Oxygen at Nanomolar Levels Reversibly Suppresses Process Rates and Gene Expression in Anammox and Denitrification in the Oxygen Minimum Zone off Northern Chile

Abstract

A major percentage (20 to 40%) of global marine fixed-nitrogen loss occurs in oxygen minimum zones (OMZs). Concentrations of O[subscript 2] and the sensitivity of the anaerobic N[subscript 2]-producing processes of anammox and denitrification determine where this loss occurs. We studied experimentally how O[subscript 2] at nanomolar levels affects anammox and denitrification rates and the transcription of nitrogen cycle genes in the anoxic OMZ off Chile. Rates of anammox and denitrification were reversibly suppressed, most likely at the enzyme level. Fifty percent inhibition of N[subscript 2] and N[subscript 2]O production by denitrification was achieved at 205 and 297 nM O[subscript 2], respectively, whereas anammox was 50% inhibited at 886 nM O2. Coupled metatranscriptomic analysis revealed that transcripts encoding nitrous oxide reductase (nosZ), nitrite reductase (nirS), and nitric oxide reductase (norB) decreased in relative abundance above 200 nM O[subscript 2]. This O[subscript 2] concentration did not suppress the transcription of other dissimilatory nitrogen cycle genes, including nitrate reductase (narG), hydrazine oxidoreductase (hzo), and nitrite reductase (nirK). However, taxonomic characterization of transcripts suggested inhibition of narG transcription in gammaproteobacteria, whereas the transcription of anammox narG, whose gene product is likely used to oxidatively replenish electrons for carbon fixation, was not inhibited. The taxonomic composition of transcripts differed among denitrification enzymes, suggesting that distinct groups of microorganisms mediate different steps of denitrification. Sulfide addition (1 µM) did not affect anammox or O[subscript 2] inhibition kinetics but strongly stimulated N[subscript 2]O production by denitrification. These results identify new O[subscript 2] thresholds for delimiting marine nitrogen loss and highlight the utility of integrating biogeochemical and metatranscriptomic analyses.