine molecules were metabolized within 6 h, when the initially organic-bound N was taken up as mineralized ammonium or even already oxidized to nitrate. This degradation of alanine as a percentage of the applied alanine is illustrated in Fig. 5. Intact alanine as well as mineralized alanine-derived N uptake was highest for lupine. Once fragmented, the uptake of C-1 was only half of that of C-3. This corresponds to the highest decomposition of C-1. This different fate of individual molecule positions demonstrates splitting of LMWOS which may have occurred in plant or soil. However, less than 1% of the alanine C was recovered in plants at all, and the majority of the alanine (~99%) remains in soil or microbial biomass. From the alanine fragments, 1.1% to 9.2% of the mineralized N was taken up by plants, whereas C incorporation in plants ranged only from 0.01 to 1.58% (Fig. 5). Consequently, only a small portion of the applied 14 C but a relatively higher portion of the applied 15 N was taken up by plants and incorporated into their biomass after 6 h. Publications and Manuscripts 282 Fig. 5 Illustration of the fate of alanine tracer molecules, which are either taken up intact or degraded/mineralized to fragments and subsequently incorporated into plant biomass or microorganisms. Microbial metabolism of alanine by microorganisms is adapted from Dippold & Kuzyakov (in press) 2.9.4 Discussion 2.9.4.1 Plant uptake of N-containing and N-free organic substances Our results showed no preferential uptake of 14 C from N-LMWOS alanine compared to 14 C from acetate – either taken up intact or as fragments - for any of the investigated plants. Biernath et al. (2008) found that maize had even higher uptake of acetate than alanine. This high uptake of acetate could mainly be attributed to passive uptake mechanisms (Rasmussen et al., 2010). As shown by Jones et al. (2005c) and Ge et al. (2009), a higher concentration of LMWOS and well-developed root systems increases plant competitiveness for LMWOS compared to microorganisms (Kuzyakov and Xu, 2013a; Xu et al., 2011). Stating a passive uptake means in this case the passive, unspecific transport of all LMWOS with the water flux towards the root surface, without any direct root- Publications and Manuscripts 283 specific regulation of amino acid transport. Specific conclusions about the contribution of various uptake systems at the plant surface cannot be stated from this study. If uptake is dominated by this passive flow of LMWOS towards the root surface, the bioavailability of alanine and acetate for plants would be the main driver for their uptake. Alanine can strongly interact with the soil matrix by its amino group, whereas acetate is less retained and consequently better available in the soil solution. In addition, microorganisms prefer alanine to acetate as a substrate (Fischer et al., 2010b; van Hees et al., 2002). We did not observe faster decomposition of acetate than alanine, but did not quantify incorporation into microbial biomass in this study. Alanine may have been preferentially incorporated by microorganisms, as previously observed by Fischer et al. (2010b), if the concentration of free alanine in the soil solution was lower than that of acetate. The combination of high root development and higher availability of acetate explains the higher uptake of acetate by maize under the dominance of passive flux of LMWOS towards the roots. Thus, a potential explanation from these results is that LMWOS are taken up by plants passively, irrespective whether they contain N or not. 2.9.4.2 Fate of functional groups of alanine in soil The loss of the carboxyl group by mineralization is higher than that of the methyl group in soil. Similar results were shown by Fischer & Kuzyakov (2010) for acetate, Nasholm et al. (2001) for glycine, Dijkstra et al. (2011a) for pyruvate, and Dippold & Kuzyakov (in press) for alanine. In all of these studies based on position-specific labeling, the mineralization of the carboxyl group was fastest compared to all other functional groups. Comparing remaining alanine14 C in soil after 6 h revealed the highest mineralization of C-1 with 68-70% followed by 30-45% and 34-52% for C-2 and C-3, respectively. This decomposition was even higher than that observed for 3 days in a field experiment being 89%, 49% and 29% for C-1, C-2 and C-3, respectively (Apostel et al., 2013). This higher mineralization reflects higher microbial activity under rhizosphere conditions compared to root-free soil (Blagodatskaya et al., 2009). The C-1 position is rapidly oxidized by decarboxylation of the C-1 group of pyruvate, the most abundant microbial transformation product of alanine, within the microbial metabolism. This is an extremely fast process in soil (Dippold and Kuzyakov, in press) and kinetics of microbial uptake and metabolization are known to be faster than plant uptake (Jones et al., 2005a). In contrast, methyl groups represent reduced C and do not need to be further reduced for many anabolic pathways (Apostel et al., 2013; Dijkstra et al., 2011b). Therefore, C-3 was less mineralized, preferentially incorporated into microbial Publications and Manuscripts 284 metabolites, also into those metabolites released by microorganisms into soil solution.. In addition, Dippold and Kuzyakov (2013) found a partial extracellular oxidation of alanine following the order C-1>C-2>C-3, which may also contribute to a higher amount of C-3fragments in soil. These fragments, if taken up by plants through passive mechanisms, cause the preferential incorporation of the C-3 position (Fig. 3 and Fig. 4). This preferential C-3 uptake as microbial metabolites seems to predominate the dark-fixation in the roots of microbially respired CO 2 (which would consequently have a C-1 enrichment). In summary, microbial uptake and utilization were the main processes affecting the fate of individual C positions of LMWOS in soil. Preferential oxidation of C-1 and preferential incorporation of C-3 by microorganisms are likely to explain the preferential loss of C-1 and accumulation of C-3 in the entire plant-soil system. 2.9.4.3 Allocation and transformation of C and N within plants The preference of crop plants for NO 3uptake has been reported in many studies (Ge et al., 2008; Hermans et al., 2006; Jones et al., 2005a) and was also confirmed in this experiment. When nitrate was used as the N source, maize reduced NO 3in shoots and in roots (Gavrichkova and Kuzyakov, 2008); lupine as well as chicory reduced NO 3mainly in roots (Gavrichkova and Kuzyakov, 2008; Goupil et al., 1998; Pate et al., 1981). Our results corroborate these findings, as the highest NO 3transport into shoots could be found in maize with a 15 N shoot/root ratio of 2.42 and lower ratios in chicory (1.35) and lupine (0.57) (Table 2). Low N allocation into shoots for chicory was also found by Ameziane et al. (1997): 8 days after labeling, chicory kept the majority of 15 N in its roots (shoot/root ratio 0.3). Reduced N sources like NH 4+ or alanine showed no clear preference for allocation from root to shoot. Svennerstam et al. (2007) found that incorporation of amino acids after intact uptake occurred as intact molecules, but could not prove this assumption as they neither labeled position-specific nor used compound-specific isotope analysis (CSIA) to measure plant amino acids. If intensive metabolization of amino acids in plants occurred, this would lead to a preferential decarboxylation and loss of C-1 and preferential incorporation of C-3, as observed in our study. Consequently, this approach would lead to an underestimation of intact uptake. There are not many studies investigating the transformation of amino acids taken up by plants by means of CSIA. The intact incorporation without further transformation was first shown by Persson and Nasholm (2001) by GC-MS. Sauheitl et al. (2009a), who performed similar experiments with GC-C-IRMS, also found no indication for oxidation of incorporated amino acids within the plant metabolism. Both studies excluded transforma- Publications and Manuscripts 285 tion of the C backbone of amino acids into other amino acids, but not into other metabolic products. Total 14 C and 15 N uptake reflected that only minor portion of the amino acids is taken up and consequently could be metabolized by plants. However, there is a remaining uncertainty of the effect of plant metabolism, which may contribute to an underestimation of the calculated intact uptake if preferential C-1 oxidation occurred in plants. Position-specific labeling in this experiment provided the first information about plant transformation of alanine by comparing the fate of individual molecule positions within the plant compartments. In lupine (Fig. 2), position C-1 was preferentially kept in the root and from position C-1 to C-3, an increasing allocation from root to shoot could be observed. Hence, either different fragments of alanine were allocated differently within the plant or intact alanine was partially split during 6 h by the plant metabolism. Ge et al. (2008) and Warren et al. (2012) found that amino acids can be transformed to other compounds to be transported to shoots. However, Warren et al (2012) and Sauheitl et al. (2009a) also indicate that transaminations are the most likely metabolic transformation within plants and that oxidation of the C skeleton is less likely. The 14 C uptake by chicory was too low compared to variations between repetitions to detect a comparable position-specific trend. Maize showed increasing amounts of 14 C from alanine from C-1 to C-3 for shoots and roots. This could either result from a preferential oxidation of C-1 and C-2 after intact uptake or from a preferred uptake of C-3 fragments and their allocation into the shoots without transformation. In summary, our results show that even if intact uptake occurs, plants tend to transform LMWOS rather quickly in their metabolism (Wegener et al., 2010) but mainly by transamination (Sauheitl et al., 2009a). The molecular nature of the newly formed metabolites can only be clarified by CSIA of the transformation products. 2.9.4.4 Intact uptake of alanine in plants Physiological ability of intact alanine uptake by plants was shown by Svennerstam et al. (2007) who identified lysine histidine transporter 1 (LHT1) as a facilitator for amino acid uptake (lysine, glycine and alanine) by the roots of Arabidopsis thaliana . Many studies evaluated the relevance of N nutrition by intact amino acid uptake under natural soil conditions by using dual-isotope but uniformly labeled 13 Cand 15 N-tracers (Bardgett et al., 2003; Nasholm and Persson, 2001; Weigelt et al., 2003). Calculating the 14 C/ 15 N ratio of plant uptake (Fig. 4) is based on this approach (Nasholm et al., 1998) and reflects the intact alanine uptake. If we would average our alanine C positions, which correspond to the uniform labeling approach, we would detect intact uptake of around 15-18% of alanine-derived N without species-specific differences (Fig. 4). Calculating the uptake of Publications and Manuscripts 286 intact alanine based on the C-1 position, i.e. the position with the lowest uptake, gives values of intact uptake of 7-14% of alanine-derived N. This demonstrates a 1.2 to 3-fold overestimation, if the calculation is based on uniform labeling results compared to position-specific labeling. In general, data for the highest intact uptake were found in boreal forests. This was partly explained by their nutrition via ecto-mycorrhization. However, many studies with grassland species and annual herbs also showed higher intact amino acid uptake than those observed in this study. This is a result of the methodological shortcomings of the uniform 13 C or 14 C labeling approaches. The use of position-specific labeling enables us to distinguish fragment uptake from whole molecule uptake and consequently demonstrates much lower uptake. Rasmussen et al. (2010) expected the highest plant uptake of the C-1 position. They postulated that the high mineralization of C-1 leads to an increase in HCO 3from C1 in the soil solution, which can be passively taken up by plants (Demidchik and Maathuis, 2007). Our results contradict this concept as we observed the highest incorporation rate with C-3. Thus, irrespective of the soil pH, a fast exchange of mineralized H 14 CO 3with atmospheric CO 2 leads to fast 14 C losses from mineralized molecule positions. The highest uptake of C-3 supports the idea of plant uptake of molecule fragments, i.e. microbial transformation products, by passive uptake mechanisms. In contrast, position-specific C-2-labeling revealed that ~20% of the glycine-derived N was taken up as the intact amino acid by Triticum aestivum (Nasholm et al., 2001). However, C-2 of glycine as a methyl group resembles C-3 of alanine which had the highest uptake. This suggests that labeling of reduced C positions (Nasholm et al., 2001) is likely to cause an overestimation of the real intact amino acid uptake. In addition, intact uptake of glycine may be facilitated compared to alanine due to decreased competitiveness of soil microorganisms for glycine (Hocking and Jeffery, 2004). In addition, its smaller molecular weight facilitates passive uptake. The applied amino acid concentration can be another aspect to explain the higher range of intact uptake observed in many previous studies. For example, Nasholm et al. (2001) applied a 1 mM tracer solution, whereas we used a much lower concentration of 50 µM. An increased amino acid concentration improves plant competitiveness due to early saturation of microbial amino acid transporters (Kuzyakov and Xu, 2013a). Thus, amino acid uptake quantified at high concentrations may not resemble natural conditions as free amino acid concentrations rarely exceed 100 µM in soils (Jones and Willett, 2006) and bioavailable amino acid concentrations are even lower (Hobbie and Hobbie, 2012). After glycine application to Plantago lanceolata , Sauheitl et al. (2009a) quantified intact uptake around 16.5% of glycine-derived N using 13 Cand 15 N-CSIA of amino acids. Publications and Manuscripts 287 This percentage is slightly above the values quantified here by position-specific 14 C labeling but up to 6fold lower than values gained by bulk isotope analysis. This confirms the overestimation of intact uptake gained by uniformly-labeling with bulk isotope analysis approaches. We also found significant species-specific differences in the proportion of intact alanine15 N to mineralized alanine15 N uptake (Table 3). Lupine had the highest uptake of intact alanine followed by chicory and maize (Table 3). Maize is known to take up either amino acids or their degradation fragments (Adamczyk et al., 2012; Godlewski and Adamczyk, 2007). The highest position-specific differences measured in our study reveals that mainly microbially transformed C-3 fragments of alanine were taken up. In contrast, lupine had a high total incorporation of alanine-derived N as well as high uptake of intact alanine. This can be attributed to its cluster roots (Hawkins et al., 2005) and to the very efficient amino acid transport systems, characteristic for legumes to facilitate transfer from the nodules of rhizobia (Day et al., 2001). Thus, plant ecophysiological characteristics can increase their chances to gain organic N. In summary, the use of position-specific 13 C or 14 C labeling improved the quantification of intact uptake of amino acids by plants by revealing the contribution of fragment uptake. The highly efficient microbial competition for alanine decreases the intact uptake by plant roots. 2.9.4.5 Relevance of amino acids as a N source for agricultural plants Within the three applied N sources, nitrate was preferred by the three plants irrespective of their ecophysiology. This preference of crops for nitrate has been shown in previous studies (Gavrichkova and Kuzyakov, 2008; Ge et al., 2009; Glass et al., 2002; Jamtgard et al., 2008) and is consistent with the soil properties in this study: The Luvisol, developed from loess, contains clay minerals (mainly illites) which can fix NH 4+ and cause lower plant availability of cationic nutrients. Species specific preferences for N sources are in accordance with previous studies in grasslands (Weigelt et al., 2005): fast growing species – in our study maize – showed the highest uptake of nitrate. The uptake of alanine15 N was in the same range as ammonium15 N. This indicates that presumably the majority of alanine15 N was very fast mineralized to and taken up as ammonium which is confirmed by results of a previous study with tundra species (Schimel and Chapin, 1996).. Thus, 15 N uptake confirms the position-specific 14 C results (Fig. 3) that mainly partially metabolized or mineralized fragments are taken up. Also, other studies have demonstrated fast transformation of N-containing LMWOS: Jone s et al. (2004) determined amino acid half-lives of 4 to 8 min in soil solution. Thus, within one Publications and Manuscripts 288 hour, applied amino acids are completely removed from soil solution and either incorporated into soil microorganisms, mineralized to ammonia or irreversibly fixed by the soil matrix. Table 1 shows a small C:N-ratio of the microbial biomass (~9.9) and thus a low N demand of the microbial community. Hence, the main fate of microorganisms using NLMWOS is the C skeleton. A similar strategy of microorganisms was observed for Pcontaining LMWOS in P-rich soils (Spohn and Kuzyakov, 2013). Thus, the majority of alanine15 N will be mineralized, released as 15 NH 4+ and then be available for plant uptake. Therefore, the N mineralization activity of the soil microbial community is a crucial factor deciding whether the incorporation of amino acids N occurs intact or mineralized. In summary, the majority of alanine-derived N was taken up by plants after mineralization and less than 1.5% of applied alanine as intact alanine. Thus, intact uptake of amino acids was the least relevant N source, contributing to less than 0.25% to the total N nutrition of the plant. The maximal relevance of amino acid-based N nutrition can be calculated assuming that all 20 proteinogenous amino acids have an uptake similar to alanine (although some of them have much lower concentrations in the soil than alanine). Thus, multiplying the alanine uptake with 20 gives an estimate of the total amino acid uptake. Comparing this with the ammonium and nitrate uptake measured in this study revealed that a maximum of 5% of plant N nutrition can be expected from all amino acids. 2.9.5 Conclusions and Outlook This study emphasizes that position-specific labeling is a novel and unique technique to gain detailed insight into the importance of organic N sources and the uptake of LMWOS by roots from soil. The precision of previous estimates of intact uptake can be strongly enhanced using this new labeling approach without performing timeand costconsuming measurements like compound-specific isotope 13 C/ 15 N analyses. The comparison of N-LMWOS versus N-free LMWOS uptake revealed no significant differences in the 14 C incorporation from these sources. This supported the concept of passive uptake as one of the main uptake mechanisms for LMWOS by plants. Position-specific 14 C labeling revealed that a minor portion of amino acids was taken up intact, whereas the majority of alanine was degraded by soil microorganisms. Some uncertainties remain as plant metabolization like root dark fixation (leading to an overestimation of intact uptake) and plant respiration (leading to an underestimation of intact uptake) cannot be quantified by this approach, too. Mineralized N as well as fragments of the C skeleton was partially available in the soil solution for root uptake. Lupine, as the representative of the legumes in this study, Publications and Manuscripts 289 confirmed the general trend for a greater preference of legumes for organic N sources compared to non-legumes which might be attributed to their ecophysiological capability for amino acid transfer between nodules and roots. Maize, a plant species with fast growth, high N demand and water uptake showed a higher contribution of passive uptake and thus uptake of microbial transformation products ( 14 C-fragment and DIN). In summary, comparing the relevance of DIN and amino acids for each of the investigated plants, irrespective of their ecophysiological specifics, the role of intact amino acid uptake within N nutrition was rather low. Our study suggests N uptake from organic sources is of minor importance for N nutrition of agricultural plants. Nevertheless, the ecophysiological role cannot be fully understood as long as the uptake and allocation mechanisms (passive/active transport, metabolization within the plant) as well as their regulating factors are not identified. Therefore, investigations with a broad spectrum of position-specific labeled LMWOS coupled with CSIA of plant and microbial transformation products are needed. Acknowledgements This study was financed by Deutsche Forschungsgemeinschaft DFG. We thank Ilse Thaufelder and Stefanie Bösel, technical staff at the University of Bayreuth and MartinLuther University of Halle, respectively, C. Cavedon for figure design and C. Werner for her thorough review. Additional peer-reviewed publications A-1 Additional peer-reviewed pubications Tian, Jing, Dippold M, Pausch J, Blagodatskaya E, Fan M, Li X, Kuzyakov Y (2013): Microbial response to rhizodeposition depending on water regimes in paddy rice soils. Soil Biology and Biochemistry: http://dx.doi.org/10.1016/j.soilbio.2013.05.021 Birk, Jago Jonathan, Dippold M, Wiesenberg GLB, Glaser B (2012): Combined quantification of faecal sterols, stanols, stanones and bile acids in soils and terrestrial sediments by gas chromatography-mass spectrometry. Journal of Chromatography A 1242: 1-10 Glaser, Bruno, Benesch M, Dippold M, Zech W. (2012): In situ N-15 and C-13 labelling of indigenous and plantation tree species in a tropical mountain forest (Munessa, Ethiopia) for subsequent litter and soil organic matter turnover studies. Organic Geochemistry 42 (12): 1461-1469 Zech, Wolfgang, Zech R, Zech M, Leiber K, Dippold M, Frechen M, Bussert R, Andreev A (2011): Obliquity forcing of Quaternary glaciation and environmental changes in NE Siberia. Quaternary International 234 (SI): 133-145 Sauheitl, Leopold, Glaser B, Dippold M, Leiber K, Weigelt A (2010): Amino acid fingerprint of a grassland soil reflects changes in plant species richness. Plant and Soil 334 (1-2): 353-363 Acknowledgements A-2 Acknowledgements Ich danke allen, die mir während meiner Doktorarbeit zur Seite standen ganz herzlich. Mein besonderer Dank gilt meinem Doktorvater Prof. Dr. Yakov Kuzyakov: durch einem auf mein Interessengebiet zugeschnittenen DFG Antrag ermöglichte er mir die Promotion in diesem spannenden Forschungsfeld zwischen Biochemie und Bodenkunde. Er lies mir stets den nötigen Freiraum eigene Ideen in allen Ebenen wissenschaftlich Arbeitens zu verwirklichen, was mich außerordentlich motivierte. Auch für die intensive Einführung in die Wissenschaftsgemeinde im Rahmen zahlreicher Tagungsteilnahmen möchte ich mich bei ihm recht herzlich bedanken. Ein außerordentlicher Dank geht an Prof. Dr. Bruno Glaser an dessen Institute ein maßgeblicher Anteil der Analytik erfolgte. Aus den zahlreichen Diskussionen und Gesprächen konnte ich sowohl wissenschaftlich als auch persönlich sehr profitieren. Im speziellen für seine Freundschaft und sein immer offenes Ohr möchte ich mich recht herzlich bedanken. Des Weiteren gilt mein Dank Prof. Huwe, der während meiner Doktorarbeit Laborräume und Gerätschaften zur Fertigstellng meiner Disseration bereitstellte. Ebensogroßer Dank gilt Prof. Dr. Christiane Werner Pinto, die mich sehr herzlich in ihrer Arbeitsgruppe aufnahm und mir in vielen Gesprächen stets hilfreich zur Seite stand. Insbesondere möchte ich Prof. Dr. Wolfgang Zech danken, durch dessen Förderung ich für die Bodenkunde begeistert wurde und auf dessen Unterstützung ich immer zurückgreifen konnte. Der DFG danke ich für die Finanzierung dieses Projekts (DFG KU 1184/19-1) zur Aufklärung positionsspezifischer Transformationen in Böden. Ein besonderes Dankeschön geht an Stefanie Bösel, deren Geschick am Isotopenmassenspektrometer wichtigste Voraussetzung für das Zustandekommen des Umfangreichen Datensatzes war. Die aus den vielen gemeinsamen Stunden am IRMS entstandene Freudschaft hat mir im Rahmen meiner Dissertation sehr geholfen. Auch Ilse Thaufelder, die immer mit Rat und Tat zur Seite stand, hat meine Laborarbeit in Bayreuth sehr erleichtert und bereichert. Spezieller Dank gilt meinen beiden Diplomarbeitsbetreuern Dr. Leopold Sauheitl und Jago Birk, denen ich das Wissen und die analytischen Fähigkeiten zur Durchführung dieser Doktorarbeit verdanke. Neben einer fortlaufend, konstruktiven Zusammenarbeit mit beiden standen sie mir auch als Freunde in jeder Lebenslage zur Seite. Ebenso in- Acknowledgements A-3 tensive Unterstützung habe ich durch PD Dr. Michael Zech erfahren, der sowohl fachlich in vielen Diskussionen zum Gelingen der Dissertation beigetragen hat als auch persönlich mein Leben sehr bereichert hat. Für die konstruktive Zusammenarbeit und ihre stets hilfreiche Freundschaft möchte ich mich außerdem bei Dr. Guido Wiesenberg und Dr. Björn Buggle bedanken. Tiefe Dankbarkeit empfinde ich für meine beiden, treuen Freundinnen Janine Sommer und Katharina Leiber, deren Freundschaft ich mir während all der Jahre unabhängig von allen äußeren Faktoren immer sicher sein konnte. Neben der aktiven Unterstützung in Feld und Labor hat v.a. die Kraft, die ich aus diesen Freundschaften ziehen konnte, das Zustandekommen dieser Doktorarbeit ermöglicht. Meinen herzlichen Dank möchte ich allen Korrekturlesern dieser Arbeit ausprechen, Leopold Sauheitl, Bruno Glaser, Michael Zech, Thomas Friedel, Carolin Apostel und meiner Schwester Christine. Ein in Worte nicht zu fassender Dank geht geht jedoch an meine Familie – meiner Schwester Christine, meinem Bruder Tobias, meinem Vater Klemens, Gerlinde, und meiner Patin Anna, die allen nächtlichen Laborschichten, ständigen Dienstreisen und tagelangen Computersessions zum Trotz mich immer unterstützt haben. Ihre Liebe und Rückendeckung in jeder Lebenslage waren die Grundlage für das Zustandekommen dieser Arbeit. (Eidesstattliche) Versicherungen und Erklärung A-4 (Eidesstattliche) Versicherungen und Erklärungen (§ 5 Nr. 4 PromO) Hiermit erkläre ich, dass keine Tatsachen vorliegen, die mich nach den gesetzlichen Bestimmungen über die Führung akademischer Grade zur Führung eines Doktorgrades unwürdig erscheinen lassen. (§ 8 S. 2 Nr. 5 PromO) Hiermit erkläre ich mich damit einverstanden, dass die elektronische Fassung meiner Dissertation unter Wahrung meiner Urheberrechte und des Datenschutzes einer gesonderten Überprüfung hinsichtlich der eigenständigen Anfertigung der Dissertation unterzogen werden kann. (§ 8 S. 2 Nr. 7 PromO) Hiermit erkläre ich eidesstattlich, dass ich die Dissertation selbstständig verfasst und keine anderen als die von mir angegebenen Quellen und Hilfsmittel benutzt habe. Ich habe die Dissertation nicht bereits zur Erlangung eines akademischen Grades anderweitig eingereicht und habe auch nicht bereits diese oder eine gleichartige Doktorprüfung endgültig nicht bestanden. (§ 8 S. 2 Nr. 9 PromO) Hiermt erkläre ich, dass ich keine Hilfe von gewerblichen Promotionsberatern bzw. -vermittlern in Anspruch genommen habe und auch künftig nicht nehmen werde. Bayreuth, 31.03.2014 Ort, Datum, Unterschrift