Skip to main content
Dryad logo

Nitrogen isotope fractionation during archaeal ammonia oxidation: coupled estimates from measurements of residual ammonium and accumulated nitrite

Citation

Wanek, Wolfgang; Mooshammer, Maria (2020), Nitrogen isotope fractionation during archaeal ammonia oxidation: coupled estimates from measurements of residual ammonium and accumulated nitrite, Dryad, Dataset, https://doi.org/10.5061/dryad.0gb5mkkz1

Abstract

The naturally occurring nitrogen (N) isotopes, 15N and 14N, exhibit different reaction rates during many microbial N transformation processes, which results in N isotope fractionation. Such isotope effects are critical parameters for interpreting natural stable isotope abundances as proxies for biological process rates in the environment across scales. The kinetic isotope effect of ammonia oxidation (AO) to nitrite (NO2-), performed by ammonia-oxidizing archaea (AOA) and bacteria (AOB), is generally ascribed to the enzyme ammonia monooxygenase (AMO), which catalyzes the first step in this process. However, the kinetic isotope effect of AMO, or εAMO , has been typically determined based on isotope kinetics during product formation (cumulative product, NO2-) alone, which may have overestimated εAMO  due to possible accumulation of chemical intermediates and alternative sinks of ammonia/ammonium (NH3/NH4+). Here, we analyzed 15N isotope fractionation during archaeal ammonia oxidation based on both isotopic changes in residual substrate (RS, NH4+) and cumulative product (CP, NO2-) pools in pure cultures of the soil strain Nitrososphaera viennensis EN76, and in highly enriched cultures of the marine strain Nitrosopumilus adriaticus NF5, under non-limiting substrate conditions. We obtained εAMO  values of 31.9-33.1‰ for both strains based on RS (δ15NH4+), and show that estimates based on CP (δ15NO2-) give larger isotope fractionation factors by 6-8‰. Complementary analyses showed that, at the end of the growth period, microbial biomass was 15N-enriched (10.1‰), whereas nitrous oxide (N2O) was highly 15N depleted (-38.1‰) relative to the initial substrate. Although we did not determine the isotope effect of NH4+ assimilation (biomass formation) and N2O production by AOA, our results nevertheless show that the discrepancy between εAMO  estimates based on RS and CP might have derived from incorporation of 15N-enriched residual NH4+ after AMO reaction into microbial biomass, and that N2O production did not affect isotope fractionation estimates significantly.

Methods

Pure cultures of N. viennensis EN76 were cultivated in freshwater medium and incubated at 37 °C, as described by Tourna et al. (2011). In a first experiment, quadruplicate cultures were supplemented with 1 mM NH4+ and 0.1 mM pyruvate; in a second experiment, quadruplicate cultures were supplemented with 2 mM NH4+ and 0.5 mM pyruvate to generate higher cell biomass and sufficient N2O concentrations for isotopic analysis, in order to determine their potential effect on εAMO . Quadruplicate enrichment cultures of N. adriaticus NF5 were cultivated in SCM medium at 30 °C as described by Bayer et al. (2016). The medium was supplemented with 1 mM NH4+ and 5% (v/v) autoclaved seawater, which was sterile-filtered (0.22 µm GTTP, Millipore). Kanamycin at a final concentration of 100 µg ml-1 was used to inhibit bacterial contaminants. At the time of the experiment (January 2013), the enrichment level of strain NF5 was ~95 %, as it contained a heterotrophic non-nitrifying/non-denitrifying contaminant of the alphaproteobacterial species Oceanicaulis alexandrii (Bayer et al., 2019).

AOA growth was monitored by measuring nitrite production using the Griess method (Hood-Nowotny et al., 2010), coupled to NH4+ consumption determined using the Berthelot method for N. viennensis cultures (Hood-Nowotny et al., 2010) and the o-Phthalaldehyde (OPA) method for N. adriaticus cultures (Goyal et al., 1988). δ15NH4+ was quantified by microdiffusion (Sørensen and Jensen, 1991) with subsequent analysis on a continuous-flow isotope ratio mass spectrometer consisting of an elemental analyzer (EA1110, CE Instruments) coupled via a ConFlo III interface (Finnigan MAT, Thermo Fisher) to the isotope ratio mass spectrometer (IRMS; DeltaPLUS, Finnigan MAT, Thermo Fisher). δ15NO2- was determined based on the reduction of NO2- to N2O by azide under acidified conditions (Lachouani et al., 2010). Concentrations and isotopic ratios of N2O were determined using a purge-and-trap GC/IRMS system (PreCon - GasBench II headspace analyzer, Delta Advantage V IRMS; Thermo Fischer, Vienna, Austria). For NH4+ and NO2- isotope measurements we run blanks, concentration standards and isotope standards varying in natural 15N abundance together with the samples through the full microdiffusion and azide procedures to allow corrections for blank contribution, incomplete reaction and procedural isotope fractionation (Lachouani et al., 2010). Nitrogen contents and δ15N signatures of AOA biomass were determined by EA-IRMS as described above. δ15N signatures [‰ vs. AIR] were calculated relative to the ratio R (15N:14N) of the atmospheric N2 standard (AIR), as  d15N = (Rsample/Rstandard - 1) x 1000.

Isotope fractionation factors (e) were calculated based on the Rayleigh closed system isotope fractionation, based on changes in the isotopic compositions of residual substrate (RS; i.e., NH4+) and cumulative product (CP; i.e., NO2-) (Mariotti et al., 1981):

103ln10-3δRS+110-3δS0+1=ε ln(f)             (1)

δCP-δS0=-ε fln⁡(f)(1-f) ,                (2)

where δS0  is δ15N of initial NH4+, δRS  is δ15NH4+, δCP  is δ15NO2- and f  is the fraction of the initial [NH4+] remaining in the culture. Plots of 10310-3δRS+110-3δS0+1  versus lnf  and of δCP-δS0  versus fln⁡(f)(1-f)  yield linear relations, with the slope representing the kinetic isotope effect based on the isotopic change in substrate (εRS)  and product (εAP) , respectively. Uncertainties of ε are expressed as standard error of the slope. Differences in isotope fractionation effects between cultures were assessed by testing significant differences between their regression plots, using R (R Development Core Team, 2012).

Funding

Austrian Science Fund, Award: P28037-B22